SLEEP

Proserpio,, Paola;Loddo,, Giuseppe;Zubler,, Frederic;Ferini-Strambi,, Luigi;Licchetta,, Laura;Bisulli,, Francesca;Tinuper,, Paolo;Agostoni, Elio, Clemente;Bassetti,, Claudio;Tassi,, Laura;Menghi,, Veronica;Provini,, Federica;Nobili,, Lino

doi: 10.1093/sleep/zsz166pmid: 31609388

Abstract Objective The differential diagnosis between sleep-related hypermotor epilepsy (SHE) and disorders of arousal (DOA) may be challenging. We analyzed the stage and the relative time of occurrence of parasomnic and epileptic events to test their potential diagnostic accuracy as criteria to discriminate SHE from DOA. Methods Video-polysomnography recordings of 89 patients with a definite diagnosis of DOA (59) or SHE (30) were reviewed to define major or minor events and to analyze their stage and relative time of occurrence. The “event distribution index” was defined on the basis of the occurrence of events during the first versus the second part of sleep period time. A group analysis was performed between DOA and SHE patients to identify candidate predictors and to quantify their discriminative performance. Results The total number of motor events (i.e. major and minor) was significantly lower in DOA (3.2 ± 2.4) than in SHE patients (6.9 ± 8.3; p = 0.03). Episodes occurred mostly during N3 and N2 in DOA and SHE patients, respectively. The occurrence of at least one major event outside N3 was highly suggestive for SHE (p = 2*e-13; accuracy = 0.898, sensitivity = 0.793, specificity = 0.949). The occurrence of at least one minor event during N3 was highly suggestive for DOA (p = 4*e-5; accuracy = 0.73, sensitivity = 0.733, specificity = 0.723). The “event distribution index” was statistically higher in DOA for total (p = 0.012) and major events (p = 0.0026). Conclusion The stage and the relative time of occurrence of minor and major motor manifestations represent useful criteria to discriminate DOA from SHE episodes. parasomnia, sleep-related hypermotor epilepsy, disorder of arousal Statement of Significance In our retrospective analysis of video-polysomnographic recording of patients with sleep-related hypermotor epilepsy (SHE) and disorder of arousals (DOA), we found that episodes occurred mostly during N3 and N2 in DOA and SHE patients, respectively. Specifically, the occurrence of at least one major event outside N3 was highly suggestive for SHE, while the occurrence of at least one minor event during N3 was highly suggestive for DOA. The stage and the relative time of occurrence of motor events have been considered from a long time as discriminant features between epilepsy and DOA. However, this represents the first study that quantified this traditional knowledge by systematic comparison and by means of statistical evaluation, confirming that its important value for differential diagnosis. Introduction Disorders of arousal (DOA) are parasomnic events, occurring during non-rapid eye movement (NREM) sleep, characterized by complex, seemingly purposeful, goal-directed behaviors [1]. Sleep-related hypermotor epilepsy (SHE), previously known as nocturnal frontal lobe epilepsy, is an epileptic syndrome with seizures characterized by variable duration and complexity occurring mostly during sleep, particularly in NREM sleep [2]. Ictal manifestations can consist in sustained dystonic posturing but also in complex, hyperkinetic, often bizarre, motor patterns, sometimes associated with affective symptoms or ambulatory behaviors [2]. Especially in these latter cases, the differential diagnosis with DOA can be challenging [3]. Different anamnestic key features have been recognized as discriminant in the differential diagnosis between SHE and DOA [4, 5] and, on their basis, clinical scales have been developed [6, 7], though limited by contradictory diagnostic accuracy [8]. To date, video-polysomnography (vPSG) represents the “gold-standard” test for diagnosing paroxysmal sleep-related events [9]. Considering that interictal and even ictal scalp EEG can be uninformative in many SHE patients [2, 10–12], a careful semiological analysis of manifestations is considered the most useful assessment to differentiate epileptic versus parasomnic events. However, some clinical characteristic (such as screaming or ambulation) may not allow making a definite diagnosis and brief arousals of epileptic and non-epileptic origin are frequently indistinguishable [13]. Moreover, the analysis of clinical patterns implies a part of subjectivity and depends on the clinician experience [14]. Therefore, objective PSG markers may help to increase the diagnostic accuracy of vPSG recordings. The stage from which nocturnal events emerge and event timing relative to sleep onset have been considered from a long time as discriminant features between epilepsy and DOA. Indeed, it is commonly stated that DOA arise more frequently from deep NREM sleep, usually in the first third of the night [1], while sleep-related seizures generally occur during light NREM sleep or during sleep–wake transitions, anytime during the sleep period [5]. However, systematic studies analyzing and comparing these PSG features in patients suffering from DOA and epileptic patients are scanty and their results are discordant. Early studies, focused only on SHE, found a predominance of sleep-related seizures during N2 in any time of the night [10, 15], in accordance with what observed in other epilepsy syndrome (not only sleep-related) [16]. At odds, a recent work showed that the majority of seizures emerged from N3, especially during the first sleep cycles [17]. With respect to DOA, results seem to be more homogenous. In 1965, Jacobson et al. demonstrated that most DOA episodes arose during the first 3 hours of the night, from sudden but incomplete arousal out of slow-wave sleep [18]; analogously, in 1995, Zucconi et al. found that 86% of DOA events occurred during N3 and 66% in the first third of sleep [19]. Finally, a more recent study based on vPSG analysis of sleep-related paroxysmal events showed that NREM parasomnic episodes arose exclusively during deep sleep, while sleep-related seizures emerged more frequently during light NREM sleep [13]. The aim of this study was to analyze the stage and the relative time of occurrence of minor and major DOA and epileptic events to test their potential diagnostic accuracy as criteria to discriminate DOA from sleep-related seizures. Methods Patients Patients were retrospectively recruited from the Sleep Medicine Center, “C. Munari” Center of Epilepsy Surgery, Niguarda Hospital, Milan and from the Department of Biomedical and Neuromotor Sciences, University of Bologna, Bologna. All patients gave written consent for the use of recordings for research and publications purposes. All vPSG recordings conducted for sleep-related paroxysmal events were blindly reviewed to select patients (>15 years) with the following criteria: full PSG with video recording available for all night duration; at least one major or minor habitual episode recorded; definite final diagnosis of SHE or DOA; and absence of other sleep disorders (i.e. sleep breathing disorders or periodic leg movements). Specifically, we included only epileptic patients with a confirmed diagnosis of SHE, according to the current criteria [2]. Patients with DOA were selected only if: they fulfill the current diagnostic criteria [1]; and there was a diagnostic consensus agreement by all clinicians involved in the diagnosis and management on the basis of anamnestic features, vPSG data, home-video recordings, clinical follow-up, and therapeutical responses (avoiding triggering factors and/or drugs). Patients with an overt comorbidity of epilepsy and DOA were excluded. Moreover, we excluded epileptic patients with a strong family history and a personal history positive for parasomnia before the age of seizure onset. Video-EEG monitoring and analysis All patients underwent one-night vPSG recording in the sleep laboratories. vPSG included electroencephalogram (international 10–20 system), electrocardiogram, electro-oculogram, chin and limb EMG, nasal airflow, two channels of breathing effort, and oximetry. All recordings were reviewed and scored manually according to the AASM criteria [20] by two sleep medicine experts (P.P. and G.L.). Participants with an apnea hypopnea index (AHI) >5 and/or periodic leg movements (PLM) index >15 were not included. All the video recordings of the sleep-related paroxysmal episodes were reviewed by experts in epilepsy and sleep medicine to define major or minor events. In particular, on the basis of current DOA classification by Loddo et al. [21], we defined simple arousal movements (pattern I) and rising arousal movements (pattern II) as minor DOA events and complex arousal ambulatory movements (pattern III) as major DOA events. In SHE patients, minor motor events [22–24] and paroxysmal arousal [2, 10, 22] were considered as minor events, while complex hypermotor seizures [2, 22] were defined as major events. For every major or minor event, we analyzed the stage and the relative time (first or second half of sleep period time [SPT]) of occurrence. In particular, we calculate the “events distribution index” based on the occurrence during the first versus second part of SPT ([number of events during first – number of events during second part of the SPT]/total number of events). Moreover, we investigated the distribution of N3 in the course of the night calculating the “N3 distribution index” defined as ([number of epochs of N3 in the first half of SPT – number of epochs of N3 in the second half of the SPT]/total number of N3 epochs). The N3 distribution index was also extracted from a population of 22 control subjects, matched for age and sex with patients. Statistical analysis First, group analysis was performed between subjects with DOA or SHE in order to identify candidate predictors. A Mann–Whitney U-test with ties correction was used for continuous or discrete (non-binary) variables between two groups. A chi-squared test was used for binary categorical variables. Differences between three groups were assessed with a Kruskal–Wallis test, with post hoc comparison using Tukey’s criterion. Statistical significance was set at p = 0.05. Second, the discriminative performance of specific continuous or discrete predictors was quantified with the area under the receiver operating characteristic (ROC) curve (AUC). In cases where a “canonical” cutoff value was also statistically relevant (e.g. high predictive value of presence vs. absence of an event during a particular sleep stage), we used it as operating point, and reported the accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for SHE. Finally, we examined the performance of combinations of predictors by using a binary classifier based on multivariate logistic regression. To avoid overfitting, we used 10-fold cross-validation. The performance was assessed with accuracy; for combinations with equal maximal accuracy, we computed the PPV and NPV for SHE. All analyses were performed in Matlab 2017a (Statistics and Machine Learning Toolbox). Results Demographics Eighty-nine patients were included, of which 38 (43%) were females. The mean (±SD) age was 29.4 ± 11.5 years (range: 14–66). About two-third of patients had DOA (59/89 = 66.3 %; 42% female; mean age 28.4 ± 10.1) and one third had SHE (30/89 = 33.7%; 47% female; mean age 30.4 ± 14.4) (Table 1). The control group consisted in 22 healthy drug-free subjects matched for age and sex (15 males, mean age 27, 1 year, range 19–40) recruited through advertisements, with no current or past history of DOA, epilepsy, or other neurological, sleep, or psychiatric disorders. Table 1. Classifier performance . DOA . SHE . p value . (N = 59) (N = 30) Sex (female) 25 (42%) 14 (47%) 0.7 Age (year) 28.4 ± 10.1 30.4 ± 14.4 0.81 Nb total events 3.2 ± 2.4 6.9 ± 8.2 0.03 Nb major events 1.15 ± 1.30 4.47 ± 5.20 1.25E-06 Nb minor events 2.05 ± 2.27 2.43 ± 5.81 0.07 Nb total events during N3 2.92 ± 2.21 1.90 ± 2.41 0.0036 Nb total events outside N3 0.29 ± 0.83 5.00 ± 7.43 8.56E-11 Nb major events during N3 1.10 ± 1.26 1.37 ± 1.96 0.91 Nb major events outside N3 0.05 ± 0.22 3.10 ± 4.89 2.19E-13 Nb minor events during N3 1.81 ± 1.94 0.53 ± 1.25 4.62E-05 Nb minor events outside N3 0.24 ± 0.82 1.90 ± 4.71 0.0023 Total events distribution index 0.50 ± 0.66 0.14 ± 0.71 0.0119 Major events distribution index 0.55 ± 0.77a -0.01 ± 0.78 0.0026 Minor events distribution index 0.36 ± 0.67 0.25 ± 0.52 0.2015 N3 distribution index 0.38 ± 0.24 0.17 ± 0.37 0.0048 . DOA . SHE . p value . (N = 59) (N = 30) Sex (female) 25 (42%) 14 (47%) 0.7 Age (year) 28.4 ± 10.1 30.4 ± 14.4 0.81 Nb total events 3.2 ± 2.4 6.9 ± 8.2 0.03 Nb major events 1.15 ± 1.30 4.47 ± 5.20 1.25E-06 Nb minor events 2.05 ± 2.27 2.43 ± 5.81 0.07 Nb total events during N3 2.92 ± 2.21 1.90 ± 2.41 0.0036 Nb total events outside N3 0.29 ± 0.83 5.00 ± 7.43 8.56E-11 Nb major events during N3 1.10 ± 1.26 1.37 ± 1.96 0.91 Nb major events outside N3 0.05 ± 0.22 3.10 ± 4.89 2.19E-13 Nb minor events during N3 1.81 ± 1.94 0.53 ± 1.25 4.62E-05 Nb minor events outside N3 0.24 ± 0.82 1.90 ± 4.71 0.0023 Total events distribution index 0.50 ± 0.66 0.14 ± 0.71 0.0119 Major events distribution index 0.55 ± 0.77a -0.01 ± 0.78 0.0026 Minor events distribution index 0.36 ± 0.67 0.25 ± 0.52 0.2015 N3 distribution index 0.38 ± 0.24 0.17 ± 0.37 0.0048 Patient demographics and differences between subjects with DOA and SHE. aOnly patients with at least one major event. Open in new tab Table 1. Classifier performance . DOA . SHE . p value . (N = 59) (N = 30) Sex (female) 25 (42%) 14 (47%) 0.7 Age (year) 28.4 ± 10.1 30.4 ± 14.4 0.81 Nb total events 3.2 ± 2.4 6.9 ± 8.2 0.03 Nb major events 1.15 ± 1.30 4.47 ± 5.20 1.25E-06 Nb minor events 2.05 ± 2.27 2.43 ± 5.81 0.07 Nb total events during N3 2.92 ± 2.21 1.90 ± 2.41 0.0036 Nb total events outside N3 0.29 ± 0.83 5.00 ± 7.43 8.56E-11 Nb major events during N3 1.10 ± 1.26 1.37 ± 1.96 0.91 Nb major events outside N3 0.05 ± 0.22 3.10 ± 4.89 2.19E-13 Nb minor events during N3 1.81 ± 1.94 0.53 ± 1.25 4.62E-05 Nb minor events outside N3 0.24 ± 0.82 1.90 ± 4.71 0.0023 Total events distribution index 0.50 ± 0.66 0.14 ± 0.71 0.0119 Major events distribution index 0.55 ± 0.77a -0.01 ± 0.78 0.0026 Minor events distribution index 0.36 ± 0.67 0.25 ± 0.52 0.2015 N3 distribution index 0.38 ± 0.24 0.17 ± 0.37 0.0048 . DOA . SHE . p value . (N = 59) (N = 30) Sex (female) 25 (42%) 14 (47%) 0.7 Age (year) 28.4 ± 10.1 30.4 ± 14.4 0.81 Nb total events 3.2 ± 2.4 6.9 ± 8.2 0.03 Nb major events 1.15 ± 1.30 4.47 ± 5.20 1.25E-06 Nb minor events 2.05 ± 2.27 2.43 ± 5.81 0.07 Nb total events during N3 2.92 ± 2.21 1.90 ± 2.41 0.0036 Nb total events outside N3 0.29 ± 0.83 5.00 ± 7.43 8.56E-11 Nb major events during N3 1.10 ± 1.26 1.37 ± 1.96 0.91 Nb major events outside N3 0.05 ± 0.22 3.10 ± 4.89 2.19E-13 Nb minor events during N3 1.81 ± 1.94 0.53 ± 1.25 4.62E-05 Nb minor events outside N3 0.24 ± 0.82 1.90 ± 4.71 0.0023 Total events distribution index 0.50 ± 0.66 0.14 ± 0.71 0.0119 Major events distribution index 0.55 ± 0.77a -0.01 ± 0.78 0.0026 Minor events distribution index 0.36 ± 0.67 0.25 ± 0.52 0.2015 N3 distribution index 0.38 ± 0.24 0.17 ± 0.37 0.0048 Patient demographics and differences between subjects with DOA and SHE. aOnly patients with at least one major event. Open in new tab At the moment of PSG recordings, 51/59 patients with DOA and 6/30 patients with SHE were not receiving any pharmacological treatment (see Supplementary Table S1 for drug treatments in the others patients). Eleven SHE patients were pharmaco-resistant. Two patients with DOA were under antiepileptic treatment as they were referred with a misdiagnosis of drug-resistant SHE. None of DOA patients had histories involving appetitive (sexual, feeding) behaviors. In total, 394 events were identified (range: 1–34 events/patient), from which 198 (50.2%) were “major events.” Of note, all patients with SHE had at least one major events, whereas 24 (40.7%) patients with DOA had no major events (p = 5.6 e-05). No difference in age (p = 0.81) or gender (p = 0.70) was observed between patients with parasomnia and SHE (Table 1). Group analysis Number of events. Significantly less total (i.e. major + minor) events were observed in patients with DOA (3.2 ± 2.4) than in patients with SHE (6.9 ± 8.3; p = 0.03). The number of major events was also significantly lower in case of DOA (1.15 ± 1.30 vs. 4.47 ± 5.20; p =1 .2*e-06). The difference remained significant when we only considered subjects with at least one major event (p = 0.017). By contrast, the difference in number of minor events was not significantly different (DOA = 2.15 ± 2.7, SHE = 2.56 ± 5.9; p = 0.07) (Table 1). Stage of occurrence. Events were not distributed equally between all sleep stages (Figure 1). In particular, minor and major episodes occurred mostly during N3 and N2 in DOA and SHE patients, respectively. Statistical analysis revealed that the total number of events occurring during N3 was slightly higher in case of DOA (2.9 ± 2.2 vs. 1.9 ± 2.4, p = 0.0036), whereas the events outside of N3, that is during N1, N2, or REM, were clearly higher in patients with epilepsy (DOA 0.29 ± 0.8, SHE 5.0 ± 7.4, p = 8*e-11). When considering major events only, the highest between-group difference was the occurrence outside of N3, which was very rare in DOA (0.05 ± 0.22), and therefore very specific for SHE (3.10 ± 4.89; p = 2*e-13). For minor events, by contrast, the highest difference between groups was observed during N3, and higher in case of DOA (DOA 1.81 ± 1.94, SHE 0.53 ± 1.25, p = 4*e-5) (Table 1). Figure 1. Open in new tabDownload slide Distribution of events according to sleep stages. Minor and major episodes occurred mostly during N3 and N2 in DOA and SHE patients, respectively. Figure 1. Open in new tabDownload slide Distribution of events according to sleep stages. Minor and major episodes occurred mostly during N3 and N2 in DOA and SHE patients, respectively. Events distribution index. Episodes of DOA tended to occur earlier in the course of the night than epileptic seizures. The “events distribution index” was statistically higher in DOA for total events (DOA 0.5 ± 0.66, SHE 0.14 ± 0.71, p = 0.012) and for major events (DOA 0.55 ± 0.77, SHE −0.01 ± 0.78, p = 0.0026, after exclusion of patients without major events); no significant difference was observed concerning the distribution of minor events (DOA 0.36 ± 0.67, SHE 0.25 ± 0.52, p = 0.21) (Table 1). N3 distribution index. The “N3 distribution index” was significantly different between the three groups (p = 0.0021) Figure 2. Post hoc analysis showed that the index was higher (meaning N3 was predominantly distributed during the first part of the night) in patients with DOA (0.38 ± 0.24) than in patients with SHE (0.17 ± 0.37, p = 0.016). Both patients with DOA (p = 0.042) and patients with SHE (p = 0.003) had a significantly lower index with respect to control subjects (0.49 ± 0.21). Figure 2. Open in new tabDownload slide N3 distribution index. Comparison of “N3 distribution index” between patients with DOA, patients with SHE and controls. Figure 2. Open in new tabDownload slide N3 distribution index. Comparison of “N3 distribution index” between patients with DOA, patients with SHE and controls. Performance of single predictors Based on the group analysis, the discriminative performance of five different variables was quantified (termed “predictors” in the following). The two first predictors had higher values in patients with SHE, namely predictor 1, the number of major events outside of N3 (AUC = 0.89), and predictor 2, the number of major events (AUC = 0.81) (Figure 3left). By contrast, the other predictors had higher values in patients with DOA (Figure 3right): predictor 3, the number of minor events during N3 (AUC = 0.75); predictor 4, the total events distribution index (AUC = 0.656); and predictor 5, the N3 distribution index (AUC = 0.684). Figure 3. Open in new tabDownload slide Performance of single predictors. Performance of single predictors assessed using the receiver operator characteristic (ROC). Left: predictors with higher value in SHE; right: predictors with higher value in DOA. Figure 3. Open in new tabDownload slide Performance of single predictors. Performance of single predictors assessed using the receiver operator characteristic (ROC). Left: predictors with higher value in SHE; right: predictors with higher value in DOA. For predictors 1 and 3, setting the cutoff at one led to a good predictive performance for SHE and for DOA, respectively. The occurrence of at least one major event outside N3 was suggestive for SHE (accuracy = 0.898, sensitivity = 0.793, specificity = 0.949, PPV = 0.885, NPV = 0.903). The occurrence of at least one minor event during N3 was suggestive for DOA (accuracy = 0.73, sensitivity = 0.733, specificity = 0.723, PPV = 0.843, NPV = 0.579). Performance of multivariate models The five predictors mentioned in the previous subsection were used in various combinations with a binary multivariate classifier. For each of the 5^2-1 = 31 different combinations of predictors, the performance of the classifier was evaluated with the accuracy (Supplementary Table S2). The maximum accuracy reached was 89.9%. It was obtained by 11 different combinations, all of which contained the first predictor (number of major events outside N3). Of note, also the “combination” consisting of this predictor alone reached the maximum accuracy. Furthermore, 10/11 combinations on first rank had the same predictive values (90% for DOA, 89% for SHE), whereas the last one had a predictive value of 89% for DOA and 92% for SHE. Discussion Our study confirms that the stage and the relative time of occurrence during sleep of minor and major episodes represent a useful information to discriminate DOA from SHE manifestations. The analysis of ictal semiology of vPSG-recorded events is currently considered the gold standard for the differential diagnosis in challenging cases. However, potential diagnostic pitfalls have been identified in the assessment of these sleep-related events [13]. Moreover, this semiological approach is partly subjective and highly dependent from clinician expertise and skills [14]. Thus, the analysis of event distribution in relation to sleep stages and relative time of occurrence during sleep could represent a further, simple criteria, easily usable also by less trained physician. However, we believe that this pragmatic tool could be helpful in order to facilitate the differentiation between DOA and SHE only if integrated with all the other well-known anamnestic and clinical features usually adopted for differential diagnosis. Our main results show that both minor and major episodes arise mostly from N3 and N2 in DOA and SHE patients, respectively, and that DOA manifestations tend to occur earlier in the course of the sleep period than sleep-related seizures. Major manifestations during N3, especially in the first third of the night, have always been considered one of the main characteristics able to differentiate DOA from sleep-related seizures. The latter typically arise from light sleep anytime during the night. So that, this aspect has been also included in the current diagnostic criteria for DOA [1]. However, this is the first study focused on the systematic comparison of the stage/time of occurrence of minor and major DOA and SHE episodes. Our findings are consistent with previous studies conducted only on parasomnic populations [18, 19]. On the other hand, studies analyzing SHE populations have reported discordant results. Specifically, a recent investigation aimed at describing the polysomnographic features and the distribution of epileptic motor events in 40 SHE patients found that most seizures occurred in deep sleep with a predominance in the first sleep cycle [17]. However, to our knowledge, this represent the only study with the majority of seizure occurring in N3, in contrast with both our results and those from all the other studies on large samples of SHE patients [10, 15]. Different studies have shown that the differential diagnosis between DOA and SHE, only based on vPSG analysis of minor motor events, is challenging or impossible, even when experts are involved [13, 14, 23, 25]. Indeed, the current criteria for diagnostic certainty of SHE stated that if the captured episodes are minor motor events or paroxysmal arousals, the clinical diagnosis may be unreliable [2]. On this basis, we decided to differentiate minor and major episodes in both DOA and SHE patients and we found that minor events occurred more frequently during N3 in DOA patients with respect to SHE, whereas the distribution of minor events during the sleep period was similar in the two groups of patients. Therefore, although the semiological features of minor episodes seems to be only partially useful for the differential diagnosis, the occurrence of at least one minor event during N3 seems to be highly suggestive for DOA. Conversely, our statistical analysis revealed that the occurrence of a major event outside N3 had the maximum accuracy, being highly suggestive for SHE. Finally, clinicians could be aware of the possibility of recording a mixture of epileptic and parasomnic minor events in the same subject. Considering the amount of recorded sleep-related manifestations, we found that significantly less total events were observed in patients with DOA than in patients with SHE. This result could be influenced by our specific SHE population, that included a relative high proportion (one third) of drug-resistant patients. However, also some untreated epileptic patients evaluated for a first diagnosis showed a high amount of sleep-related manifestations. Another potential explanation is that DOA events are much more difficult to record in the sleep laboratory, especially in the absence of a specific protocol [26]. Moreover, parasomnic manifestations during vPSG are generally less intense and complex with respect to the variety and number of behaviors reported during the anamnestic evaluation or documented by home-video recordings [13, 27, 28]. Finally, our population did not included DOA patients with appetitive manifestations (i.e. sexsomnia) which seem to have often higher- even nightly- frequency of events, similar to that of epileptic patients [29]. We found that DOA patients showed a lower “N3 distribution index” with respect to the control group, suggesting a less evident predominance of slow-wave sleep during the first part of the night. Abnormalities in slow-wave activity (SWA, 1–4 Hz) power in DOA have been described in some studies. Indeed, Gaudreau et al. [30] found that sleepwalkers compared with control subjects had a significantly lower level of SWA during the first NREM period, where most awakenings take place. Moreover, a lower SWA decline overnight has been described in patients with DOA than in control subjects [27, 31]. A recent study with high-density EEG revealed that the decrease in SWA power in DOA patients is not global, but localized mainly to motor, premotor, and cingulate areas [32]. By analyzing the sleep structure of SHE patients, some authors found that the sleep macrostructure is similar in control and SHE subjects [15, 24]. However, significant variations in the sleep microstructure, expressed by an increase of arousal-related phasic events, have been reported [33]. To date, no data are available regarding the dynamic of SWA in SHE patients. We found that the “N3 distribution index” is lower in SHE with respect to DOA patients. We can hypothesize that the high amount of epileptic attacks occurring particularly during light sleep could lead to a sleep fragmentation thus preventing to reach a stable deep sleep and interfering with the homeostatic decay of SWA. Finally, it is worth to remind that also the chronic use of antiepileptic drugs can affect sleep architecture [34]. The main limitations of our work are the retrospective design of the study and the relative small sample of patients. Indeed, the low number of subjects did not allow us to correct the results for possible confounding factors, such as drugs and comorbidities. On the other hand, one of the main strength point of our work is the very stringent exclusion and inclusion criteria, especially for DOA patients, in order to avoid misdiagnosis. Moreover, considering that different studies demonstrated a frequent comorbity between SHE and DOA [10, 35], we cannot completely exclude that, in some of our included patients, the “major events” were clearly related to SHE, but some “minor events” produced by DOA (or vice versa). However, different studies showed that in the majority of patients with this comorbidity, a time gap—usually long—separated the last parasomnic manifestations from the onset of the first SHE episode [10, 35]. Anyway, to try to reduce this bias, we excluded epileptic patients with a strong family history and a personal history positive for parasomnia before the age of seizure onset. Our results should be confirmed in further prospective studies in a larger population, including also pediatric cases. Conflict of interest statement. None declared. References 1. American Academy of Sleep Medicine. International Classification of sleep disorders-third edition (ICSD-3) . Darien, Illinois ; 2014 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 2. Tinuper P , et al. Definition and diagnostic criteria of sleep-related hypermotor epilepsy . Neurology . 2016 ; 86 ( 19 ): 1834 – 1842 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Tinuper P , et al. From nocturnal frontal lobe epilepsy to sleep-related hypermotor epilepsy: a 35-year diagnostic challenge . Seizure . 2017 ; 44 : 87 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Derry CP , et al. Paroxysmal motor disorders of sleep: the clinical spectrum and differentiation from epilepsy . Epilepsia . 2006 ; 47 ( 11 ): 1775 – 1791 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Tinuper P , et al. Movement disorders in sleep: guidelines for differentiating epileptic from non-epileptic motor phenomena arising from sleep . Sleep Med Rev . 2007 ; 11 ( 4 ): 255 – 267 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Derry CP , et al. Distinguishing sleep disorders from seizures: diagnosing bumps in the night . Arch Neurol . 2006 ; 63 ( 5 ): 705 – 709 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Bisulli F , et al. Diagnostic accuracy of a Structured Interview for Nocturnal Frontal Lobe Epilepsy (SINFLE): a proposal for developing diagnostic criteria . Sleep Med . 2012 ; 13 ( 1 ): 81 – 87 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Manni R , et al. The FLEP scale in diagnosing nocturnal frontal lobe epilepsy, NREM and REM parasomnias: data from a tertiary sleep and epilepsy unit . Epilepsia . 2008 ; 49 ( 9 ): 1581 – 1585 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Aurora RN , et al. Practice parameters for the non-respiratory indications for polysomnography and multiple sleep latency testing for children . Sleep . 2012 ; 35 ( 11 ): 1467 – 1473 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Provini F , et al. Nocturnal frontal lobe epilepsy. A clinical and polygraphic overview of 100 consecutive cases . Brain . 1999 ; 122 ( Pt 6 ): 1017 – 1031 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Nobili L , et al. Nocturnal frontal lobe epilepsy: intracerebral recordings of paroxysmal motor attacks with increasing complexity . Sleep . 2003 ; 26 ( 7 ): 883 – 886 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 12. Gibbs SA , Figorilli M , Casaceli G , Proserpio P , Nobili L . Sleep Related Hypermotor Seizures with a Right Parietal Onset . J Clin Sleep Med . 2015 ; 11 : 953 – 955 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 13. Derry CP , et al. NREM arousal parasomnias and their distinction from nocturnal frontal lobe epilepsy: a video EEG analysis . Sleep . 2009 ; 32 ( 12 ): 1637 – 1644 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Vignatelli L , et al. Interobserver reliability of video recording in the diagnosis of nocturnal frontal lobe seizures . Epilepsia . 2007 ; 48 ( 8 ): 1506 – 1511 . Google Scholar Crossref Search ADS PubMed WorldCat 15. Oldani A , et al. Autosomal dominant nocturnal frontal lobe epilepsy. A video-polysomnographic and genetic appraisal of 40 patients and delineation of the epileptic syndrome . Brain . 1998 ; 121 ( Pt 2 ): 205 – 223 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Ng M , et al. Why are seizures rare in rapid eye movement sleep? Review of the frequency of seizures in different sleep stages . Epilepsy Res Treat . 2013 ; 2013 : 932790 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 17. Parrino L , et al. Distinctive polysomnographic traits in nocturnal frontal lobe epilepsy . Epilepsia . 2012 ; 53 ( 7 ): 1178 – 1184 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Jacobson A , et al. Somnambulism: all-night electroencephalographic studies . Science . 1965 ; 148 ( 3672 ): 975 – 977 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Zucconi M , et al. Arousal fluctuations in non-rapid eye movement parasomnias: the role of cyclic alternating pattern as a measure of sleep instability . J Clin Neurophysiol . 1995 ; 12 ( 2 ): 147 – 154 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Berry R , Brooks R , Gamaldo CE , et al. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications, Version 2.2 . Darien, IL : American Academy of Sleep Medicine ; 2015 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 21. Loddo G , et al. Specific motor patterns of arousal disorders in adults: a video-polysomnographic analysis of 184 episodes . Sleep Med . 2018 ; 41 : 102 – 109 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Nobili L , et al. Surgical treatment of drug-resistant nocturnal frontal lobe epilepsy . Brain . 2007 ; 130 ( Pt 2 ): 561 – 573 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Terzaghi M , et al. Sleep-related minor motor events in nocturnal frontal lobe epilepsy . Epilepsia . 2007 ; 48 ( 2 ): 335 – 341 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Oldani A , et al. Autosomal dominant nocturnal frontal lobe epilepsy: electroclinical picture . Epilepsia . 1996 ; 37 ( 10 ): 964 – 976 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Nobili L , et al. Intracerebral recordings of minor motor events, paroxysmal arousals and major seizures in nocturnal frontal lobe epilepsy . Neurol Sci . 2005 ; 26 ( Suppl. 3 ): s215 – s219 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Pilon M , et al. Precipitating factors of somnambulism: impact of sleep deprivation and forced arousals . Neurology . 2008 ; 70 ( 24 ): 2284 – 2290 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Nobili L . Can homemade video recording become more than a screening tool? Sleep . 2009 ; 32 ( 12 ): 1544 – 1545 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Mwenge B , et al. Sleepwalking: long-term home video monitoring . Sleep Med . 2013 ; 14 ( 11 ): 1226 – 1228 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Voges BR , Schmitt FC , House PM , Stodieck SR , Schenck CH . Complex sexual behaviors during sleep as a manifestation of epilepsy: a case series . Sleep . 2019 Mar 1; 42 ( 3 ). OpenURL Placeholder Text WorldCat 30. Gaudreau H , et al. Dynamics of slow-wave activity during the NREM sleep of sleepwalkers and control subjects . Sleep . 2000 ; 23 ( 6 ): 755 – 760 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Guilleminault C , et al. Sleep and wakefulness in somnambulism: a spectral analysis study . J Psychosom Res . 2001 ; 51 ( 2 ): 411 – 416 . Google Scholar Crossref Search ADS PubMed WorldCat 32. Castelnovo A , et al. Scalp and source power topography in sleepwalking and sleep terrors: a high-density EEG study . Sleep . 2016 ; 39 ( 10 ): 1815 – 1825 . Google Scholar Crossref Search ADS PubMed WorldCat 33. Zucconi M , et al. The macrostructure and microstructure of sleep in patients with autosomal dominant nocturnal frontal lobe epilepsy . J Clin Neurophysiol . 2000 ; 17 ( 1 ): 77 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Jain SV , Glauser TA . Effects of epilepsy treatments on sleep architecture and daytime sleepiness: an evidence-based review of objective sleep metrics . Epilepsia . 2014 ; 55 ( 1 ): 26 – 37 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Bisulli F , et al. Increased frequency of arousal parasomnias in families with nocturnal frontal lobe epilepsy: a common mechanism? Arousal parasomnias in NFLE families . Epilepsia . 2010 ; 51 ( 9 ): 1852 – 1860 . Google Scholar Crossref Search ADS PubMed WorldCat © Sleep Research Society 2019. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected] This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Malekshahi,, Azim;Chaudhary,, Ujwal;Jaramillo-Gonzalez,, Andres;Lucas Luna,, Alberto;Rana,, Aygul;Tonin,, Alessandro;Birbaumer,, Niels;Gais,, Steffen

doi: 10.1093/sleep/zsz185pmid: 31665518

Abstract Persons in the completely locked-in state (CLIS) suffering from amyotrophic lateral sclerosis (ALS) are deprived of many zeitgebers of the circadian rhythm: While cognitively intact, they are completely paralyzed, eyes mostly closed, with artificial ventilation and artificial nutrition, and social communication extremely restricted or absent. Polysomnographic recordings in eight patients in CLIS, however, revealed the presence of regular episodes of deep sleep during night time in all patients. It was also possible to distinguish an alpha-like state and a wake-like state. Classification of rapid eye movement (REM) sleep is difficult because of absent eye movements and absent muscular activity. Four out of eight patients did not show any sleep spindles. Those who have spindles also show K-complexes and thus regular phases of sleep stage 2. Thus, despite some irregularities, we found a surprisingly healthy sleep pattern in these patients. movement disorders, neurological disorders, sleep/wake physiology, circadian rhythms, completely locked-in state, polysomnography Statement of Significance The presence of circadian variation in EEG activity points to a conserved sleep-wake cycle in amyotrophic lateral sclerosis (ALS) patients in completely locked-in state (CLIS). There are marked differences between these patients and healthy participants, e.g. a complete loss of sleep spindles in some patients and the presence of sinusoid, high-amplitude theta activity instead of alpha activity. Slow-wave generation, on the other hand, seems intact in all patients. Rapid eye movement (REM) sleep is present in at least some patients, but it cannot be ascertained in all patients. The existence of intact sleep in patients in CLIS is another important sign that their wakefulness is likewise intact. It also indicates that providing undisturbed sleep opportunity will be important for the patients’ mental well-being. Introduction The sleep-wake cycle is a crucial component of neural and bodily development and restitution involved in a plethora of homeostatic processes [1]. Different exogenous zeitgebers, such as daylight, social stimulation, motor activity or food intake, synchronize the various central and peripheral endogenous oscillators [2–7] In the locked-in condition and the completely locked-in state (CLIS) [8], most of these zeitgebers are absent or attenuated due to complete immobility and absence of muscular activity, artificial respiration, artificial feeding and artificial lighting [9–12] Patients in CLIS suffering from advanced motor neuron diseases such as amyotrophic lateral sclerosis (ALS), who have an inability to open or close their eyes voluntarily, have their eyes closed most of the day and during the night to avoid drying of the immobile eyeball and cornea. In some cases, the complete closure of the eyelids remains impossible, leading to corneal lesions due to extensive drying with little or no differentiated vision left. Artificial feeding and breathing through a tracheostoma impose an extreme regularity on the internal organ systems, impeding the differentiation between day and nighttime. Social stimulation is certainly more frequent during the day but 24-hour care entails multiple activities also during the night (suction of saliva, excrements, eye care, position changes to avoid decubitus, change of tubes and feeding devices, etc.). The reduction of active social interaction and the complete lack of (verbal or signing) communication further minimize the differences between daytime and nighttime stimulation. Muscular paralysis also dramatically reduces the amount of afferent proprioceptive inputs towards the spinal cord and brain. Patients are bedridden and completely motionless over months and years, except during short periods of passive position changes. Although effects of constant conditions on homeostatic and circadian processes and on sleep have been investigated in constant routine studies and forced desynchrony protocols [13], in CLIS patients who constantly live in a controlled environment, the effects on their circadian rhythm are largely unknown [14]. Although, the scientific literature is rich on circadian rhythm and sleep-wake cycle in healthy populations [15–19] and patients with other neurological disorders [14, 20–25], there exists virtually no information about circadian rhythm and sleep-wake cycle in patients in the CLIS. For brain-computer interface (BCI) communication with these patients, it is necessary to discriminate periods of sleep and wakefulness in the individual EEG. It has been argued that the main reason for decreased or random performance in BCI communication might be the lack of attention and the presence of micro sleep during the presentation of the questions [11]. In a study of one CLIS and two LIS patients over a whole year it has been shown that a reduced P300 amplitude during the BCI task predicted lower performance, again suggesting reduced wakefulness and attention as a major limiting factor for BCI applications in these severely compromised patients [26]. To our knowledge, the sleep pattern of only one patient from our lab shortly after the transition from the locked-in state to the CLIS has been reported [27]. An irregular sleep pattern with daytime episodes of slow-wave sleep disrupted slow-wave sleep during the night and irregular appearance of different sleep stages during day and night was reported in this patient. The overall slow-wave sleep duration was normal for the age of the patient. It is of clinical and theoretical importance to study sleep in a larger sample of CLIS patients. The presence of distinguishable sleep and wake periods in these patients would constitute another piece of evidence for their cognitive functioning. Undisturbed sleep could also serve to prevent depression and maintain sufficiently high quality of life in these patients [28]. Hence, to elucidate their sleep-wake cycle, we recorded sleep in eight CLIS patients. We will outline criteria, which can be used to delineate the sleep stages, and describe and discuss the sleep cycles of the individual patients. Already at this point we need to mention the main limitations of this research, which are dictated by the clinical condition: the absence of eye movements and of changes in muscular tone due to the complete paralysis complicate scoring of rapid eye movement (REM) sleep. Mainly because of this limitation, it is impossible to label each 30-second epoch of the night with a sleep stage. We, therefore, refrain from presenting hypnograms and percentages of sleep stages, but only describe which sleep stages occur in a specific patient. Materials and Methods The Internal Review Board of the Medical Faculty of the University of Tubingen approved the experiment reported in this study and the patients’ legal representatives gave written informed consent for the study with permission to publish the results. The study is in full compliance with the ethical practice of Medical Faculty of the University of Tubingen. The clinical trial registration number is ClinicalTrials.gov identifier: NCT02980380. Patients The details of patients 1, 3, and 4 are described in Ref. [11] as patients F, G, and W, respectively. The remaining patients are described below. Patient 5 (male, 50 years old, CLIS) was diagnosed with bulbar sporadic ALS in May 2008, as locked-in in 2009, and as completely locked-in May 2010, based on the diagnosis of neurologists and on our recordings (see below and Ref. [11]). He has been artificially ventilated since September 2009, fed through a percutaneous endoscopic gastrostomy tube since October 2009, and is in home care. No communication with eye movements, other muscles, or assistive communication devices was possible since 2010. Patient 6 (male, 38 years old, CLIS) was diagnosed with bulbar ALS in 2009. He lost speech and capability to move by 2010. He has been artificially ventilated since September 2010 and is in home care. No communication with eye movements, other muscles, or assistive communication devices was possible since 2012. Patient 7 (female, 57 years old, CLIS) was diagnosed with Mills’ syndrome of ALS with atypical progression at the beginning of 2010. She lost speech and capability to walk by 2011. She has been fed through a percutaneous endoscopic gastrostomy tube since June 2010, artificially ventilated since June 2010, and was in home care until she passed away in 2017. She started using assistive communication devices employing eye movement for communication in 2011. Eye-tracker-based communication failed at the beginning of 2015. The family and caretakers communicated with her since the middle of 2015 based on her thumb-movements, which after a year became unreliable. Patient 9 (male, 23 years old, CLIS) was diagnosed with juvenile ALS with FUS mutation heterozygote on Exon 14: c.1504delG, gene mutation diagnosed in 2013. He has been artificially ventilated since August 2014 and is in home care. He started communication using MyTobii eye-tracking device in January 2015. He was able to communicate with MyTobii until December 2015 after which the family members attempted to communicate by training him to move his facial muscles near the nose to answer “yes” but the response was unreliable. No communication was possible since June 2016. Patient 10 (male, 25 years old, locked-in state on the verge of CLIS) was diagnosed with familial juvenile ALS with ALS 6-FUS gene mutation in December 2012. He was completely paralyzed within a year after diagnosis, has been artificially ventilated since November 2013, and is in home care. He was able to communicate with eye-tracking from early 2014 to August 2016 but was unable to use the eye-tracking device after the loss of eye control in August 2016. No communication with eye movements, other muscles, or assistive communication devices was possible since 2016. In none of these patients, voluntary eye movement responses to questions were recorded in any of the recording sessions. None of these patients showed any brain disease unrelated to ALS. CLIS onset was on average 38 months after initial diagnosis, which corresponds to expectations [29]. The proportion of juvenile-onset ALS was higher in our sample than in the general population of ALS patients [30]. Sleep recording Sleep EEG was recorded consecutively for two nights from each patient, except patient 4 and 7 (one night only). The first recording was carried out in order to adapt the patients to the electrodes. The second night EEG, EMG, and EOG were recorded for later sleep scoring. As shown in Table 1, sleep EEG data was acquired for more than 12 hours for all the patients, except patient 4. Table 1. Summary of findings for each patient Patient . Start of recording . End of recording . W [active] . W [inactive] . α-like freq. . SWS . REM sleep . Sleep spindles . 1 10:34 pm 11:14 am + + 4–6 Hz + (+) - 3 10:00 pm 10:17 am (+) + 4–7 Hz + + (few) 4 10:14 pm 07:33 am + + 2–4 Hz + (+) + 5 08:32 pm 05:11 pm + + 3–6 Hz + + - 6 08:38 pm 03:45 pm + + 2–4 Hz + (uncertain) - 7 09:10 pm 08:50 am (+) + 1–3 Hz + (+) + 9 02:04 pm 09:04 am + + 3–5 Hz + (+) + 10 06:49 pm 09:53 am (+) + 2–4 Hz + (uncertain) - Patient . Start of recording . End of recording . W [active] . W [inactive] . α-like freq. . SWS . REM sleep . Sleep spindles . 1 10:34 pm 11:14 am + + 4–6 Hz + (+) - 3 10:00 pm 10:17 am (+) + 4–7 Hz + + (few) 4 10:14 pm 07:33 am + + 2–4 Hz + (+) + 5 08:32 pm 05:11 pm + + 3–6 Hz + + - 6 08:38 pm 03:45 pm + + 2–4 Hz + (uncertain) - 7 09:10 pm 08:50 am (+) + 1–3 Hz + (+) + 9 02:04 pm 09:04 am + + 3–5 Hz + (+) + 10 06:49 pm 09:53 am (+) + 2–4 Hz + (uncertain) - +: clear signs of sleep stage present; (+): some signs of sleep stage present; (uncertain): possible signs of sleep stage present; -: no sign of sleep stage present. Open in new tab Table 1. Summary of findings for each patient Patient . Start of recording . End of recording . W [active] . W [inactive] . α-like freq. . SWS . REM sleep . Sleep spindles . 1 10:34 pm 11:14 am + + 4–6 Hz + (+) - 3 10:00 pm 10:17 am (+) + 4–7 Hz + + (few) 4 10:14 pm 07:33 am + + 2–4 Hz + (+) + 5 08:32 pm 05:11 pm + + 3–6 Hz + + - 6 08:38 pm 03:45 pm + + 2–4 Hz + (uncertain) - 7 09:10 pm 08:50 am (+) + 1–3 Hz + (+) + 9 02:04 pm 09:04 am + + 3–5 Hz + (+) + 10 06:49 pm 09:53 am (+) + 2–4 Hz + (uncertain) - Patient . Start of recording . End of recording . W [active] . W [inactive] . α-like freq. . SWS . REM sleep . Sleep spindles . 1 10:34 pm 11:14 am + + 4–6 Hz + (+) - 3 10:00 pm 10:17 am (+) + 4–7 Hz + + (few) 4 10:14 pm 07:33 am + + 2–4 Hz + (+) + 5 08:32 pm 05:11 pm + + 3–6 Hz + + - 6 08:38 pm 03:45 pm + + 2–4 Hz + (uncertain) - 7 09:10 pm 08:50 am (+) + 1–3 Hz + (+) + 9 02:04 pm 09:04 am + + 3–5 Hz + (+) + 10 06:49 pm 09:53 am (+) + 2–4 Hz + (uncertain) - +: clear signs of sleep stage present; (+): some signs of sleep stage present; (uncertain): possible signs of sleep stage present; -: no sign of sleep stage present. Open in new tab The sleep polysomnography was recorded with a multi-channel EEG amplifier (Brain Amp DC, Brain Products, Germany) from 11 Ag/AgCl passive electrodes mounted on a head cap. Six electrodes (F3, F4, C3, C4, O1, and O2) were used to acquire EEG signals, four electrodes were used to acquire the vertical and horizontal EOGs, and one electrode was used to acquire chin-EMG. EEG-channels were referenced to an electrode on the right mastoid and grounded to the electrode placed at the Fz location of the scalp. Electrode impedances were kept below 10 kΩ and the EEG signal was sampled at 500 Hz. Preprocessing The recorded EEG and EOG signals were low-pass filtered using a finite impulse response low-pass filter of 30 Hz. For the EMG, the signal was filtered with a 50 Hz notch filter. Sleep scoring The EEG of patients in CLIS is dissimilar in large extents to that of healthy participants. These patients show, e.g. regular high-amplitude oscillatory activity in the theta (4–8 Hz) frequency range for longer periods of time, which cannot be found in healthy participants. Moreover, they have no detectable eye movements during wakefulness and no muscle tone, due to their condition. Thus, the standard sleep-scoring criteria, particularly for REM sleep, have to be adapted to score the sleep stages from patients in CLIS. Still, initial visual inspection of the whole night EEG shows already obvious variations of EEG patterns over time. It can, therefore, be hypothesized that the states of arousal and consciousness also fluctuate throughout the night. Based on time of day and on interaction with the patient (movement artifacts), we had a starting point for a rough assumption of sleep and wakefulness: during the night, longer quiet periods without movement would have a higher probability of representing sleep than morning recordings and periods directly after movement (suction of saliva, repositioning of the patient). We then applied the criteria of Rechtschaffen and Kales [31] to the recordings to determine whether some features of sleep could be found. This scoring was performed visually on 30-s epochs. It resulted in the detection of a number of EEG patterns that generalized over patients, some of which resembled the classical sleep stages. Scoring was done by an experienced sleep scorer (S.G.) and discussed in detail among all authors. We describe the criteria that had to be adjusted as well as the criteria that could be applied directly in the results section. Signal analysis Power spectra were computed using Welch’s method with a resolution of 0.5 Hz for each 30-second epoch for all EEG channels. Channels containing obvious artifacts were excluded from further analysis. Whole-night spectrograms were calculated based on the median of all artifact-free channels. The median was used to further suppress artifacts and unusually high or low power values. Spectrograms were normalized by subtracting the median and dividing by the value of the 75th quantile of each frequency, analogous to Z-standardization, in order to have comparable ranges. The scale of the spectrograms is, therefore, dimensionless. Additionally, for all patients, the power spectra of a period of quiet wakefulness is presented to show peaks in EEG power. As the heartbeat was clearly reflected in the EMG trace of two patients (Pat. 1 and Pat. 5), we used these recordings to calculate the continuous heart rate. The 50 Hz notch-filtered EMG channel was band-pass filtered between 5 and 70 Hz using an IIR filter. Individual heart beats were detected as spikes in the signal. Spikes that were too large or too small were removed as outliers. Then, heart rate was calculated from the distance between spikes. Finally, outliers in heart rate were removed and the whole time series filtered with a 30-point moving average filter. Results Three main observations could be made during visual sleep scoring: (1) All patients had cyclic changes in their brain activity throughout the recording in the sense that we found alternating periods with and without slow-wave activity. (2) The discrepancy to healthy sleep EEG differed between patients. (3) Some commonalities in EEG anomalies across patients could be found. Because we observed that each patient has his or her own idiosyncrasies in the sleep-wake cycle, the result of each patient will be described first in this section. Subsequently, we will describe common patterns between participants and suggest heuristics for the scoring of sleep in CLIS patients. Exemplary epochs for individual sleep stages and whole-night power spectrograms can be found for each patient in the Supplementary Material. Exemplary data for patient 9 is presented in Figure 1. Figure 1. Open in new tabDownload slide Normalized spectrogram (upper left) and segments of selected sleep epochs of patient 9. Yellow color in the spectrogram reflects higher than average activity in a frequency band; blue color signifies lower than average activity. Red lines on the spectrogram indicate epochs belonging to particular sleep stages. Twelve-second segments of these epochs are shown in the other panels. Here, vertical lines represent 1-s intervals. The EEG during active wakefulness (W[A]) is largely similar to that of healthy participants. EEG activity during inactive wakefulness (W[I]) is around 5 Hz, but resembles alpha activity rather than typical theta activity. Stage 2 sleep (S2) with spindles and K-complexes as well as slow-wave sleep (SWS) can be clearly discerned in this patient. Note the large eye movements during REM sleep (R), which the patient is unable to perform voluntarily during wakefulness. Recordings of the other patients are available in the Supplementary Material, following the same arrangement. Figure 1. Open in new tabDownload slide Normalized spectrogram (upper left) and segments of selected sleep epochs of patient 9. Yellow color in the spectrogram reflects higher than average activity in a frequency band; blue color signifies lower than average activity. Red lines on the spectrogram indicate epochs belonging to particular sleep stages. Twelve-second segments of these epochs are shown in the other panels. Here, vertical lines represent 1-s intervals. The EEG during active wakefulness (W[A]) is largely similar to that of healthy participants. EEG activity during inactive wakefulness (W[I]) is around 5 Hz, but resembles alpha activity rather than typical theta activity. Stage 2 sleep (S2) with spindles and K-complexes as well as slow-wave sleep (SWS) can be clearly discerned in this patient. Note the large eye movements during REM sleep (R), which the patient is unable to perform voluntarily during wakefulness. Recordings of the other patients are available in the Supplementary Material, following the same arrangement. Patient 9 This patient shows clear high amplitude slow waves during the night, starting at around 1:30 am (Figure 1). These waves are in shape and in their clustered occurrence indistinguishable from slow waves of healthy patients. In the spectrogram, they are visible as a strong <2 Hz band. For longer periods of up to 45 minutes, the criteria for slow-wave sleep (SWS), sleep stages 3 (S3) or S4 are reached. For about 1 hour before the onset of S3, individual slow waves (K-complexes) and sleep spindles can be found (S2). In-between periods of S2, epochs dominated by bursts of regular, sinusoidal EEG activity of 3–5 Hz with medium to high amplitude can be found. These bursts have varying length, typically between 5 and 20 seconds. Because of its regularity and the duration of bursts, this activity is strongly reminiscent of alpha activity (see Figure 1, section W[I]). It also occurs for longer periods of several hours in the afternoon and in the evening before the first signs of S2. Although having theta band frequency, these waves have no similarity to the typical, irregular, sawtooth-shaped theta activity of healthy participants. Except for its frequency, this activity, therefore, in all other aspects seems to be equivalent to alpha activity in healthy patients. We, therefore, use this alpha-like activity to score periods of inactive, relaxed wakefulness (W[I]). Starting around 6:00 am, after signs of S2 have diminished, longer periods of low amplitude, high-frequency activity appear. Frequencies are dominated by beta-band activity, with a varying degree of intermixed irregular theta activity. Because of its similarity to typical wake activity and because of its strongest occurrence in the morning after sleep, we score this activity pattern as active wakefulness (W[A]). In the morning, this high-frequency activity is occasionally interrupted by 5- to 20-second periods of regular theta activity as described above (W[I]). In the afternoon, the pattern inverses and long periods of W[I] are interrupted by epochs of W[A]. Unrelated to S2 or SWS sleep phases, there are some epochs that show eye movements that resemble those typical for REM sleep. These occur mainly in the afternoon in the middle of longer periods of W[I], and are accompanied by a brief discontinuation of the alpha-like regular theta activity and appearance of irregular theta and beta waves. Although the timing of these periods is not typical for REM sleep, we suggest that these signs are indicative of REM sleep-like processes and score the corresponding epochs as REM sleep (R). Patient 1 This patient shows clear high amplitude slow waves (SWS) during the night, starting at around 2:00 am (Supplementary Figure S1). These waves are co-occurring with sinusoidal 4–7 Hz activity in some epochs. There are two more periods of SWS around 5:30 am and 7:30 am. There are no signs of sleep spindles throughout the night. The EEG in the evening before occurrence of SWS is dominated by the same regular, alpha-like 4–7 Hz (theta) activity (W[I]) as in patient 9, although with a lower amplitude. It is intermixed with high-frequency, low-amplitude activity (W[A]) to varying degrees. Interspersed throughout the recording are periods of dominant irregular theta activity during which REMs are manifest. Although the timing of these periods is not typical for REM sleep, we again suggest that these signs are indicative of REM sleep-like processes and score the corresponding epochs as R. Heart rate tendentially increases or reaches maxima over periods during which we did not detect any SWS (e.g. 10:30 pm – 02:00 am and 08:30 am – 10:30 am). During periods of SWS (marked by strong < 1.5 Hz activity) heart rate tendentially decreases. Patient 3 The EEG of this patient is dominated by regular, high-amplitude, alpha-like 4–7 Hz (theta) activity (W[I]), particularly in the evening and in some parts of the night (Supplementary Figure S2). Starting at 2:00 am, slow waves occur, first together with the alpha-like activity then replacing it more and more. There are several periods of SWS, alternating with periods of W[I]. In the morning, activity becomes more irregular (high-frequency beta and irregular theta activity), especially after intervals of (external) movement (W[A]); but still these periods contain amounts of alpha-like theta activity. After awakening and morning hygiene there is a 2-hour period of strong, continuous REM, accompanied mainly by irregular (sawtooth) theta activity and only little regular (sinusoidal) alpha-like theta activity. In this patient, the eye movement and EEG activity provide strong evidence for REM sleep. There is also evidence for a few sleep spindles in the morning after the period of morning hygiene and before REM sleep. Patient 4 We find two >1-hour periods of extensive SWS, at 11:00 pm and at 2:00 am (Supplementary Figure S3). During SWS, strong sleep spindle activity (12–14 Hz) is present. Strong spindle activity is also found during a 1-hour period around 4:00 am, which also presents K-complexes (S2). In-between SWS periods, there are periods dominated by 5- to 20-second bursts of regular, sinusoidal 2–4 Hz activity, which resemble the 4–7 Hz activity described in the patients above. These also occur in the evening, intermixed with high-frequency, low-amplitude activity. Because it shows all properties of alpha activity except the frequency range, we score these periods as W[I]. In the evening before sleep, the EEG is dominated by irregular theta and low-amplitude beta activity W[A]. In the morning starting around 6:00 am, we see similar irregular theta and low-amplitude beta activity, but accompanied by (mainly small and a few larger) REMs. Although difficult to discriminate from W[A], the quieter EMG (no external disturbance), the REMs, and the absence of alpha-like activity lead us to score R. Patient 5 This patient shows several periods of SWS between 10:00 pm and 5:00 am (Supplementary Figure S4). The slow waves are slower than usual and have a higher amplitude. In the evening before sleep and for 1 hour after the end of SWS in the morning, the EEG is dominated by 3–6 Hz alpha-like activity (W[I]). Starting at 6:00 am, the EEG becomes more mixed with irregular theta, beta, and only little alpha-like activity (W[A]). Throughout the day, periods of W[A] and W[I] alternate. In the afternoon at 05:00 pm, a period of clear REM sleep with strong REMs, irregular theta and low amplitude beta activity was found, which continued for >15 minutes until the end of the recording. No sleep spindles were found in this patient. Heart rate is minimal during the night period, during which strong <1.5 Hz activity is recorded (09:30 pm – 7:30 am). It increases in the morning during interaction with the patient. Patient 6 In patient 6, there are two clear periods of SWS between 4:30 am and 7:00 am (Supplementary Figure S5). In between these two periods, we find several quiet epochs of low-amplitude, high-frequency activity that contain a certain degree of small eye movements. Because they do not resemble periods of wakefulness in this patient, we label these epochs tentatively as uncertain R. The rest of the day alternates between periods of W[I] (mainly in the evening) and W[A] (more abundant in the morning), showing more or less alpha-like 2–4 Hz activity, respectively. Again, no sleep spindles were found. Patient 7 This patient shows about 1 hour of SWS with very slow, high-amplitude EEG around midnight (Supplementary Figure S6). Before this period, a few minutes of S2 sleep containing sleep spindles was found. This period is also represented with a yellow spot in the 12–15 Hz band in the spectrogram of this patient. The rest of the recording contains mainly mixed frequency activity in the 1–3 Hz, theta, and beta bands. The amount of low-frequency activity varies throughout the recording, probably reflecting W[I] and W[A]. However, these variations are only gradual, so that a clear distinction between these two stages is difficult in this patient. Moreover, the 1–3 Hz low-frequency activity is less regular and sinusoidal than the alpha-like activity of the previous patients. Although it lies in the delta band and could be taken for slow-wave activity, it occurs also during periods where caretakers interact with the patients, and it much faster than the patients SWS activity, which can be clearly delineated in the patient’s spectrogram. We, therefore, believe this activity corresponds to the alpha-like activity in the other patients. During an undisturbed period at the beginning of the night, we found irregular theta activity together with some REMs and an absence of 1–3 Hz activity. We score this as a brief period of R. Patient 10 The spectrogram of this patient shows three periods of SWS between 3:00 am and 7:00 am (Supplementary Figure S7). During these periods, very slow (<1 Hz) activity can be seen in the sleep recording. These slow waves have the typical shape of sleep slow waves. In the evening before 3:00 am, the EEG is dominated by strong 2–4 Hz activity, which, as in patient 7, can be distinguished from SWS. Because of its regular sinusoidal shape, we score it as alpha-like activity and W[I]. In the morning, alpha-like activity alternates with slightly faster regular and irregular theta activity, which we score as W[A]. During a period with only a minimum of 2–4 Hz activity, we see a few small eye movements that could represent REMs. We, therefore, score some epochs as uncertain R. Sleep spindles were not found. The power spectra of channel C3 of all eight patients can be found in Supplementary Figure S8. The figure shows a single peak in the spectrum in most patients. This peak is not found in the alpha band, but in a lower range between 2 and 7 Hz depending on the patient. The individual alpha-like frequency band for each patient is shown in Table 1. Comparing all six recorded channels, we did not observe any obvious topography (e.g. anterior-posterior) of alpha-like activity across the scalp. Discussion Polysomnographic recordings in CLIS patients show that these patients have a circadian sleep-wake pattern. All patients show <1 Hz slow-wave activity during one or several periods during the night. In most patients, the dominant activity outside of SWS was a regular, sinusoidal 4–6 Hz or 2–4 Hz activity, which resembles alpha activity in its distribution and burst-like behavior. This inactive wakefulness could be discriminated from active wakefulness in most patients, with active wakefulness showing irregular, higher frequency activity. Sleep spindles were absent in half of the patients. REM sleep was clearly present in two patients and probably present in most. Impaired REM sleep and lack of sleep spindles were independent, as the two patients showing the strongest REM sleep did not present any sleep spindles. Slow activity dominates the EEG of CLIS patients during sleep and wakefulness. The slow waves of SWS have their typical frequency in some patients, in others they appear to be distinctly slower. However, all patients have slow waves in the range below 1 Hz, which can be easily detected by visual scoring or automatic analysis in the EEG. The timing of this activity, which occurs mainly after midnight and before 6:00 am in the morning, and which represents the slowest activity for each patient, leaves little doubt, that this activity actually represents SWS. All patients thus showed one or more periods of clear S3 or S4 sleep. Consistent with expectations, in both patients for which heart rate data was available, heart rate was decreasing or minimal during periods scored as SWS. During the 12-hour section of the circadian rhythm that we have recorded, we found longer stretches with slow-wave activity, which were mostly split into several consecutive periods. Night-time periods with and without slow-wave activity did not differ systematically with regard to light or other stimulation, except that sometimes a period of slow-wave activity ended with EMG activation (patient being moved). Our findings speak to an astonishingly well-conserved circadian rhythm, at least within our period of observation. It is, however, possible that there are additional periods of sleep during daytime that we did not record in the present study. Moreover, there might be periods of light S2 sleep, which we have not detected because of the lack of sleep spindles in most patients. There is, however, another type of slow activity, which can be confounded with SWS in more than half of the patients. Most of the recordings are dominated by a regular oscillation in the upper delta/lower theta range (2–7 Hz). Several reasons speak against this activity being sleep-related. First, it appears at all times of day throughout the recordings: more strongly in the evenings, but also in the mornings. It often continues throughout periods of interaction with the patient. Second, it is distinct from the slower <1 Hz oscillations, which occur during the night, and mostly disappear during these periods. Third, this oscillation has a regular, sinusoidal shape, occurs in 5- to 20-second bursts, and has a waxing and waning amplitude during these bursts. Its shape clearly distinguishes it from theta and delta activity, which is usually more sawtooth or rectangular shaped. In most patients, it is also reflected by a single, pronounced peak in the power spectrum. Apart from the lower frequency, it is therefore strongly reminiscent of periods of alpha, which occur during resting wakefulness with eyes closed in healthy participants. CLIS patients have their eyes closed most of the time to prevent drying of the eyeballs. It could, therefore, be expected to see strong alpha oscillations, but none are found in the 8–12 Hz range in any patient. Previous studies in locked-in ALS patients also showed a significant reduction in the alpha band over the central electrodes [32]. Moreover, our observations confirm the observations of Hohmann et al. [33], who showed in two CLIS patients that alpha frequency activity has completely disappeared and shifted towards lower frequencies. We, therefore, hypothesize that the continuous alpha activity is gradually slowing in frequency throughout the progression of the disease, may be due to lack of sensory stimulation or overuse of alpha-generating circuitry. This, however, remains open for further investigation. In fact, there are reports of “alpha coma” and dominant alpha-theta activity in locked-in patients that are not completely locked-in [34, 35]. The alpha-like theta/delta oscillation might, therefore, serve as an indicator of absence of slow-wave sleep. In healthy participants, alpha activity serves, together with rolling eye movements, as an indicator of falling asleep (S1). No CLIS patient showed rolling eye movements. Because of the persistence of alpha-like activity, the impossibility to describe consistent and distinct features of wake EEG in these patients, and the lack of light S2 sleep in most patients, it was impossible to distinguish a transitory S1 sleep stage in CLIS patients. We have therefore described the state of continuous (>50% of an epoch) alpha-like activity as inactive wakefulness (W[I]). During most parts of the day, this stage alternates with periods of active wakefulness, which has faster more irregular activity with lower amplitudes. Sleep spindles (11–17 Hz), are the hallmark of light S2 sleep. These were completely absent in four out of eight patients. The lack of sleep spindles mirrors the findings by Pavlov et al. [36], who also found few or no spindles in most of their non-responsive (vegetative state) patients. It is yet unclear when and why sleep spindles cease to occur in these patients. As spindles have been linked to the reprocessing of new memories during sleep [37], one might speculate that the daily routine of the patients with lack of change or new information renders sleep spindles superfluous. It is conceivable that the frequency of spindle oscillations also slows with the progression of the disease, but we have not found any indication of such a slowing. In fact, in those patients with spindles, these had all typical characteristics of sleep spindles in healthy participants. K-complexes, which often have a higher frequency than full slow-waves, were difficult to delineate. The presence of high-amplitude delta and theta band activity during wakefulness prevents the clear definition of individual K-complexes. However, no periods resembling S2 sleep (low-amplitude activity with a few, clearly delineated high-amplitude waves) were found in patients that did not also show sleep spindles. Scoring of REM sleep according to standard rules relies largely on eye movements and muscle tone changes. Although patients in CLIS cannot produce any voluntary eye movements or muscle contractions, we found strong REMs during sleep in two patients and rudimentary eye movements during sleep in another four patients. This finding indicates that voluntary and autonomous eye movement control can be independent in CLIS patients. REM sleep-like irregular theta activity with medium amplitude was present during these REM periods. Scoring of REM sleep by EEG alone, which is feasible by experienced scorers in healthy patients, was impossible in our patients. The EEG could only be used in conjunction with the eye movement signal, and confident judgments of REM sleep were only possible in a few patients. Heart rate, which was unfortunately available only in patients 1 and 5, might be an additional helpful parameter for scoring REM sleep. In both patients, heart rate during periods scored as REM sleep or potential REM sleep was higher than during periods scored as sleep, and variability of the heart rate was greater. Further studies should consider placing explicit ECG electrodes. Although we did not find any increase in spectral power during REM sleep, we noticed a distinctive decline in 2–3 Hz activity in the spectrograms of those patients with strong REMs. We found similar periods of distinctly decreased power in this frequency band in most patients. Visual inspection of the EEG during these periods visually confirmed REM sleep-like theta activity with small deflections in the EOG trace that could not be accounted for other than by eye movements (e.g. no corresponding activity in the EEG traces, no movement artifacts). These eye movements are much smaller and fewer than those habitually seen in healthy participants, but still, as there are no other sources of artifacts in CLIS patients, they can be taken as signs of autonomous eye movement activity. We can thus confirm the presence of REM sleep in some CLIS patients and the presence of possible REM sleep in most. However, new rules for scoring REM sleep, perhaps based on spectral EEG power, should be developed to increase sleep scoring accuracy in these patients. The present data indicates that REM sleep seems to be uncoupled from the typical NREM-REM cycle and to occur at different times of the day. This pattern might be related to the lack of movement and structure in the daily routine of the patients or to an uncoupling of circadian and ultradian cycles. Comparison with other bed-ridden patients might illuminate this aspect of CLIS sleep. The analysis of the nighttime EEG demonstrates that patients in CLIS with ALS show slow-wave sleep episodes comparable to the healthy aged population. The two younger patients (9 and 10), suffering from a genetically determined ALS, do not substantially differ from the older ALS patients. Overall, sleep is fragmented, as already noticed [27], in these completely inactive patients, which spend most of their time, some of them since more than 5 years, in a completely paralyzed state in their bed, with only some transfer to the wheelchair during daytime. This constant, mostly bedridden routine and closed eyes, weakening the influence of light as the most potent zeitgeber, might be one reason for the disruption of the NREM-REM cycle and the absence of some of the typical sleep signs. Further studies should include neuroimaging to allow a more mechanistic investigation of the relation between pathologic changes in the brain and in sleep. Still, the maintenance of SWS in all patients might be an important factor contributing to preserve the quality of life in patients in an advanced locked-in state. Undisturbed nighttime sleep should, therefore, be aimed at in CLIS patient care. Funding The work of the authors is supported by the Deutsche Forschungsgemeinschaft (DFG, Kosellek) DFG BI 195/77-1, BMBF (German Ministry of Education and Research) 16SV7701 CoMiCon, LUMINOUS-H2020- FETOPEN-2014- 2015-RIA (686764), and Wyss Center for Bio and Neuroengineering, Geneva. Conflict of interest statement. None declared. References 1. Jennett B , et al. Persistent vegetative state after brain damage. A syndrome in search of a name . Lancet. 1972 ; 1 ( 7753 ): 734 – 737 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Boivin DB , et al. Dose-response relationships for resetting of human circadian clock by light . Nature . 1996 ; 379 ( 6565 ): 540 – 542 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Hastings MH , et al. A clockwork web: circadian timing in brain and periphery, in health and disease . Nat Rev Neurosci . 2003 ; 4 ( 8 ): 649 – 661 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Challet E , et al. Synchronization of the molecular clockwork by light- and food-related cues in mammals . Biol Chem . 2003 ; 384 ( 5 ): 711 – 719 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Mistlberger RE , et al. Social influences on mammalian circadian rhythms: animal and human studies . Biol Rev Camb Philos Soc. 2004 ; 79 ( 3 ): 533 – 556 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Bell-Pedersen D , et al. Circadian rhythms from multiple oscillators: lessons from diverse organisms . Nat Rev Genet . 2005 ; 6 ( 7 ): 544 – 556 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Golombek DA , et al. Physiology of circadian entrainment . Physiol Rev . 2010 ; 90 ( 3 ): 1063 – 1102 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Bauer G , et al. Varieties of the locked-in syndrome . J Neurol . 1979 ; 221 ( 2 ): 77 – 91 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Birbaumer N . Breaking the silence: brain-computer interfaces (BCI) for communication and motor control . Psychophysiology . 2006 ; 43 ( 6 ): 517 – 532 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Chaudhary U , et al. Brain-computer interfaces for communication and rehabilitation . Nat Rev Neurol . 2016 ; 12 ( 9 ): 513 – 525 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Chaudhary U , et al. Brain-Computer Interface-Based Communication in the Completely Locked-In State . PLoS Biol . 2017 ; 15 ( 1 ): e1002593 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Chaudhary U , et al. Brain-computer interfaces in the completely locked-in state and chronic stroke . Prog Brain Res . 2016 ; 228 : 131 – 161 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Blatter K , et al. Circadian rhythms in cognitive performance: methodological constraints, protocols, theoretical underpinnings . Physiol Behav . 2007 ; 90 ( 2-3 ): 196 – 208 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Cologan V , et al. Sleep in disorders of consciousness . Sleep Med Rev . 2010 ; 14 ( 2 ): 97 – 105 . Google Scholar Crossref Search ADS PubMed WorldCat 15. Czeisler CA , et al. Human sleep: its duration and organization depend on its circadian phase . Science . 1980 ; 210 ( 4475 ): 1264 – 1267 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Czeisler CA , et al. Stability, precision, and near-24-hour period of the human circadian pacemaker . Science . 1999 ; 284 ( 5423 ): 2177 – 2181 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Van Someren EJ . Circadian rhythms and sleep in human aging . Chronobiol Int . 2000 ; 17 ( 3 ): 233 – 243 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Dijk DJ , Lockley SW . Integration of human sleep-wake regulation and circadian rhythmicity . J Appl Physiol (1985) . 2002 ; 92 ( 2 ): 852 – 862 . doi: 10.1152/japplphysiol.00924.2001 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 19. Saper CB , et al. Hypothalamic regulation of sleep and circadian rhythms . Nature . 2005 ; 437 ( 7063 ): 1257 – 1263 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Hulihan JF Jr , et al. Electroencephalographic sleep patterns in post-anoxic stupor and coma . Neurology . 1994 ; 44 ( 4 ): 758 – 760 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Isono M , et al. Sleep cycle in patients in a state of permanent unconsciousness . Brain Inj . 2002 ; 16 ( 8 ): 705 – 712 . doi: 10.1080/02699050210127303 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 22. Germain A , et al. Circadian rhythm disturbances in depression . Hum Psychopharmacol . 2008 ; 23 ( 7 ): 571 – 585 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Wulff K , et al. Sleep and circadian rhythm disruption in psychiatric and neurodegenerative disease . Nat Rev Neurosci . 2010 ; 11 ( 8 ): 589 – 599 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Videnovic A , et al. Circadian and sleep disorders in Parkinson’s disease . Exp Neurol . 2013 ; 243 : 45 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Musiek ES , et al. Sleep, circadian rhythms, and the pathogenesis of Alzheimer disease . Exp Mol Med . 2015 ; 47 : e148 . Google Scholar Crossref Search ADS PubMed WorldCat 26. De Massari D , et al. Brain communication in the locked-in state . Brain . 2013 ; 136 ( Pt 6 ): 1989 – 2000 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Soekadar SR , et al. Fragmentation of slow wave sleep after onset of complete locked-in state . J Clin Sleep Med . 2013 ; 9 ( 9 ): 951 – 953 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 28. Riemann D , et al. Sleep and depression–results from psychobiological studies: an overview . Biol Psychol . 2001 ; 57 ( 1-3 ): 67 – 103 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Chiò A , et al. ; Eurals Consortium . Prognostic factors in ALS: a critical review . Amyotroph Lateral Scler . 2009 ; 10 ( 5-6 ): 310 – 323 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Raymond J , et al. Clinical characteristics of a large cohort of US participants enrolled in the National Amyotrophic Lateral Sclerosis (ALS) Registry, 2010–2015 . Amyotroph Lateral Scler Frontotemporal Degener . 2019 ; 20 ( 5-6 ): 413 – 420 . doi: 10.1080/21678421.2019.1612435 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 31. Rechtschaffen A , Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects . Los Angeles, CA : Brain Information Service/Brain Research Institute, UCLA ; 1968 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 32. Mai R , et al. Quantitative electroencephalography in amyotrophic lateral sclerosis . Electroencephalogr Clin Neurophysiol . 1998 ; 106 ( 4 ): 383 – 386 . Google Scholar Crossref Search ADS PubMed WorldCat 33. Hohmann MR , et al. Case series: slowing alpha rhythm in late-stage ALS patients . Clin Neurophysiol . 2018 ; 129 ( 2 ): 406 – 408 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Jacome DE , et al. Unreactive EEG: pattern in locked-in syndrome . Clin Electroencephalogr . 1990 ; 21 ( 1 ): 31 – 36 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Bragatti JA , et al. Alpha coma and locked-in syndrome . J Clin Neurophysiol . 2007 ; 24 ( 3 ): 308 . doi: 10.1097/WNP.0b013e31803bb72c . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 36. Pavlov YG , et al. Night sleep in patients with vegetative state . J Sleep Res . 2017 ; 26 ( 5 ): 629 – 640 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Gais S , et al. Sleep after learning aids memory recall . Learn Mem . 2006 ; 13 ( 3 ): 259 – 262 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes These authors contributed equally. Data collection. Principal investigator of the project. © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail [email protected]. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Barateau,, Lucie;Chenini,, Sofiene;Evangelista,, Elisa;Jaussent,, Isabelle;Lopez,, Regis;Dauvilliers,, Yves

doi: 10.1093/sleep/zsz187pmid: 31418025

Abstract Study Objectives (1) To compare the presence of autonomic symptoms using the validated SCOPA-AUT questionnaire in untreated patients with narcolepsy type 1 (NT1) to healthy controls, (2) to study the determinants of a high total SCOPA-AUT score in NT1, and (3) to evaluate the effect of drug intake on SCOPA-AUT results in NT1. Methods The SCOPA-AUT questionnaire that evaluates gastrointestinal, urinary, cardiovascular, thermoregulatory, pupillomotor, and sexual dysfunction was completed by 92 consecutive drug-free adult NT1 patients (59 men, 39.1 ± 15.6 years old) and 109 healthy controls (63 men, 42.6 ± 18.2 years old). A subgroup of 59 NT1 patients completed the questionnaire a second time, under medication (delay between two evaluations: 1.28 ± 1.14 years). Results Compared to controls, NT1 patients were more frequently obese, had more dyslipidemia, with no difference for age and gender. The SCOPA-AUT score of NT1 was higher than in controls in crude and adjusted models. Patients experienced more problems than controls in all subdomains. A higher score in NT1 was associated with older age, longer disease duration, altered quality of life and more depressive symptoms, but not with orexin levels and disease severity. Among patients evaluated twice, the SCOPA-AUT score total did not differ according to treatment status, neither did each subdomain. Conclusion We captured a frequent and large spectrum of clinical autonomic dysfunction in NT1, with impairment in all SCOPA-AUT domains, without key impact of medication intake. This assessment may allow physicians to screen and treat various symptoms, often not spontaneously reported but associated with poor quality of life. narcolepsy, orexin/hypocretin, autonomic dysfunction, SCOPA-AUT, cataplexy Statement of Significance Using a reliable and valid instrument, we found a frequent and large spectrum of clinical autonomic dysfunction, with impairment in every subdomains (gastrointestinal, urinary, cardiovascular, thermoregulatory, pupillomotor, and sexual areas) of the SCOPA-AUT questionnaire, in a large population of well-characterized patients with NT1 compared to a control population. Higher scores on SCOPA-AUT were associated with older age, poorer quality of life and more depressive symptoms in patients with NT1, without significant effect of the drugs intake for narcolepsy. This assessment may allow physicians to screen and treat various symptoms, often not spontaneously reported but associated with poor quality of life. Introduction Narcolepsy type 1 (NT1) is a rare neurological disorder, caused by the selective and irreversible destruction of a small population of neurons in the lateral and posterior hypothalamus that release a neuropeptide, hypocretin/orexin (ORX) [1, 2]. The main symptoms of narcolepsy are excessive daytime sleepiness and cataplexy, but patients also present hypnagogic hallucinations, sleep paralysis, disturbed nighttime sleep with multiple nocturnal arousals and dissociated rapid eye movement (REM) sleep features, as REM sleep without atonia and REM sleep behavior disorder (RBD) [2, 3]. ORX is implicated in various physiological functions such as sleep and wake regulation, energy homeostasis, feeding, thermoregulation, neuroendocrine, and cardiovascular control [4, 5]. ORX neurons also project to brainstem structures involved in autonomic regulation [6–8]. Several preclinical evidences favour a direct effect of ORX on the modulation of autonomic control [8]. Using different assessments (microneurography, ambulatory blood pressure [BP] monitoring), several human studies found abnormal sympathetic activation in drug-free narcoleptic patients, together with a non-dipper BP profile, metabolic and heart rate variability alterations [9–11]. In addition, clinical autonomic dysfunctions were often reported in NT1: abnormal pupillometry function, erectile dysfunction, fainting spells, and impaired autonomic control of the cardiovascular system [9]. Autonomic dysfunctions may cause serious health problems, and are often associated with depressive symptoms and altered quality of life [12]. Therapeutic interventions may help patients to manage those autonomic symptoms; however, most drugs prescribed for narcolepsy, namely psychostimulants and anticataplectics may also interfere with the autonomic nervous system functioning [13, 14]. Moreover, several frequent comorbidities in NT1 such as obesity, RBD, and depressive symptoms, may impact on the autonomic nervous system regulation. Nowadays, autonomic symptoms can be screened by questionnaires such as the “scales for outcomes in Parkinson’s disease - autonomic” (SCOPA-AUT), validated in Parkinson’s disease [15], used in other synucleinopathies [16], and in sleep disorders, idiopathic RBD (iRBD) [17], restless legs syndrome (RLS) [18], and obstructive sleep apnea syndrome (OSAS) [19]. However, large-scale assessment of autonomic symptoms, has never been performed in large population of NT1, nor the effect of drugs intake has been studied. In the present study, we aimed (1) to compare autonomic symptoms using the SCOPA-AUT questionnaire in a large sample of untreated NT1 patients to healthy controls, taking into account potential confounders, (2) to study the clinical, biological and neurophysiological determinants of a high total SCOPA-AUT score in NT1, and (3) to assess the effect of stimulants and anticataplectic medication on autonomic symptoms in NT1. Methods Participants We recruited 92 consecutive drug-free adult patients (63.1% men, mean age 39.13 ± 15.69 years old) with a diagnosis of NT1 according to International Classification of Sleep Disorders-3 [20] in the National Reference Center for Narcolepsy, Montpellier, France, between 2014 and 2018. The diagnosis of NT1 was based on the presence of excessive daytime sleepiness (EDS), a mean sleep latency below 8 minutes on the Multiple Sleep Latency Test (MSLT), with at least 2 sleep-onset REM periods (SOREMPs), typical cataplexy and/or low ORX-A cerebrospinal fluid (CSF) levels (<110 pg/mL). Immunogenetic (HLA typing) was obtained in all patients, all were positive for HLA-DQB1*06:02. CSF ORX-A levels were determined in 71 patients (77.2%), all had low levels (<110 pg/mL). CSF ORX-A levels were determined in duplicate using the I125-radioimmunoassay (RIA) kit from Phoenix Pharmaceuticals, Inc, according to the manufacturer’s recommendations. A subgroup of 59 NT1 patients completed the questionnaire a second time, under medication. All drugs and doses were recorded and classified as: (1) psychostimulants (modafinil, methylphenidate or pitolisant, n = 56, 95%), (2) sodium oxybate (n = 13, 22%), and (3) anticataplectic medication (serotonin-norepinephrine reuptake inhibitors [SNRI] and serotonin specific reuptake inhibitors [SSRI], n = 29, 49%). We also recruited 109 community-dwelling adults (57.8% men, mean age 42.61 ± 18.20 years old) through advertisements and local association networks, during the same period, from the general population. They had no complaints of EDS (defined by Epworth Sleepiness Scale [ESS] score < 11), and no insomnia (defined by Insomnia Severity Index [ISI] score ≤ 14), were not taking any psychotropic medications or drugs that may influence sleep and autonomic regulation, and did not have any significant medical, neurological or psychiatric conditions. All participants could speak and understand French. To avoid potential bias, no details were given to participants about the exact objectives of this study. The methods were carried out in accordance with the approved guidelines. This study was approved by the institutional review board of the University of Montpellier, France. All participants consented to participate in the study. Main study outcome: assessment of clinical autonomic symptoms by the SCOPA-AUT scale The SCOPA-AUT scale is a 25-item self-administered questionnaire that assesses gastrointestinal (7 items), urinary (6 items), cardiovascular (3 items), thermoregulatory (4 items), pupillomotor (1 item) and sexual (2 items for men, and 2 others for women) dysfunctions. Each item is scored from 0 (never) to 3 (often), and refers to symptoms within the past month, except for syncope (past 6 months). The SCOPA-AUT total score ranges from 0 to 69, with the highest score reflecting worse autonomic functioning. Other clinical data All participants had a semi-structured medical interview, a clinical examination, and completed questionnaires. Age at inclusion, gender, body mass index (BMI), alcohol and tobacco regular use, physical activity and educational level were recorded for all participants. Daytime sleepiness was quantified by ESS [21] and insomnia symptoms by ISI [22]. A homemade structured questionnaire recorded the presence of cardiovascular comorbidities, defined as the presence of diabetes type 2 or use of antidiabetic drugs, cardiac disorders (i.e. coronary artery disease, chronic heart failure, arrhythmia), hypertension or treatment with antihypertensive drugs, dyslipidemia or treatment with lipid-lowering drugs, or stroke. Current use of all drugs, and especially those that may interfere with autonomic function was assessed in all subjects. In the patients group, symptoms of depression were detected with the Beck Depression Inventory (BDI) and quality of life was assessed using the European Quality of Life five-dimensions (EQ-5D) instrument, including a health self-classification system (EQ-5D utility values) and a visual analog scale (EQ-5D VAS) [23]. We also recorded age at onset of sleepiness and cataplexy, cataplexy frequency, presence of hypnagogic hallucinations, sleep paralysis, and REM sleep behavior disorder. The global severity of the disease was assessed by Narcolepsy Severity Scale (NSS) [24] in a subgroup of 66 patients. Neurophysiological data All patients with NT1 underwent a video-polysomnography (PSG) followed by MSLT. Proportions of each stage of sleep, micro-arousals, periodic limb movements during sleep (PLMS), and apnea-hypopnea index were recorded. REM sleep without atonia was quantified, defined as the presence of a tonic muscle activity > 30% of the total REM sleep duration, and/or phasic electromyography (EMG) activity during REM sleep > 15% [25]. Mean sleep latency and number of SOREMPs on the MSLT were recorded. Statistical analysis Categorical variables were presented as percentages, quantitative variables as mean and SD. Patients with NT1 and controls were compared using logistic regression models. Associations were quantified using odds ratios (ORs) and their confidence intervals (CI). Demographic and clinical variables associated with NT1 in univariate analysis (at p < 0.10; Table 1) were included in logistic regression models to estimate the adjusted OR and their 95% CI for the associations between SCOPA dimensions and NT1. The same methodology was used to study the relationships between the patients’ clinical and neurophysiological characteristics and a high SCOPA-AUT total score (defined as the highest tertile of the distribution in NT1 patients). The dependent t-test was used to compare differences between continuous variables at two different conditions (before treatment, and after treatment) and the agreement between the two conditions was evaluated with the Cohen’s Kappa coefficient for dichotomous variables. The significance level was set at p < 0.05. Analyses were performed using SAS statistical software (version 9.4; SAS, Cary, NC). Table 1. Demographic and clinical characteristics of drug-free patients with NT1 compared to controls . Controls, N = 109 . NT1, N = 92 . OR [95% CI] . p-value . n . % . n . % . Demographic and clinical characteristics  Gender, men 63 57.80 59 63.13 0.77 [0.43;1.36] 0.36  Age (years)a 42.61 (±18.20) 39.13 (±15.69) 0.99 [0.97;1.00] 0.15  BMI (kg/m2)a 23.98 (±3.27) 26.97 (±4.66) 1.21 [1.12;1.32] <0.0001  BMI (kg/m2)   <25 74 67.89 35 38.04 1 <0.0001   25–30 30 27.52 35 38.04 2.47 [1.31;4.64]   ≥30 5 4.59 22 23.91 9.30 [3.25;26.6] Metabolic and cardiovascular comorbidities  Dyslipidemia, yes 7 6.54 18 20.93 3.78 [1.50;9.54] 0.005  Diabetes type 2, yes 4 3.70 6 7.06 1.97 [0.54;7.24] 0.30  Hypertension, yes 9 8.33 8 9.52 1.16 [0.43;3.14] 0.77  Cardiac disorders, yes 0 0.00 5 5.43 NA . Controls, N = 109 . NT1, N = 92 . OR [95% CI] . p-value . n . % . n . % . Demographic and clinical characteristics  Gender, men 63 57.80 59 63.13 0.77 [0.43;1.36] 0.36  Age (years)a 42.61 (±18.20) 39.13 (±15.69) 0.99 [0.97;1.00] 0.15  BMI (kg/m2)a 23.98 (±3.27) 26.97 (±4.66) 1.21 [1.12;1.32] <0.0001  BMI (kg/m2)   <25 74 67.89 35 38.04 1 <0.0001   25–30 30 27.52 35 38.04 2.47 [1.31;4.64]   ≥30 5 4.59 22 23.91 9.30 [3.25;26.6] Metabolic and cardiovascular comorbidities  Dyslipidemia, yes 7 6.54 18 20.93 3.78 [1.50;9.54] 0.005  Diabetes type 2, yes 4 3.70 6 7.06 1.97 [0.54;7.24] 0.30  Hypertension, yes 9 8.33 8 9.52 1.16 [0.43;3.14] 0.77  Cardiac disorders, yes 0 0.00 5 5.43 NA NA, test non applicable. aContinuous variables are expressed as means (±SD). Open in new tab Table 1. Demographic and clinical characteristics of drug-free patients with NT1 compared to controls . Controls, N = 109 . NT1, N = 92 . OR [95% CI] . p-value . n . % . n . % . Demographic and clinical characteristics  Gender, men 63 57.80 59 63.13 0.77 [0.43;1.36] 0.36  Age (years)a 42.61 (±18.20) 39.13 (±15.69) 0.99 [0.97;1.00] 0.15  BMI (kg/m2)a 23.98 (±3.27) 26.97 (±4.66) 1.21 [1.12;1.32] <0.0001  BMI (kg/m2)   <25 74 67.89 35 38.04 1 <0.0001   25–30 30 27.52 35 38.04 2.47 [1.31;4.64]   ≥30 5 4.59 22 23.91 9.30 [3.25;26.6] Metabolic and cardiovascular comorbidities  Dyslipidemia, yes 7 6.54 18 20.93 3.78 [1.50;9.54] 0.005  Diabetes type 2, yes 4 3.70 6 7.06 1.97 [0.54;7.24] 0.30  Hypertension, yes 9 8.33 8 9.52 1.16 [0.43;3.14] 0.77  Cardiac disorders, yes 0 0.00 5 5.43 NA . Controls, N = 109 . NT1, N = 92 . OR [95% CI] . p-value . n . % . n . % . Demographic and clinical characteristics  Gender, men 63 57.80 59 63.13 0.77 [0.43;1.36] 0.36  Age (years)a 42.61 (±18.20) 39.13 (±15.69) 0.99 [0.97;1.00] 0.15  BMI (kg/m2)a 23.98 (±3.27) 26.97 (±4.66) 1.21 [1.12;1.32] <0.0001  BMI (kg/m2)   <25 74 67.89 35 38.04 1 <0.0001   25–30 30 27.52 35 38.04 2.47 [1.31;4.64]   ≥30 5 4.59 22 23.91 9.30 [3.25;26.6] Metabolic and cardiovascular comorbidities  Dyslipidemia, yes 7 6.54 18 20.93 3.78 [1.50;9.54] 0.005  Diabetes type 2, yes 4 3.70 6 7.06 1.97 [0.54;7.24] 0.30  Hypertension, yes 9 8.33 8 9.52 1.16 [0.43;3.14] 0.77  Cardiac disorders, yes 0 0.00 5 5.43 NA NA, test non applicable. aContinuous variables are expressed as means (±SD). Open in new tab Results Patients with NT1 and controls were not different for age and gender, but patients had a higher BMI, were more frequently obese, and had more dyslipidemia (Table 1). The results were thus adjusted for these variables in the subsequent analyses. Frequency of hypertension and diabetes was comparable between groups, and no history of stroke was recorded in both populations. Few cardiac disorders were recorded among NT1 patients: rhythms disorders (n = 2), heart failure (n = 1), and myocardial infarction (n = 2). SCOPA-AUT scores in untreated NT1 patients and controls Narcoleptic patients had more autonomic symptoms than controls, measured by the total SCOPA-AUT score (Table 2, model 1). Fifty-six NT1 patients (60.9%) and 19 controls (17.6%) had a score ≥14 (highest tertile of the whole population) (p < 0.0001). All the associations remained significant after adjustment for BMI (Table 2, model 2), and also for BMI and dyslipidemia (p < 0.0001). Similar results were found when comparing 56 NT1 patients and 56 controls matched for age, gender, and BMI (data not shown). Results were also significant when obese participants were excluded from the analyses. Table 2. SCOPA-AUT Total Scores and Subdomain Scores in drug-free patients with NT1 compared to controls . Controls, N = 109 . NT1, N = 92 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . SCOPA-AUT Total Score  SCOPA-AUT Total Scorea 5.36 (±5.05) 12.92 (±6.84) 1.23 [1.16;1.31] <0.0001 1.21 [1.14;1.29] <0.0001  SCOPA-AUT Total Scoreb   <4 50 46.30 7 7.61 1 <0.0001 1 <0.0001   [4–11] 39 36.11 29 31.52 5.31 [2.11;13.4] 4.79 [1.85;12.4]   ≥14 19 17.59 56 60.87 21.05[8.17;54.3] 16.77 [6.36;44.3] SCOPA-AUT Subdomain Scores  SCOPA gastrointestinala 0.99 (±1.18) 2.61 (±2.40) 1.82 [1.45;2.28] <0.0001 1.87 [1.47;2.37] <0.0001   SCOPA gastrointestinal ≥ 1 57 52.29 81 88.04 6.72 [3.23;14.0] <0.0001 6.63 [3.01;14.6] <0.0001  SCOPA urinarya 1.99 (±2.43) 3.87 (±2.92) 1.32 [1.17;1.50] <0.0001 1.26 [1.11;1.44] 0.0003   SCOPA urinary ≥ 1 73 66.97 85 92.39 5.99 [2.51;14.3] <0.0001 5.25 [2.13;12.9] <0.0001  SCOPA cardiovasculara 0.23 (±0.52) 0.89 (±1.17) 2.93 [1.83;4.69] <0.0001 2.91 [1.77;4.76] <0.0001   SCOPA cardiovascular ≥ 1 21 19.44 48 52.17 4.52 [2.41;8.47] <0.0001 4.43 [2.26;8.66] <0.0001  SCOPA thermoregulatorya 1.30 (±1.75) 3.02 (±2.39) 1.54 [1.30;1.82] <0.0001 1.42 [1.19;1.68] <0.0001   SCOPA thermoregulatory ≥ 1 54 50.00 80 86.96 6.67 [3.26;13.6] <0.0001 5.82 [2.77;12.2] <0.0001  SCOPA pupillomotora 0.46 (±0.69) 1.02 (±0.97) 2.24 [1.54;3.24] <0.0001 2.31 [1.54;3.46] <0.0001   SCOPA pupillomotor ≥ 1 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002 3.44 [1.83;6.48] 0.0001  SCOPA sexual in mena 0.27 (±0.65) 1.02 (±1.49) 2.11 [1.28;3.47] 0.003 1.82 [1.10;3.00] 0.02   SCOPA sexual in men ≥ 1 9 16.07 25 48.08 4.84 [1.97;11.9] 0.0006 3.82 [1.47;9.91] 0.006  SCOPA sexual in womena 0.49 (±0.84) 2.05 (±1.50) 2.92 [1.61;5.28] 0.0004 2.80 [1.54;5.08] 0.0007   SCOPA sexual in women ≥ 1 13 35.14 17 80.95 7.84 [2.18;28.3] 0.002 7.16 [1.93;26.5] 0.003 . Controls, N = 109 . NT1, N = 92 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . SCOPA-AUT Total Score  SCOPA-AUT Total Scorea 5.36 (±5.05) 12.92 (±6.84) 1.23 [1.16;1.31] <0.0001 1.21 [1.14;1.29] <0.0001  SCOPA-AUT Total Scoreb   <4 50 46.30 7 7.61 1 <0.0001 1 <0.0001   [4–11] 39 36.11 29 31.52 5.31 [2.11;13.4] 4.79 [1.85;12.4]   ≥14 19 17.59 56 60.87 21.05[8.17;54.3] 16.77 [6.36;44.3] SCOPA-AUT Subdomain Scores  SCOPA gastrointestinala 0.99 (±1.18) 2.61 (±2.40) 1.82 [1.45;2.28] <0.0001 1.87 [1.47;2.37] <0.0001   SCOPA gastrointestinal ≥ 1 57 52.29 81 88.04 6.72 [3.23;14.0] <0.0001 6.63 [3.01;14.6] <0.0001  SCOPA urinarya 1.99 (±2.43) 3.87 (±2.92) 1.32 [1.17;1.50] <0.0001 1.26 [1.11;1.44] 0.0003   SCOPA urinary ≥ 1 73 66.97 85 92.39 5.99 [2.51;14.3] <0.0001 5.25 [2.13;12.9] <0.0001  SCOPA cardiovasculara 0.23 (±0.52) 0.89 (±1.17) 2.93 [1.83;4.69] <0.0001 2.91 [1.77;4.76] <0.0001   SCOPA cardiovascular ≥ 1 21 19.44 48 52.17 4.52 [2.41;8.47] <0.0001 4.43 [2.26;8.66] <0.0001  SCOPA thermoregulatorya 1.30 (±1.75) 3.02 (±2.39) 1.54 [1.30;1.82] <0.0001 1.42 [1.19;1.68] <0.0001   SCOPA thermoregulatory ≥ 1 54 50.00 80 86.96 6.67 [3.26;13.6] <0.0001 5.82 [2.77;12.2] <0.0001  SCOPA pupillomotora 0.46 (±0.69) 1.02 (±0.97) 2.24 [1.54;3.24] <0.0001 2.31 [1.54;3.46] <0.0001   SCOPA pupillomotor ≥ 1 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002 3.44 [1.83;6.48] 0.0001  SCOPA sexual in mena 0.27 (±0.65) 1.02 (±1.49) 2.11 [1.28;3.47] 0.003 1.82 [1.10;3.00] 0.02   SCOPA sexual in men ≥ 1 9 16.07 25 48.08 4.84 [1.97;11.9] 0.0006 3.82 [1.47;9.91] 0.006  SCOPA sexual in womena 0.49 (±0.84) 2.05 (±1.50) 2.92 [1.61;5.28] 0.0004 2.80 [1.54;5.08] 0.0007   SCOPA sexual in women ≥ 1 13 35.14 17 80.95 7.84 [2.18;28.3] 0.002 7.16 [1.93;26.5] 0.003 Model 1: crude association. Model 2: adjustment for BMI. aContinuous variables are expressed as means (±SD). bTertiles of the whole population study (NT1 and controls). Open in new tab Table 2. SCOPA-AUT Total Scores and Subdomain Scores in drug-free patients with NT1 compared to controls . Controls, N = 109 . NT1, N = 92 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . SCOPA-AUT Total Score  SCOPA-AUT Total Scorea 5.36 (±5.05) 12.92 (±6.84) 1.23 [1.16;1.31] <0.0001 1.21 [1.14;1.29] <0.0001  SCOPA-AUT Total Scoreb   <4 50 46.30 7 7.61 1 <0.0001 1 <0.0001   [4–11] 39 36.11 29 31.52 5.31 [2.11;13.4] 4.79 [1.85;12.4]   ≥14 19 17.59 56 60.87 21.05[8.17;54.3] 16.77 [6.36;44.3] SCOPA-AUT Subdomain Scores  SCOPA gastrointestinala 0.99 (±1.18) 2.61 (±2.40) 1.82 [1.45;2.28] <0.0001 1.87 [1.47;2.37] <0.0001   SCOPA gastrointestinal ≥ 1 57 52.29 81 88.04 6.72 [3.23;14.0] <0.0001 6.63 [3.01;14.6] <0.0001  SCOPA urinarya 1.99 (±2.43) 3.87 (±2.92) 1.32 [1.17;1.50] <0.0001 1.26 [1.11;1.44] 0.0003   SCOPA urinary ≥ 1 73 66.97 85 92.39 5.99 [2.51;14.3] <0.0001 5.25 [2.13;12.9] <0.0001  SCOPA cardiovasculara 0.23 (±0.52) 0.89 (±1.17) 2.93 [1.83;4.69] <0.0001 2.91 [1.77;4.76] <0.0001   SCOPA cardiovascular ≥ 1 21 19.44 48 52.17 4.52 [2.41;8.47] <0.0001 4.43 [2.26;8.66] <0.0001  SCOPA thermoregulatorya 1.30 (±1.75) 3.02 (±2.39) 1.54 [1.30;1.82] <0.0001 1.42 [1.19;1.68] <0.0001   SCOPA thermoregulatory ≥ 1 54 50.00 80 86.96 6.67 [3.26;13.6] <0.0001 5.82 [2.77;12.2] <0.0001  SCOPA pupillomotora 0.46 (±0.69) 1.02 (±0.97) 2.24 [1.54;3.24] <0.0001 2.31 [1.54;3.46] <0.0001   SCOPA pupillomotor ≥ 1 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002 3.44 [1.83;6.48] 0.0001  SCOPA sexual in mena 0.27 (±0.65) 1.02 (±1.49) 2.11 [1.28;3.47] 0.003 1.82 [1.10;3.00] 0.02   SCOPA sexual in men ≥ 1 9 16.07 25 48.08 4.84 [1.97;11.9] 0.0006 3.82 [1.47;9.91] 0.006  SCOPA sexual in womena 0.49 (±0.84) 2.05 (±1.50) 2.92 [1.61;5.28] 0.0004 2.80 [1.54;5.08] 0.0007   SCOPA sexual in women ≥ 1 13 35.14 17 80.95 7.84 [2.18;28.3] 0.002 7.16 [1.93;26.5] 0.003 . Controls, N = 109 . NT1, N = 92 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . SCOPA-AUT Total Score  SCOPA-AUT Total Scorea 5.36 (±5.05) 12.92 (±6.84) 1.23 [1.16;1.31] <0.0001 1.21 [1.14;1.29] <0.0001  SCOPA-AUT Total Scoreb   <4 50 46.30 7 7.61 1 <0.0001 1 <0.0001   [4–11] 39 36.11 29 31.52 5.31 [2.11;13.4] 4.79 [1.85;12.4]   ≥14 19 17.59 56 60.87 21.05[8.17;54.3] 16.77 [6.36;44.3] SCOPA-AUT Subdomain Scores  SCOPA gastrointestinala 0.99 (±1.18) 2.61 (±2.40) 1.82 [1.45;2.28] <0.0001 1.87 [1.47;2.37] <0.0001   SCOPA gastrointestinal ≥ 1 57 52.29 81 88.04 6.72 [3.23;14.0] <0.0001 6.63 [3.01;14.6] <0.0001  SCOPA urinarya 1.99 (±2.43) 3.87 (±2.92) 1.32 [1.17;1.50] <0.0001 1.26 [1.11;1.44] 0.0003   SCOPA urinary ≥ 1 73 66.97 85 92.39 5.99 [2.51;14.3] <0.0001 5.25 [2.13;12.9] <0.0001  SCOPA cardiovasculara 0.23 (±0.52) 0.89 (±1.17) 2.93 [1.83;4.69] <0.0001 2.91 [1.77;4.76] <0.0001   SCOPA cardiovascular ≥ 1 21 19.44 48 52.17 4.52 [2.41;8.47] <0.0001 4.43 [2.26;8.66] <0.0001  SCOPA thermoregulatorya 1.30 (±1.75) 3.02 (±2.39) 1.54 [1.30;1.82] <0.0001 1.42 [1.19;1.68] <0.0001   SCOPA thermoregulatory ≥ 1 54 50.00 80 86.96 6.67 [3.26;13.6] <0.0001 5.82 [2.77;12.2] <0.0001  SCOPA pupillomotora 0.46 (±0.69) 1.02 (±0.97) 2.24 [1.54;3.24] <0.0001 2.31 [1.54;3.46] <0.0001   SCOPA pupillomotor ≥ 1 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002 3.44 [1.83;6.48] 0.0001  SCOPA sexual in mena 0.27 (±0.65) 1.02 (±1.49) 2.11 [1.28;3.47] 0.003 1.82 [1.10;3.00] 0.02   SCOPA sexual in men ≥ 1 9 16.07 25 48.08 4.84 [1.97;11.9] 0.0006 3.82 [1.47;9.91] 0.006  SCOPA sexual in womena 0.49 (±0.84) 2.05 (±1.50) 2.92 [1.61;5.28] 0.0004 2.80 [1.54;5.08] 0.0007   SCOPA sexual in women ≥ 1 13 35.14 17 80.95 7.84 [2.18;28.3] 0.002 7.16 [1.93;26.5] 0.003 Model 1: crude association. Model 2: adjustment for BMI. aContinuous variables are expressed as means (±SD). bTertiles of the whole population study (NT1 and controls). Open in new tab Patients experienced more problems than controls in the six subdomains of the scale, with major differences between groups in crude and adjusted models (Table 2). Similarly, the proportion of narcoleptic patients with a score greater than zero was also higher than controls in every subdomain, even after adjustment on BMI. We found that 88% of patients reported symptoms consistent with autonomic gastrointestinal dysfunctions, 92.4% with urinary, 52.2% with cardiovascular, 87% with thermoregulatory, 64.2% with pupillomotor, and from 48% (in men) to 81% (in women) with sexual dysfunctions (Table 2). The proportions of subjects in the two groups with a score greater than zero for each of the 25 items of the scale are shown in Table 3. Differences were significant for 15 items, always higher in NT1, and persisted after adjustment on BMI. Among the three questions focusing on the cardiovascular subdomain: two items (i.e. items 14 and 15 for lightheaded standing up and lightheaded standing for some time, respectively) differed between the groups with higher scores in patients with NT1 than controls, but not the one related to the presence of syncope (i.e. item 16). Table 3. Score greater than zero for each item of SCOPA-AUT of drug-free patients with NT1 compared to controls . Controls, N = 109 . NT1, N = 92 . Model 1 . n . % . n . % . OR [95% CI] . p-value . 1. Swallowing/choking 12 11.01 16 17.39 1.70 [0.76;3.81] 0.20 2. Drooling 13 11.93 33 35.87 4.13 [2.01;8.48] 0.0001* 3. Dysphagia 13 11.93 12 13.04 1.11 [0.48;2.56] 0.81 4. Early abdominal fullness 28 25.69 51 55.43 3.60 [1.99;6.52] <0.0001* 5. Constipation 16 14.68 31 33.70 2.95 [1.49;5.86] 0.002* 6. Straining for defecation 20 18.35 33 35.87 2.49 [1.31;4.75] 0.006* 7. Fecal incontinence 2 1.83 6 6.52 3.73 [0.73;19.0] 0.11 8. Urinary urgency 17 15.60 23 25.00 1.80 [0.90;3.63] 0.10 9. Urinary incontinence 9 8.26 17 18.48 2.52 [1.06;5.96] 0.04 10. Incomplete emptying 17 15.60 35 38.04 3.32 [1.71;6.47] 0.0004* 11. Weak stream of urine 24 22.02 27 29.35 1.47 [0.78;2.78] 0.24 12. Frequency 58 53.21 55 59.78 1.31 [0.75;2.29] 0.35 13. Nocturia 48 44.04 78 84.78 7.08 [3.58;14.0] <0.0001* 14. Lightheaded standing up 13 12.04 37 40.22 4.92 [2.41;10.0] <0.0001* 15. Lightheaded standing for sometime 9 8.33 22 23.91 3.46 [1.50;7.96] 0.004* 16. Syncope 2 1.85 4 4.35 2.41 [0.43;13.5] 0.32 17. Hyperhidrosis during the day 20 18.52 46 50.00 4.40 [2.33;8.30] <0.0001* 18. Hyperhidrosis during the night 29 26.85 37 40.22 1.83 [1.01;3.32] 0.05 19. Oversensitive to bright light 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002* 20. Cold intolerance 39 35.78 52 56.52 2.33 [1.32;4.12] 0.004* 21. Heat intolerance 23 21.10 47 51.09 3.90 [2.11;7.23] <0.0001* 22. Erection problem-male 6 10.71 20 37.74 5.05 [1.83;13.9] 0.002* 23. Ejaculation problem-male 8 14.29 13 25.00 2.00 [0.75;5.31] 0.16 24. Vaginal lubrication-female 7 17.50 14 63.64 8.25 [2.51;27.2] 0.0005* 25. Problem with orgasm-female 10 27.03 14 66.67 5.40 [1.69;17.3] 0.004* . Controls, N = 109 . NT1, N = 92 . Model 1 . n . % . n . % . OR [95% CI] . p-value . 1. Swallowing/choking 12 11.01 16 17.39 1.70 [0.76;3.81] 0.20 2. Drooling 13 11.93 33 35.87 4.13 [2.01;8.48] 0.0001* 3. Dysphagia 13 11.93 12 13.04 1.11 [0.48;2.56] 0.81 4. Early abdominal fullness 28 25.69 51 55.43 3.60 [1.99;6.52] <0.0001* 5. Constipation 16 14.68 31 33.70 2.95 [1.49;5.86] 0.002* 6. Straining for defecation 20 18.35 33 35.87 2.49 [1.31;4.75] 0.006* 7. Fecal incontinence 2 1.83 6 6.52 3.73 [0.73;19.0] 0.11 8. Urinary urgency 17 15.60 23 25.00 1.80 [0.90;3.63] 0.10 9. Urinary incontinence 9 8.26 17 18.48 2.52 [1.06;5.96] 0.04 10. Incomplete emptying 17 15.60 35 38.04 3.32 [1.71;6.47] 0.0004* 11. Weak stream of urine 24 22.02 27 29.35 1.47 [0.78;2.78] 0.24 12. Frequency 58 53.21 55 59.78 1.31 [0.75;2.29] 0.35 13. Nocturia 48 44.04 78 84.78 7.08 [3.58;14.0] <0.0001* 14. Lightheaded standing up 13 12.04 37 40.22 4.92 [2.41;10.0] <0.0001* 15. Lightheaded standing for sometime 9 8.33 22 23.91 3.46 [1.50;7.96] 0.004* 16. Syncope 2 1.85 4 4.35 2.41 [0.43;13.5] 0.32 17. Hyperhidrosis during the day 20 18.52 46 50.00 4.40 [2.33;8.30] <0.0001* 18. Hyperhidrosis during the night 29 26.85 37 40.22 1.83 [1.01;3.32] 0.05 19. Oversensitive to bright light 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002* 20. Cold intolerance 39 35.78 52 56.52 2.33 [1.32;4.12] 0.004* 21. Heat intolerance 23 21.10 47 51.09 3.90 [2.11;7.23] <0.0001* 22. Erection problem-male 6 10.71 20 37.74 5.05 [1.83;13.9] 0.002* 23. Ejaculation problem-male 8 14.29 13 25.00 2.00 [0.75;5.31] 0.16 24. Vaginal lubrication-female 7 17.50 14 63.64 8.25 [2.51;27.2] 0.0005* 25. Problem with orgasm-female 10 27.03 14 66.67 5.40 [1.69;17.3] 0.004* Model 1: crude association. *Still significant after adjustment for BMI. Open in new tab Table 3. Score greater than zero for each item of SCOPA-AUT of drug-free patients with NT1 compared to controls . Controls, N = 109 . NT1, N = 92 . Model 1 . n . % . n . % . OR [95% CI] . p-value . 1. Swallowing/choking 12 11.01 16 17.39 1.70 [0.76;3.81] 0.20 2. Drooling 13 11.93 33 35.87 4.13 [2.01;8.48] 0.0001* 3. Dysphagia 13 11.93 12 13.04 1.11 [0.48;2.56] 0.81 4. Early abdominal fullness 28 25.69 51 55.43 3.60 [1.99;6.52] <0.0001* 5. Constipation 16 14.68 31 33.70 2.95 [1.49;5.86] 0.002* 6. Straining for defecation 20 18.35 33 35.87 2.49 [1.31;4.75] 0.006* 7. Fecal incontinence 2 1.83 6 6.52 3.73 [0.73;19.0] 0.11 8. Urinary urgency 17 15.60 23 25.00 1.80 [0.90;3.63] 0.10 9. Urinary incontinence 9 8.26 17 18.48 2.52 [1.06;5.96] 0.04 10. Incomplete emptying 17 15.60 35 38.04 3.32 [1.71;6.47] 0.0004* 11. Weak stream of urine 24 22.02 27 29.35 1.47 [0.78;2.78] 0.24 12. Frequency 58 53.21 55 59.78 1.31 [0.75;2.29] 0.35 13. Nocturia 48 44.04 78 84.78 7.08 [3.58;14.0] <0.0001* 14. Lightheaded standing up 13 12.04 37 40.22 4.92 [2.41;10.0] <0.0001* 15. Lightheaded standing for sometime 9 8.33 22 23.91 3.46 [1.50;7.96] 0.004* 16. Syncope 2 1.85 4 4.35 2.41 [0.43;13.5] 0.32 17. Hyperhidrosis during the day 20 18.52 46 50.00 4.40 [2.33;8.30] <0.0001* 18. Hyperhidrosis during the night 29 26.85 37 40.22 1.83 [1.01;3.32] 0.05 19. Oversensitive to bright light 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002* 20. Cold intolerance 39 35.78 52 56.52 2.33 [1.32;4.12] 0.004* 21. Heat intolerance 23 21.10 47 51.09 3.90 [2.11;7.23] <0.0001* 22. Erection problem-male 6 10.71 20 37.74 5.05 [1.83;13.9] 0.002* 23. Ejaculation problem-male 8 14.29 13 25.00 2.00 [0.75;5.31] 0.16 24. Vaginal lubrication-female 7 17.50 14 63.64 8.25 [2.51;27.2] 0.0005* 25. Problem with orgasm-female 10 27.03 14 66.67 5.40 [1.69;17.3] 0.004* . Controls, N = 109 . NT1, N = 92 . Model 1 . n . % . n . % . OR [95% CI] . p-value . 1. Swallowing/choking 12 11.01 16 17.39 1.70 [0.76;3.81] 0.20 2. Drooling 13 11.93 33 35.87 4.13 [2.01;8.48] 0.0001* 3. Dysphagia 13 11.93 12 13.04 1.11 [0.48;2.56] 0.81 4. Early abdominal fullness 28 25.69 51 55.43 3.60 [1.99;6.52] <0.0001* 5. Constipation 16 14.68 31 33.70 2.95 [1.49;5.86] 0.002* 6. Straining for defecation 20 18.35 33 35.87 2.49 [1.31;4.75] 0.006* 7. Fecal incontinence 2 1.83 6 6.52 3.73 [0.73;19.0] 0.11 8. Urinary urgency 17 15.60 23 25.00 1.80 [0.90;3.63] 0.10 9. Urinary incontinence 9 8.26 17 18.48 2.52 [1.06;5.96] 0.04 10. Incomplete emptying 17 15.60 35 38.04 3.32 [1.71;6.47] 0.0004* 11. Weak stream of urine 24 22.02 27 29.35 1.47 [0.78;2.78] 0.24 12. Frequency 58 53.21 55 59.78 1.31 [0.75;2.29] 0.35 13. Nocturia 48 44.04 78 84.78 7.08 [3.58;14.0] <0.0001* 14. Lightheaded standing up 13 12.04 37 40.22 4.92 [2.41;10.0] <0.0001* 15. Lightheaded standing for sometime 9 8.33 22 23.91 3.46 [1.50;7.96] 0.004* 16. Syncope 2 1.85 4 4.35 2.41 [0.43;13.5] 0.32 17. Hyperhidrosis during the day 20 18.52 46 50.00 4.40 [2.33;8.30] <0.0001* 18. Hyperhidrosis during the night 29 26.85 37 40.22 1.83 [1.01;3.32] 0.05 19. Oversensitive to bright light 40 37.04 59 64.13 3.04 [1.71;5.42] 0.0002* 20. Cold intolerance 39 35.78 52 56.52 2.33 [1.32;4.12] 0.004* 21. Heat intolerance 23 21.10 47 51.09 3.90 [2.11;7.23] <0.0001* 22. Erection problem-male 6 10.71 20 37.74 5.05 [1.83;13.9] 0.002* 23. Ejaculation problem-male 8 14.29 13 25.00 2.00 [0.75;5.31] 0.16 24. Vaginal lubrication-female 7 17.50 14 63.64 8.25 [2.51;27.2] 0.0005* 25. Problem with orgasm-female 10 27.03 14 66.67 5.40 [1.69;17.3] 0.004* Model 1: crude association. *Still significant after adjustment for BMI. Open in new tab Determinants of high SCOPA-AUT scores in untreated NT1 patients Narcoleptic patients with the highest SCOPA-AUT scores (≥15, i.e. the highest tertile of patients with NT1) were older, had a longer disease duration, more depressive symptoms, and a poorer quality of life (on EQ-5D Utility), in crude and adjusted model on age (Table 4, models 1 and 2). Similar trends were found for clinical RBD and insomnia symptoms. No associations were found for metabolic and cardiovascular comorbidities, severity of NT1 assessed by the NSS, CSF ORX-A levels, REM sleep without atonia, and PLMs and apnea-hypopnea indexes assessed continuously or by category (Table 4). We found no differences on SCOPA-AUT cardiovascular subscore between NT1 patients with and without clinically diagnosed RBD or REM sleep without atonia. No differences on SCOPA-AUT total or cardiovascular scores were found between patients with undetectable (<10 pg/ml, n = 9) vs. low (10–110 pg/ml, n = 62) CSF ORX-A levels. Table 4. Determinants of a high total SCOPA-AUT score in drug-free patients with NT1 . SCOPA-AUT Total Score . . . . <15 N = 58 . ≥15a N = 34 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . Clinical and biological characteristics  Gender, men 40 68.97 19 55.88 1 0.21 1 0.19  Age (years)b 35.78 (±13.80) 44.85 (±17.22) 1.04 [1.01;1.07] 0.01  BMI (kg/m2)b 26.67 (±4.11) 27.48 (±5.52) 1.04 [0.95;1.14] 0.43 1.02 [0.93;1.12] 0.68  Age at onset of EDS (years)b 24.12 (±9.78) 26.06 (±13.57) 1.02 [0.98;1.05] 0.43 0.99 [0.95;1.04] 0.72  Duration of narcolepsy (years)b 11.66 (±12.87) 18.88 (±17.23) 1.03 [1.00;1.06] 0.03  REM sleep behavior disorder, yes 17 33.33 14 53.85 2.33 [0.89;6.13] 0.09 2.76 [0.97;7.83] 0.06  CSF ORX-A levels (pg/mL)b 23.98 (±23.36) 34.91 (±36.67) 1.01 [1.00;1.03] 0.13 1.01 [0.99;1.03] 0.16 Metabolic and cardiovascular comorbidities  Cardiac disorders, yes 2 3.70 3 10.00 3.93 [0.91;16.9] 0.07 1.71 [0.31;9.28] 0.54  Hypertension, yes 3 5.56 5 16.67 3.40 [0.75;15.4] 0.11 1.51 [0.27;8.55] 0.64  Diabetes type 2, yes 3 5.45 3 10.00 1.93 [0.36;10.2] 0.44 0.62 [0.09;4.43] 0.63  Dyslipidemia, yes 8 14.29 10 33.33 3.00 [1.03;8.71] 0.04 1.87 [0.58;6.09] 0.30 Self-reported questionnaires  ESS scoreb 18.02 (±3.69) 17.94 (±4.19) 0.99 [0.89;1.11] 0.93 0.98 [0.87;1.10] 0.76  ISI scoreb 14.30 (±5.59) 16.77 (±5.63) 1.09 [1.00;1.18] 0.06 1.09 [1.00;1.20] 0.06  EQ5D-Utilityb 0.81 (±0.13) 0.63 (±0.21) 0.00 [0.00;0.05] 0.0005 0.00 [0.00;0.09] 0.002  EQ5D-VASb 61.44 (±19.55) 53.04 (±19.82) 0.98 [0.95;1.00] 0.08 0.98 [0.95;1.00] 0.08  BDI scoreb 16.07 (±8.73) 21.91 (±11.78) 1.06 [1.01;1.11] 0.01 1.06 [1.01;1.11] 0.02  NSS scoreb 32.10 (±9.23) 36.25 (±10.50) 1.05 [0.99;1.11] 0.12 1.04 [1.01–1.07] 0.15 Polysomnographic characteristics  Total sleep time (min)b 400.06 (±77.62) 402.18 (±94.16) 1.00 [0.99;1.01] 0.92 1.00 [1.00;1.01] 0.38  Sleep efficiency (%)b 80.45 (±12.42) 79.57 (±16.46) 1.00 [0.96;1.03] 0.80 1.01 [0.97;1.05] 0.59  Stage 2 (%)b 47.83 (±8.98) 48.82 (±9.98) 1.01 [0.96;1.07] 0.67 1.02 [0.96;1.08] 0.55  Stage 3 (%)b 18.70 (±7.77) 15.69 (±6.51) 0.94 [0.87;1.02] 0.12 0.97 [0.89;1.04] 0.37  REM sleep (%)b 22.74 (±8.09) 24.68 (±8.68) 1.03 [0.97;1.09] 0.36 1.02 [0.96;1.09] 0.55  Wake after sleep onset (min)b 83.96 (±49.20) 89.55 (±65.22) 1.00 [0.99;1.01] 0.68 1.00 [0.99;1.01] 0.76  AHI/hourb 9.05 (±10.31) 7.44 (±7.68) 0.98 [0.93;1.04] 0.51 0.95 [0.89;1.02] 0.18  PLMs index/hourb 12.64 (±14.18) 18.76 (±35.87) 1.01 [0.99;1.03] 0.33 1.00 [0.98;1.03] 0.80  Microarousal index/hourb 19.34 (±9.78) 21.26 (±12.97) 1.02 [0.97;1.06] 0.48 1.00 [0.95;1.05] 0.93 MSLT  Mean sleep latency (min)b 4.25 (±3.26) 5.59 (±3.80) 1.12 [0.96;1.30] 0.15 1.09 [0.94;1.27] 0.26  Number of SOREMb 3.69 (±1.33) 3.24 (±1.34) 0.78 [0.53;1.14] 0.19 0.89 [0.58;1.35] 0.57 EMG activity in REM sleep  Phasic EMG activity (%)b 5.22 (±4.78) 5.59 (±5.69) 1.01 [0.92;1.12] 0.77 1.00 [0.90;1.11] 0.97  Tonic EMG activity (%)b 7.45 (±10.63) 6.23 (±7.77) 0.99 [0.93;1.04] 0.63 0.99 [0.93;1.05] 0.67 . SCOPA-AUT Total Score . . . . <15 N = 58 . ≥15a N = 34 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . Clinical and biological characteristics  Gender, men 40 68.97 19 55.88 1 0.21 1 0.19  Age (years)b 35.78 (±13.80) 44.85 (±17.22) 1.04 [1.01;1.07] 0.01  BMI (kg/m2)b 26.67 (±4.11) 27.48 (±5.52) 1.04 [0.95;1.14] 0.43 1.02 [0.93;1.12] 0.68  Age at onset of EDS (years)b 24.12 (±9.78) 26.06 (±13.57) 1.02 [0.98;1.05] 0.43 0.99 [0.95;1.04] 0.72  Duration of narcolepsy (years)b 11.66 (±12.87) 18.88 (±17.23) 1.03 [1.00;1.06] 0.03  REM sleep behavior disorder, yes 17 33.33 14 53.85 2.33 [0.89;6.13] 0.09 2.76 [0.97;7.83] 0.06  CSF ORX-A levels (pg/mL)b 23.98 (±23.36) 34.91 (±36.67) 1.01 [1.00;1.03] 0.13 1.01 [0.99;1.03] 0.16 Metabolic and cardiovascular comorbidities  Cardiac disorders, yes 2 3.70 3 10.00 3.93 [0.91;16.9] 0.07 1.71 [0.31;9.28] 0.54  Hypertension, yes 3 5.56 5 16.67 3.40 [0.75;15.4] 0.11 1.51 [0.27;8.55] 0.64  Diabetes type 2, yes 3 5.45 3 10.00 1.93 [0.36;10.2] 0.44 0.62 [0.09;4.43] 0.63  Dyslipidemia, yes 8 14.29 10 33.33 3.00 [1.03;8.71] 0.04 1.87 [0.58;6.09] 0.30 Self-reported questionnaires  ESS scoreb 18.02 (±3.69) 17.94 (±4.19) 0.99 [0.89;1.11] 0.93 0.98 [0.87;1.10] 0.76  ISI scoreb 14.30 (±5.59) 16.77 (±5.63) 1.09 [1.00;1.18] 0.06 1.09 [1.00;1.20] 0.06  EQ5D-Utilityb 0.81 (±0.13) 0.63 (±0.21) 0.00 [0.00;0.05] 0.0005 0.00 [0.00;0.09] 0.002  EQ5D-VASb 61.44 (±19.55) 53.04 (±19.82) 0.98 [0.95;1.00] 0.08 0.98 [0.95;1.00] 0.08  BDI scoreb 16.07 (±8.73) 21.91 (±11.78) 1.06 [1.01;1.11] 0.01 1.06 [1.01;1.11] 0.02  NSS scoreb 32.10 (±9.23) 36.25 (±10.50) 1.05 [0.99;1.11] 0.12 1.04 [1.01–1.07] 0.15 Polysomnographic characteristics  Total sleep time (min)b 400.06 (±77.62) 402.18 (±94.16) 1.00 [0.99;1.01] 0.92 1.00 [1.00;1.01] 0.38  Sleep efficiency (%)b 80.45 (±12.42) 79.57 (±16.46) 1.00 [0.96;1.03] 0.80 1.01 [0.97;1.05] 0.59  Stage 2 (%)b 47.83 (±8.98) 48.82 (±9.98) 1.01 [0.96;1.07] 0.67 1.02 [0.96;1.08] 0.55  Stage 3 (%)b 18.70 (±7.77) 15.69 (±6.51) 0.94 [0.87;1.02] 0.12 0.97 [0.89;1.04] 0.37  REM sleep (%)b 22.74 (±8.09) 24.68 (±8.68) 1.03 [0.97;1.09] 0.36 1.02 [0.96;1.09] 0.55  Wake after sleep onset (min)b 83.96 (±49.20) 89.55 (±65.22) 1.00 [0.99;1.01] 0.68 1.00 [0.99;1.01] 0.76  AHI/hourb 9.05 (±10.31) 7.44 (±7.68) 0.98 [0.93;1.04] 0.51 0.95 [0.89;1.02] 0.18  PLMs index/hourb 12.64 (±14.18) 18.76 (±35.87) 1.01 [0.99;1.03] 0.33 1.00 [0.98;1.03] 0.80  Microarousal index/hourb 19.34 (±9.78) 21.26 (±12.97) 1.02 [0.97;1.06] 0.48 1.00 [0.95;1.05] 0.93 MSLT  Mean sleep latency (min)b 4.25 (±3.26) 5.59 (±3.80) 1.12 [0.96;1.30] 0.15 1.09 [0.94;1.27] 0.26  Number of SOREMb 3.69 (±1.33) 3.24 (±1.34) 0.78 [0.53;1.14] 0.19 0.89 [0.58;1.35] 0.57 EMG activity in REM sleep  Phasic EMG activity (%)b 5.22 (±4.78) 5.59 (±5.69) 1.01 [0.92;1.12] 0.77 1.00 [0.90;1.11] 0.97  Tonic EMG activity (%)b 7.45 (±10.63) 6.23 (±7.77) 0.99 [0.93;1.04] 0.63 0.99 [0.93;1.05] 0.67 AHI = apnea hypopnea index, ORX = orexin. Model 1: crude association. Model 2: adjustment for age. aHighest Tertile of SCOPA-AUT total score in NT1 group. bContinuous variables are expressed as means (±SD). Open in new tab Table 4. Determinants of a high total SCOPA-AUT score in drug-free patients with NT1 . SCOPA-AUT Total Score . . . . <15 N = 58 . ≥15a N = 34 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . Clinical and biological characteristics  Gender, men 40 68.97 19 55.88 1 0.21 1 0.19  Age (years)b 35.78 (±13.80) 44.85 (±17.22) 1.04 [1.01;1.07] 0.01  BMI (kg/m2)b 26.67 (±4.11) 27.48 (±5.52) 1.04 [0.95;1.14] 0.43 1.02 [0.93;1.12] 0.68  Age at onset of EDS (years)b 24.12 (±9.78) 26.06 (±13.57) 1.02 [0.98;1.05] 0.43 0.99 [0.95;1.04] 0.72  Duration of narcolepsy (years)b 11.66 (±12.87) 18.88 (±17.23) 1.03 [1.00;1.06] 0.03  REM sleep behavior disorder, yes 17 33.33 14 53.85 2.33 [0.89;6.13] 0.09 2.76 [0.97;7.83] 0.06  CSF ORX-A levels (pg/mL)b 23.98 (±23.36) 34.91 (±36.67) 1.01 [1.00;1.03] 0.13 1.01 [0.99;1.03] 0.16 Metabolic and cardiovascular comorbidities  Cardiac disorders, yes 2 3.70 3 10.00 3.93 [0.91;16.9] 0.07 1.71 [0.31;9.28] 0.54  Hypertension, yes 3 5.56 5 16.67 3.40 [0.75;15.4] 0.11 1.51 [0.27;8.55] 0.64  Diabetes type 2, yes 3 5.45 3 10.00 1.93 [0.36;10.2] 0.44 0.62 [0.09;4.43] 0.63  Dyslipidemia, yes 8 14.29 10 33.33 3.00 [1.03;8.71] 0.04 1.87 [0.58;6.09] 0.30 Self-reported questionnaires  ESS scoreb 18.02 (±3.69) 17.94 (±4.19) 0.99 [0.89;1.11] 0.93 0.98 [0.87;1.10] 0.76  ISI scoreb 14.30 (±5.59) 16.77 (±5.63) 1.09 [1.00;1.18] 0.06 1.09 [1.00;1.20] 0.06  EQ5D-Utilityb 0.81 (±0.13) 0.63 (±0.21) 0.00 [0.00;0.05] 0.0005 0.00 [0.00;0.09] 0.002  EQ5D-VASb 61.44 (±19.55) 53.04 (±19.82) 0.98 [0.95;1.00] 0.08 0.98 [0.95;1.00] 0.08  BDI scoreb 16.07 (±8.73) 21.91 (±11.78) 1.06 [1.01;1.11] 0.01 1.06 [1.01;1.11] 0.02  NSS scoreb 32.10 (±9.23) 36.25 (±10.50) 1.05 [0.99;1.11] 0.12 1.04 [1.01–1.07] 0.15 Polysomnographic characteristics  Total sleep time (min)b 400.06 (±77.62) 402.18 (±94.16) 1.00 [0.99;1.01] 0.92 1.00 [1.00;1.01] 0.38  Sleep efficiency (%)b 80.45 (±12.42) 79.57 (±16.46) 1.00 [0.96;1.03] 0.80 1.01 [0.97;1.05] 0.59  Stage 2 (%)b 47.83 (±8.98) 48.82 (±9.98) 1.01 [0.96;1.07] 0.67 1.02 [0.96;1.08] 0.55  Stage 3 (%)b 18.70 (±7.77) 15.69 (±6.51) 0.94 [0.87;1.02] 0.12 0.97 [0.89;1.04] 0.37  REM sleep (%)b 22.74 (±8.09) 24.68 (±8.68) 1.03 [0.97;1.09] 0.36 1.02 [0.96;1.09] 0.55  Wake after sleep onset (min)b 83.96 (±49.20) 89.55 (±65.22) 1.00 [0.99;1.01] 0.68 1.00 [0.99;1.01] 0.76  AHI/hourb 9.05 (±10.31) 7.44 (±7.68) 0.98 [0.93;1.04] 0.51 0.95 [0.89;1.02] 0.18  PLMs index/hourb 12.64 (±14.18) 18.76 (±35.87) 1.01 [0.99;1.03] 0.33 1.00 [0.98;1.03] 0.80  Microarousal index/hourb 19.34 (±9.78) 21.26 (±12.97) 1.02 [0.97;1.06] 0.48 1.00 [0.95;1.05] 0.93 MSLT  Mean sleep latency (min)b 4.25 (±3.26) 5.59 (±3.80) 1.12 [0.96;1.30] 0.15 1.09 [0.94;1.27] 0.26  Number of SOREMb 3.69 (±1.33) 3.24 (±1.34) 0.78 [0.53;1.14] 0.19 0.89 [0.58;1.35] 0.57 EMG activity in REM sleep  Phasic EMG activity (%)b 5.22 (±4.78) 5.59 (±5.69) 1.01 [0.92;1.12] 0.77 1.00 [0.90;1.11] 0.97  Tonic EMG activity (%)b 7.45 (±10.63) 6.23 (±7.77) 0.99 [0.93;1.04] 0.63 0.99 [0.93;1.05] 0.67 . SCOPA-AUT Total Score . . . . <15 N = 58 . ≥15a N = 34 . Model 1 . Model 2 . n . % . n . % . OR [95% CI] . p-value . OR [95% CI] . p-value . Clinical and biological characteristics  Gender, men 40 68.97 19 55.88 1 0.21 1 0.19  Age (years)b 35.78 (±13.80) 44.85 (±17.22) 1.04 [1.01;1.07] 0.01  BMI (kg/m2)b 26.67 (±4.11) 27.48 (±5.52) 1.04 [0.95;1.14] 0.43 1.02 [0.93;1.12] 0.68  Age at onset of EDS (years)b 24.12 (±9.78) 26.06 (±13.57) 1.02 [0.98;1.05] 0.43 0.99 [0.95;1.04] 0.72  Duration of narcolepsy (years)b 11.66 (±12.87) 18.88 (±17.23) 1.03 [1.00;1.06] 0.03  REM sleep behavior disorder, yes 17 33.33 14 53.85 2.33 [0.89;6.13] 0.09 2.76 [0.97;7.83] 0.06  CSF ORX-A levels (pg/mL)b 23.98 (±23.36) 34.91 (±36.67) 1.01 [1.00;1.03] 0.13 1.01 [0.99;1.03] 0.16 Metabolic and cardiovascular comorbidities  Cardiac disorders, yes 2 3.70 3 10.00 3.93 [0.91;16.9] 0.07 1.71 [0.31;9.28] 0.54  Hypertension, yes 3 5.56 5 16.67 3.40 [0.75;15.4] 0.11 1.51 [0.27;8.55] 0.64  Diabetes type 2, yes 3 5.45 3 10.00 1.93 [0.36;10.2] 0.44 0.62 [0.09;4.43] 0.63  Dyslipidemia, yes 8 14.29 10 33.33 3.00 [1.03;8.71] 0.04 1.87 [0.58;6.09] 0.30 Self-reported questionnaires  ESS scoreb 18.02 (±3.69) 17.94 (±4.19) 0.99 [0.89;1.11] 0.93 0.98 [0.87;1.10] 0.76  ISI scoreb 14.30 (±5.59) 16.77 (±5.63) 1.09 [1.00;1.18] 0.06 1.09 [1.00;1.20] 0.06  EQ5D-Utilityb 0.81 (±0.13) 0.63 (±0.21) 0.00 [0.00;0.05] 0.0005 0.00 [0.00;0.09] 0.002  EQ5D-VASb 61.44 (±19.55) 53.04 (±19.82) 0.98 [0.95;1.00] 0.08 0.98 [0.95;1.00] 0.08  BDI scoreb 16.07 (±8.73) 21.91 (±11.78) 1.06 [1.01;1.11] 0.01 1.06 [1.01;1.11] 0.02  NSS scoreb 32.10 (±9.23) 36.25 (±10.50) 1.05 [0.99;1.11] 0.12 1.04 [1.01–1.07] 0.15 Polysomnographic characteristics  Total sleep time (min)b 400.06 (±77.62) 402.18 (±94.16) 1.00 [0.99;1.01] 0.92 1.00 [1.00;1.01] 0.38  Sleep efficiency (%)b 80.45 (±12.42) 79.57 (±16.46) 1.00 [0.96;1.03] 0.80 1.01 [0.97;1.05] 0.59  Stage 2 (%)b 47.83 (±8.98) 48.82 (±9.98) 1.01 [0.96;1.07] 0.67 1.02 [0.96;1.08] 0.55  Stage 3 (%)b 18.70 (±7.77) 15.69 (±6.51) 0.94 [0.87;1.02] 0.12 0.97 [0.89;1.04] 0.37  REM sleep (%)b 22.74 (±8.09) 24.68 (±8.68) 1.03 [0.97;1.09] 0.36 1.02 [0.96;1.09] 0.55  Wake after sleep onset (min)b 83.96 (±49.20) 89.55 (±65.22) 1.00 [0.99;1.01] 0.68 1.00 [0.99;1.01] 0.76  AHI/hourb 9.05 (±10.31) 7.44 (±7.68) 0.98 [0.93;1.04] 0.51 0.95 [0.89;1.02] 0.18  PLMs index/hourb 12.64 (±14.18) 18.76 (±35.87) 1.01 [0.99;1.03] 0.33 1.00 [0.98;1.03] 0.80  Microarousal index/hourb 19.34 (±9.78) 21.26 (±12.97) 1.02 [0.97;1.06] 0.48 1.00 [0.95;1.05] 0.93 MSLT  Mean sleep latency (min)b 4.25 (±3.26) 5.59 (±3.80) 1.12 [0.96;1.30] 0.15 1.09 [0.94;1.27] 0.26  Number of SOREMb 3.69 (±1.33) 3.24 (±1.34) 0.78 [0.53;1.14] 0.19 0.89 [0.58;1.35] 0.57 EMG activity in REM sleep  Phasic EMG activity (%)b 5.22 (±4.78) 5.59 (±5.69) 1.01 [0.92;1.12] 0.77 1.00 [0.90;1.11] 0.97  Tonic EMG activity (%)b 7.45 (±10.63) 6.23 (±7.77) 0.99 [0.93;1.04] 0.63 0.99 [0.93;1.05] 0.67 AHI = apnea hypopnea index, ORX = orexin. Model 1: crude association. Model 2: adjustment for age. aHighest Tertile of SCOPA-AUT total score in NT1 group. bContinuous variables are expressed as means (±SD). Open in new tab Effects of medications on autonomic symptoms in patients with NT1 Fifty-nine patients with NT1 were evaluated twice, untreated then treated, with a mean delay of 1.28 ± 1.14 years. There was no difference for BMI between the two conditions. In this dependent sample, the SCOPA-AUT total score (taken as continuous variable or in tertiles) did not differ significantly between groups, neither did each subdomain of the scale (Table 5). However, when each item of the questionnaire was analyzed separately, a worsening in sexual dysfunction was found in men only (ejaculation problem), with 42.9% of patients reporting such symptom under medication vs. 21.9% without treatment (p = 0.02). Table 5. Comparison between 59 patients with NT1 evaluated twice, in drug-free and treated condition Variable . Drug-free NT1 patients . Treated NT1 patients . p-value . n . % . n . % . SCOPA-AUT Total Scorea 12.64 (±6.98) 13.34 (±7.75) 0.40 SCOPA gastrointestinala 2.42 (±2.25) 2.53 (±2.49) 0.53  SCOPA gastrointestinal ≥ 1 51 86.44 45 76.27 0.06 SCOPA urinarya 3.78 (±3.10) 4.19 (±3.38) 0.48  SCOPA urinary ≥ 1 53 89.83 54 91.53 0.65 SCOPA cardiovasculara 0.93 (±1.28) 0.93 (±1.14) 0.89  SCOPA cardiovascular ≥ 1 30 50.85 32 54.24 0.62 SCOPA thermoregulatorya 3.19 (±2.67) 3.20 (±2.48) 0.87  SCOPA thermoregulatory ≥ 1 49 83.05 50 84.75 0.74 SCOPA pupillomotora 1.05 (±1.07) 1.05 (±1.01) 0.99  SCOPA pupillomotor ≥ 1 35 59.32 38 64.41 0.43 SCOPA sexual in womena 2.23 (±1.54) 1.54 (±1.20) 0.05  SCOPA sexual in women ≥ 1 11 84.62 10 76.92 0.56 SCOPA sexual in mena 0.78 (±1.50) 1.26 (±1.70) 0.07  SCOPA sexual in men ≥ 1 10 31.25 16 45.71 0.13 Variable . Drug-free NT1 patients . Treated NT1 patients . p-value . n . % . n . % . SCOPA-AUT Total Scorea 12.64 (±6.98) 13.34 (±7.75) 0.40 SCOPA gastrointestinala 2.42 (±2.25) 2.53 (±2.49) 0.53  SCOPA gastrointestinal ≥ 1 51 86.44 45 76.27 0.06 SCOPA urinarya 3.78 (±3.10) 4.19 (±3.38) 0.48  SCOPA urinary ≥ 1 53 89.83 54 91.53 0.65 SCOPA cardiovasculara 0.93 (±1.28) 0.93 (±1.14) 0.89  SCOPA cardiovascular ≥ 1 30 50.85 32 54.24 0.62 SCOPA thermoregulatorya 3.19 (±2.67) 3.20 (±2.48) 0.87  SCOPA thermoregulatory ≥ 1 49 83.05 50 84.75 0.74 SCOPA pupillomotora 1.05 (±1.07) 1.05 (±1.01) 0.99  SCOPA pupillomotor ≥ 1 35 59.32 38 64.41 0.43 SCOPA sexual in womena 2.23 (±1.54) 1.54 (±1.20) 0.05  SCOPA sexual in women ≥ 1 11 84.62 10 76.92 0.56 SCOPA sexual in mena 0.78 (±1.50) 1.26 (±1.70) 0.07  SCOPA sexual in men ≥ 1 10 31.25 16 45.71 0.13 aContinuous variables are expressed as means (±SD). Open in new tab Table 5. Comparison between 59 patients with NT1 evaluated twice, in drug-free and treated condition Variable . Drug-free NT1 patients . Treated NT1 patients . p-value . n . % . n . % . SCOPA-AUT Total Scorea 12.64 (±6.98) 13.34 (±7.75) 0.40 SCOPA gastrointestinala 2.42 (±2.25) 2.53 (±2.49) 0.53  SCOPA gastrointestinal ≥ 1 51 86.44 45 76.27 0.06 SCOPA urinarya 3.78 (±3.10) 4.19 (±3.38) 0.48  SCOPA urinary ≥ 1 53 89.83 54 91.53 0.65 SCOPA cardiovasculara 0.93 (±1.28) 0.93 (±1.14) 0.89  SCOPA cardiovascular ≥ 1 30 50.85 32 54.24 0.62 SCOPA thermoregulatorya 3.19 (±2.67) 3.20 (±2.48) 0.87  SCOPA thermoregulatory ≥ 1 49 83.05 50 84.75 0.74 SCOPA pupillomotora 1.05 (±1.07) 1.05 (±1.01) 0.99  SCOPA pupillomotor ≥ 1 35 59.32 38 64.41 0.43 SCOPA sexual in womena 2.23 (±1.54) 1.54 (±1.20) 0.05  SCOPA sexual in women ≥ 1 11 84.62 10 76.92 0.56 SCOPA sexual in mena 0.78 (±1.50) 1.26 (±1.70) 0.07  SCOPA sexual in men ≥ 1 10 31.25 16 45.71 0.13 Variable . Drug-free NT1 patients . Treated NT1 patients . p-value . n . % . n . % . SCOPA-AUT Total Scorea 12.64 (±6.98) 13.34 (±7.75) 0.40 SCOPA gastrointestinala 2.42 (±2.25) 2.53 (±2.49) 0.53  SCOPA gastrointestinal ≥ 1 51 86.44 45 76.27 0.06 SCOPA urinarya 3.78 (±3.10) 4.19 (±3.38) 0.48  SCOPA urinary ≥ 1 53 89.83 54 91.53 0.65 SCOPA cardiovasculara 0.93 (±1.28) 0.93 (±1.14) 0.89  SCOPA cardiovascular ≥ 1 30 50.85 32 54.24 0.62 SCOPA thermoregulatorya 3.19 (±2.67) 3.20 (±2.48) 0.87  SCOPA thermoregulatory ≥ 1 49 83.05 50 84.75 0.74 SCOPA pupillomotora 1.05 (±1.07) 1.05 (±1.01) 0.99  SCOPA pupillomotor ≥ 1 35 59.32 38 64.41 0.43 SCOPA sexual in womena 2.23 (±1.54) 1.54 (±1.20) 0.05  SCOPA sexual in women ≥ 1 11 84.62 10 76.92 0.56 SCOPA sexual in mena 0.78 (±1.50) 1.26 (±1.70) 0.07  SCOPA sexual in men ≥ 1 10 31.25 16 45.71 0.13 aContinuous variables are expressed as means (±SD). Open in new tab Discussion Using a reliable and valid instrument, the SCOPA-AUT questionnaire, we found a high frequency of autonomic symptoms in various domains in a large population of well-characterized patients with NT1 compared to a control population. Higher scores on SCOPA-AUT were associated with older age, poorer quality of life and more depressive symptoms in patients with NT1, without major effect of the drugs intake for narcolepsy. The SCOPA-AUT questionnaire was first validated in Parkinson’s disease [15] and includes items considered relevant by patients and specialists to assess the dysautonomic-related complaints in multiple domains. The questionnaire was further used in other synucleinopathies [16], and sleep disorders such as RLS [18], iRBD [17], and OSAS [19]. As no pathological cut-off of the scale was validated in the literature, we chose the last tertile of the SCOPA-AUT total score of the present sample to define highest scores. Interestingly, the mean total score of patients with NT1 was lower than scores reported in the literature for other diseases (12.92 ± 6.84 for NT1 vs. 18.8 ± 8.5 for Parkinson’s disease, 19.9 ± 10.5 for RLS, 14.5 ± 12.21 for OSAS, and 12.44 ± 6.99 for iRBD), a rather expected difference that may be related to the young age of patients with NT1 compared to other conditions. We highlighted a frequent and large spectrum of clinical autonomic dysfunctions in NT1, with impairment in every subdomains of the SCOPA-AUT questionnaire, namely on gastrointestinal, urinary, cardiovascular, thermoregulatory, pupillomotor, and sexual areas. Some clinically relevant dysautonomic symptoms were already described in NT1 and could be explained, at least partly, by ORX deficiency. Orexin neurons project widely to brain, brainstem areas, and intermediolateral nucleus neurons of the spinal cord, involved in autonomic and cardiovascular regulation [6, 7]. However, frequent autonomic symptoms were also reported in non-ORX deficient sleep disorders that suggests a close relationship between sleep and autonomic activity [26]. Indeed, the pupillomotor symptoms (increased sensitivity to bright light) reported by 64% of patients, could be explained by the lack of stimulation of ORX-containing nerve fibers on preganglionic neurons of the ciliary ganglion, but also to daytime sleepiness. Gastro-intestinal disturbances are particularly frequent in our population of NT1 (88% of cases) that may relate to vagal system dysfunction, with a role of ORX on its regulation [27, 28]. Symptoms related to abnormal temperature regulation, with hyperhidrosis during the day and heat intolerance were found in 87% of NT1 patients. This finding is not surprising since the known relationship between ORX function, temperature and sleep architecture and continuity, and the altered diurnal profile of skin temperature already reported in NT1 [29, 30]. Most of patients with NT1 (92.4%) reported urinary symptoms (especially nocturia and feeling of incomplete emptying of bladder), being potentially related to high fragmented sleep at night or comorbid sleep apnea syndrome [31]. More than half of patients with NT1 (52.2%) also reported abnormalities in the cardiovascular domain, in line with preclinical and clinical studies linking autonomic cardiovascular dysfunction to ORX deficiency [8, 32, 33]. Our results thus emphasized the impaired autonomic control of the cardiovascular system already reported during sleep and wakefulness in NT1 [34, 35]. Therefore, the SCOPA-AUT questionnaire may help to detect cardiovascular autonomic dysfunction at baseline, with the greatest symptoms burden representing orthostatic intolerance that favor an impairment of the sympathetic tone regulation. We suggest to reinforce the follow-up of this at-risk population of NT1 with autonomic cardiovascular complaints to further prevent abnormal blood pressure regulation and related cardio-vascular morbidity and mortality [11, 13, 36]. The last domain explored with the SCOPA-AUT questionnaire concerns the sexual dysfunction that affects 48% of men and 81% of women with NT1. The commonest symptoms were erection problem for men, and vaginal lubrification for women. Interestingly, for all subdomains in the SCOPA-AUT, we found no significant effect of medication intake on autonomic symptoms in the subgroup of 59 patients evaluated twice. The absence of changes of dysautonomia after several months of stable medication for narcolepsy was unexpected. We may however speculate on several hypotheses. Autonomic impairment could be more a trait than a state in NT1. Accordingly, we recently found an absence of cardiac sympathetic denervation in narcolepsy, using the MIBG cardiac scintigraphy, with an absence of effect of treatment intake for narcolepsy at time of evaluation [37]. Increased of diastolic blood pressure and heart rate found in NT1 patients treated by psychostimulants than in untreated [13] were often imperceptible by patients in daily life. In contrast, the impact of anticataplectic drugs (e.g. SNRI and SSRI) on sexual dysfunction was often reported by patients. Indeed, when we analysed separately the items of sexual dysfunction for men, the proportion of patients with ejaculation problem doubled under medication, a result in line with clinical experience. However, there is no significant change in the total SCOPA-AUT score as a function of medication, a reassuring finding for patients. Our results are consistent with a small case-control study, evaluating autonomic symptoms with the SCOPA-AUT questionnaire in 15 narcoleptic patients [38]. However in this study, the authors did not differentiate narcolepsy with and without cataplexy, and patients had no ORX measurement in the CSF. The control group differed for BMI, which could be a confounding factor, and most narcoleptic patients (n = 13) were evaluated with their treatment for narcolepsy (modafinil, methylphenidate, and sodium oxybate). Here, we assessed the presence of clinical autonomic symptoms in a large sample of NT1 patients from a National Reference Center for narcolepsy, we compared the results to healthy controls, and took into account a large number of potential confounders, as obesity, a frequent comorbidity in narcolepsy [39], also known to be associated with an increased sympathetic activity [40]. It is important to emphasize that all subdomains of the SCOPA-AUT questionnaire are not equally relevant and clinically significant for patients with NT1. Some items and subdomains could be relevant in their daily life and management. Others are less crucial in clinical practice, but important for research, to screen and better understand the autonomic impairment in a large population of well-characterized patients with NT1. The dysfunction of temperature regulation previously described in narcolepsy is not necessarily perceived by the patients, and should not be systematically managed. So is the increased sensitivity to bright light. Conversely, sexual dysfunction may be of major importance, as often not reported spontaneously by drug-free patients, but being detected by this self-reported questionnaire. The presence of sexual impairment could guide the physician, to propose an anticataplectic medication rather than another, with less sexual adverse effect. The cardiovascular dysfunctions must be interpreted in the clinical context, for each patient (i.e. presence of cardiovascular comorbidities, antihypertensive medication), so must be urinary symptoms (as nocturia, a symptom of sleep fragmentation and sleep apnea). Gastro-intestinal disturbances are particularly frequent in our sample, somehow nonspecific, but their systematic screening is interesting, related to quality of life and management of patients. In patients with NT1, a higher SCOPA-AUT score (i.e. the highest tertile of the sample) was associated with older age and longer narcolepsy duration, but not with ORX measurements, EDS, BMI, and narcolepsy severity. A trend was found between high SCOPA-AUT score and clinically defined RBD in NT1, in agreement with recent studies showing sympathetic adrenergic abnormalities and higher autonomic symptoms in iRBD patients using the similar SCOPA-AUT questionnaire. However, we found no association between clinically defined RBD or REM sleep without atonia and mean scores on either total or cardiovascular SCOPA-AUT, with key clinical and pathophysiological differences between iRBD and RBD associated with NT1 [37, 41]. Patients with NT1 with higher SCOPA-AUT score had a worse quality of life and more depressive symptoms. Whether these symptoms are causes or consequences of the autonomic disturbances in NT1 remains unclear. Several studies already underlined the high frequency of depressive symptoms and its major impact on health-related quality of life, especially in patients with narcolepsy with cataplexy [42, 43]. The present study has some limitations. We must be careful when interpreting our results, as even this scale has demonstrated its clinical utility in assessing autonomic symptoms in other sleep and neurological disorders, it was not designed specifically for narcoleptic patients. The SCOPA-AUT questionnaire was initially created and validated for Parkinson’s disease, a neurodegenerative disorder that affects patients older than patients with NT1. Some subdomains, particularly the urinary domain items, may be less appropriate for this young population. As perspective for future research, alternative scales could be used, such as the composite autonomic symptom score 31 (COMPASS-31) [44], which addresses different autonomic changes on cardiovagal, adrenergic and sudomotor domains. At last, some subdomains of the SCOPA-AUT turned out to be not sensitive enough for other neurodegenerative disorders (i.e. multiple system atrophy) [45]. The relationships between some items of the SCOPA-AUT questionnaire with pure autonomic symptoms may be questionable in patients with NT1; however this instrument was validated and covers the whole spectrum of problems related to major autonomic areas. We did not assess for direct measures of the autonomic impairment to evaluate and correlate clinical autonomic manifestations with objective measurements of dysautonomia. We found that autonomic symptom load was unassociated with ORX levels, the precise causes for autonomic dysfunction in NT1 remain unclear that require further investigations. Finally, we found no significant effect of medication intake on global autonomic symptoms in the subgroup of NT1 patients evaluated twice, but reported some changes on sexual dysfunction in men. This absence of key global differences may relate to the low sample size (n = 59), the short time window (1.28 ± 1.14 years), or to some changes being potentially imperceptible by patients in daily life or not captured by this questionnaire. To conclude, using the validated SCOPA-AUT scale, we captured the whole spectrum of clinical autonomic dysfunction in NT1, with impairments in all domains explored. This scale appears to be a simple and useful instrument for the assessment of clinical autonomic dysfunction in NT1. The absence of effect of medications for narcolepsy favours of a trait—more than a state—dysfunction in that disease. This comprehensive assessment of dysautonomia has a clinical interest for patients with NT1, as associated with altered quality of life, allowing physicians to screen and treat various symptoms that are often not reported spontaneously by patients. Funding This is not an industry-supported study. Acknowledgments We thank all collaborators within the National Reference Center for Narcolepsy, in Montpellier, France, especially Séverine Béziat for her contribution to the data management, Sabine Scholtz for her contribution to the systematic neurophysiological assessment of narcoleptic patients, and Carole Pesenti for her contribution to the systematic clinical assessment of patients. A special thanks to Ali Ahmad, student of the University of Sciences of Montpellier who actively participated in this study during his internship. At last we are indebted to all the participants of the study: the narcoleptic patients and the controls. This study would not have been possible without their contributions. Conflict of interest statement. Y.D. received funds for seminars, board engagements, and travel to conferences by UCB Pharma, Jazz, Theranexus, Idorsia, Flamel, and Bioprojet. R.L. received funds for speaking by UCB Pharma and Shire. L.B., S.C., E.E., and I.J. report no disclosures. References 1. Peyron C , et al. A mutation in a case of early onset narcolepsy and a generalized absence of hypocretin peptides in human narcoleptic brains . Nat Med. 2000 ; 6 ( 9 ): 991 – 997 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Dauvilliers Y , et al. Narcolepsy with cataplexy . Lancet. 2007 ; 369 ( 9560 ): 499 – 511 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Dauvilliers Y , et al. Rapid eye movement sleep behavior disorder and rapid eye movement sleep without atonia in narcolepsy . Sleep Med. 2013 ; 14 ( 8 ): 775 – 781 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Peyron C , et al. Neurons containing hypocretin (orexin) project to multiple neuronal systems . J Neurosci. 1998 ; 18 ( 23 ): 9996 – 10015 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Date Y , et al. Orexins, orexigenic hypothalamic peptides, interact with autonomic, neuroendocrine and neuroregulatory systems . Proc Natl Acad Sci U S A. 1999 ; 96 ( 2 ): 748 – 753 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Dun NJ , et al. Orexins: a role in medullary sympathetic outflow . Regul Pept. 2000 ; 96 ( 1-2 ): 65 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Grimaldi D , et al. Orexin/hypocretin system and autonomic control: new insights and clinical correlations . Neurology . 2014 ; 82 : 271 – 278 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Berteotti C , et al. The link between narcolepsy and autonomic cardiovascular dysfunction: a translational perspective . Clin Auton Res . 2018 ; 28 ( 6 ): 545 – 555 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Plazzi G , et al. Autonomic disturbances in narcolepsy . Sleep Med Rev. 2011 ; 15 ( 3 ): 187 – 196 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Donadio V , et al. Lower wake resting sympathetic and cardiovascular activities in narcolepsy with cataplexy . Neurology. 2014 ; 83 ( 12 ): 1080 – 1086 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Dauvilliers Y , et al. Non-dipping blood pressure profile in narcolepsy with cataplexy . PLoS One. 2012 ; 7 ( 6 ): e38977 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Tomic S , et al. Impact of autonomic dysfunctions on the quality of life in Parkinson’s disease patients . Acta Neurol Belg. 2017 ; 117 ( 1 ): 207 – 211 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Bosco A , et al. Effect of psychostimulants on blood pressure profile and endothelial function in narcolepsy . Neurology . 2018 ; 90 ( 6 ): e479 – e491 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Ferguson JM . The effects of antidepressants on sexual functioning in depressed patients: a review . J Clin Psychiatry . 2001 ; 62 ( Suppl 3 ): 22 – 34 . Google Scholar Crossref Search ADS PubMed WorldCat 15. Visser M , et al. Assessment of autonomic dysfunction in Parkinson’s disease: the SCOPA-AUT . Mov Disord. 2004 ; 19 ( 11 ): 1306 – 1312 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Berganzo K , et al. SCOPA-AUT scale in different parkinsonisms and its correlation with (123) I-MIBG cardiac scintigraphy . Parkinsonism Relat Disord. 2012 ; 18 ( 1 ): 45 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Ferini-Strambi L , et al. Autonomic symptoms in idiopathic REM behavior disorder: a multicentre case-control study . J Neurol. 2014 ; 261 ( 6 ): 1112 – 1118 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Acar BA , et al. Patients with primary restless legs syndrome have higher prevalence of autonomic dysfunction and irritable bowel syndrome . Singapore Med J. 2018 ; 59 ( 10 ): 539 – 544 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Izzi F , et al. Does continuous positive airway pressure treatment affect autonomic nervous system in patients with severe obstructive sleep apnea? Sleep Med. 2018 ; 42 : 68 – 72 . Google Scholar Crossref Search ADS PubMed WorldCat 20. American Academy of Sleep Medicine . International Classification of Sleep Disorders , 3rd ed. (ICSD-3). Darien, IL: American Academy of Sleep Medicine ; 2014 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 21. Johns MW . A new method for measuring daytime sleepiness: the Epworth sleepiness scale . Sleep. 1991 ; 14 ( 6 ): 540 – 545 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Bastien CH , et al. Validation of the Insomnia Severity Index as an outcome measure for insomnia research . Sleep Med. 2001 ; 2 ( 4 ): 297 – 307 . Google Scholar Crossref Search ADS PubMed WorldCat 23. EuroQol Group . EuroQol--a new facility for the measurement of health-related quality of life . Health Policy Amst Neth. 1990 ; 16 : 199 – 208 . Crossref Search ADS WorldCat 24. Dauvilliers Y , et al. Measurement of narcolepsy symptoms: the narcolepsy severity scale . Neurology. 2017 ; 88 ( 14 ): 1358 – 1365 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Montplaisir J , et al. Polysomnographic diagnosis of idiopathic REM sleep behavior disorder . Mov Disord. 2010 ; 25 ( 13 ): 2044 – 2051 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Calandra-Buonaura G , et al. Cardiovascular autonomic dysfunctions and sleep disorders . Sleep Med Rev. 2016 ; 26 : 43 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Namkoong C , et al. NERP-2 regulates gastric acid secretion and gastric emptying via the orexin pathway . Biochem Biophys Res Commun. 2017 ; 485 ( 2 ): 409 – 413 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Okumura T , et al. Role of brain orexin in the pathophysiology of functional gastrointestinal disorders . J Gastroenterol Hepatol. 2011 ; 26 ( Suppl 3 ): 61 – 66 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Fronczek R , et al. Altered skin-temperature regulation in narcolepsy relates to sleep propensity . Sleep. 2006 ; 29 ( 11 ): 1444 – 1449 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Dimicco JA , et al. The dorsomedial hypothalamus: a new player in thermoregulation . Am J Physiol Regul Integr Comp Physiol. 2007 ; 292 ( 1 ): R47 – R63 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Roth T , et al. Disrupted nighttime sleep in narcolepsy . J Clin Sleep Med . 2013 ; 9 : 955 – 965 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 32. Antunes VR , et al. Orexins/hypocretins excite rat sympathetic preganglionic neurons in vivo and in vitro . Am J Physiol Regul Integr Comp Physiol. 2001 ; 281 ( 6 ): R1801 – R1807 . Google Scholar Crossref Search ADS PubMed WorldCat 33. Sorensen GL , et al. Attenuated heart rate response is associated with hypocretin deficiency in patients with narcolepsy . Sleep. 2013 ; 36 ( 1 ): 91 – 98 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Grimaldi D , et al. Abnormal sleep-cardiovascular system interaction in narcolepsy with cataplexy: effects of hypocretin deficiency in humans . Sleep. 2012 ; 35 ( 4 ): 519 – 528 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Grimaldi D , et al. Spectral analysis of heart rate variability reveals an enhanced sympathetic activity in narcolepsy with cataplexy . Clin Neurophysiol. 2010 ; 121 ( 7 ): 1142 – 1147 . Google Scholar Crossref Search ADS PubMed WorldCat 36. Jennum P , et al. Comorbidity and mortality of narcolepsy: a controlled retro- and prospective national study . Sleep. 2013 ; 36 ( 6 ): 835 – 840 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Barateau L , et al. Exploration of cardiac sympathetic adrenergic nerve activity in narcolepsy . Clin Neurophysiol. 2019 ; 130 ( 3 ): 412 – 418 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Klein G , et al. Dysautonomia in narcolepsy: evidence by questionnaire assessment . J Clin Neurol. 2014 ; 10 ( 4 ): 314 – 319 . Google Scholar Crossref Search ADS PubMed WorldCat 39. Schuld A , et al. Increased body-mass index in patients with narcolepsy . Lancet. 2000 ; 355 ( 9211 ): 1274 – 1275 . Google Scholar Crossref Search ADS PubMed WorldCat 40. Grassi G , et al. Sympathetic activation in obese normotensive subjects . Hypertension. 1995 ; 25 ( 4 Pt 1 ): 560 – 563 . Google Scholar Crossref Search ADS PubMed WorldCat 41. Barateau L , et al. Cardiac sympathetic activity differentiates idiopathic and symptomatic rapid eye movement sleep behaviour disorder . Sci Rep. 2018 ; 8 ( 1 ): 7304 . Google Scholar Crossref Search ADS PubMed WorldCat 42. Dauvilliers Y , et al. Psychological health in central hypersomnias: the French Harmony study . J Neurol Neurosurg Psychiatry. 2009 ; 80 ( 6 ): 636 – 641 . Google Scholar Crossref Search ADS PubMed WorldCat 43. Ingravallo F , et al. The burden of narcolepsy with cataplexy: how disease history and clinical features influence socio-economic outcomes . Sleep Med. 2012 ; 13 ( 10 ): 1293 – 1300 . Google Scholar Crossref Search ADS PubMed WorldCat 44. Sletten DM , et al. COMPASS 31: a refined and abbreviated composite autonomic symptom score . Mayo Clin Proc. 2012 ; 87 ( 12 ): 1196 – 1201 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Damon-Perrière N , et al. Assessment of the Scopa-Aut questionnaire in multiple system atrophy: relation to UMSARS scores and progression over time . Parkinsonism Relat Disord. 2012 ; 18 ( 5 ): 612 – 615 . Google Scholar Crossref Search ADS PubMed WorldCat © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail [email protected]. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Acosta, Francisco M; Sanchez-Delgado, Guillermo; Martinez-Tellez, Borja; Migueles, Jairo H; Amaro-Gahete, Francisco J; Rensen, Patrick C N; Llamas-Elvira, Jose M; Blondin, Denis P; Ruiz, Jonatan R

doi: 10.1093/sleep/zsz177pmid: 31555815

Abstract Study Objectives Short sleep duration and sleep disturbances have been related to obesity and metabolic disruption. However, the behavioral and physiological mechanisms linking sleep and alterations in energy balance and metabolism are incompletely understood. In rodents, sleep regulation is closely related to appropriate brown adipose tissue (BAT) thermogenic activity, but whether the same is true in humans has remained unknown. The present work examines whether sleep duration and quality are related to BAT volume and activity (measured by 18F-FDG) and BAT radiodensity in humans. Methods A total of 118 healthy adults (69% women, 21.9 ± 2.2 years, body mass index: 24.9 ± 4.7 kg/m2) participated in this cross-sectional study. Sleep duration and other sleep variables were measured using a wrist-worn accelerometer for seven consecutive days for 24 hours per day. The Pittsburgh Sleep Quality Index was used to assess sleep quality. All participants then underwent a personalized cold exposure to determine their BAT volume, activity, and radiodensity (a proxy of the intracellular triglyceride content), using static positron emission tomography combined with computed tomography (PET/CI) scan. Results Neither sleep duration nor quality was associated with BAT volume or activity (the latter represented by the mean and peak standardized 18F-FDG uptake values) or radiodensity (all p > .1). The lack of association remained after adjusting the analyses for sex, date of PET/CT, and body composition. Conclusions Although experiments in rodent models indicate a strong relationship to exist between sleep regulation and BAT function, it seems that sleep duration and quality may not be directly related to the BAT variables examined in the present work. Clinical Trial Registration NCT02365129 (ClinicalTrials.gov). brown fat, cold-induced thermogenesis, energy balance, glucose uptake, sleep curtailment, thermoregulation Statement of Significance The behavioral and physiological mechanisms linking sleep and alterations in energy balance and metabolism are not well understood. This study uncovers for the first time that neither sleep duration nor quality is related to brown adipose tissue (BAT) volume, activity, or radiodensity after cold exposure, in a large cohort of young healthy adults. This suggests that the relationship between short sleep duration and poor sleep quality with the obesity pandemic and the increase in cardiometabolic disease is more likely to be influenced by other mechanisms rather than BAT function. Future studies should examine whether BAT function assessed by radio-imaging techniques (or continuously measured through indirect markers) is specifically related to sleep stages using polysomnography records, and whether it is altered after sleep deprivation. Introduction Sleep is an active, regulated metabolic state essential for health [1, 2]. An extensive body of epidemiological and experimental evidence has shown that sleep curtailment and disturbances are related to an increased risk of obesity and the disruption of metabolic and endocrine functions [2–4], becoming a new avenue for intervention. However, the behavioral and physiological mechanisms linking sleep and alterations in energy balance and metabolism are not well understood. Brown adipose tissue (BAT) is a specialized thermogenic organ that dissipates heat, especially during cold exposure, a process mediated by uncoupling protein 1 (UCP1) [5]. In rodents, BAT is characterized by its strong thermogenic capacity, but also by its contribution to metabolic homeostasis via the uptake of nutrients [5, 6] and its role as an endocrine organ [7]. Until a decade ago, it was thought that it is present only in small rodents and neonates, but a number of studies simultaneously confirmed its existence and metabolic activity in adult humans [8–11]. Since its “rediscovery,” manipulating human BAT activity has been contemplated as means of combating obesity and diabetes, although recent evidence calls into question whether its impact on energy balance is as substantial as initially thought [6, 12, 13]. The systems that regulate energy balance and metabolic homeostasis are often linked to the neural circuits that modulate sleep duration and quality [14]. For instance, the dorsomedial nucleus in the hypothalamus, which projects into different nuclei and areas related to energy expenditure, feeding, and sleep regulation [15–18], plays a key role in BAT sympathetic premotor neuron excitement. It is also well known that the sleep and thermoregulatory mechanisms are closely related [19–22]. In adult humans, an increase in the distal skin temperature during the night (which is phased-opposed to the decrease in core temperature) [23] is associated with shortened sleep latency and increases in sleep duration and depth [24, 25]. Therefore, it is biologically plausible that sleep regulation and BAT thermogenic function are related in humans. The first observations of this potential relationship arose from experiments in sleep-deprived rodents; these animals showed a hyperphagic response and an increase in energy expenditure despite a falling body temperature [26, 27]. This led to the hypothesis that BAT is activated to compensate for the heat loss typically observed in sleep deprivation states [27]. Accordingly, Balzano et al. [26] confirmed the 5′-deiodinase activity of BAT to be prompted when rats were sleep deprived. Later experiments [28, 29] in rodents showed that intact BAT thermogenesis is required for restorative sleep responses after induced sleep loss and that BAT has an important function as a sleep-inducing signaling organ. In fact, sleep deprivation induces a sixfold increase in UCP-1 mRNA expression in the BAT of wild-type mice [28]. Interestingly, the activation of BAT is related to rodent rapid and nonrapid eye movement (REM and NREM, respectively) sleep phases under normal and inflammatory conditions [28, 30–32]. There is, therefore, evidence that supports the idea of crosstalk between sleep regulation and BAT function—at least in these animals. Whether these observations also apply to healthy humans remains to be seen. The present study examines whether sleep duration and quality are related to BAT volume and activity (both determined via 18F-FDG uptake) and radiodensity (a proxy of the intracellular triglyceride content) [33] after personalized cold exposure in a cohort of young healthy adults. Unfortunately, nearly all the cross-sectional and longitudinal studies [1–4, 14, 34–36] that have examined the evidence for a relationship between sleep curtailment/other sleep variables and obesity have suffered from (1) the lack of an objective assessment of these variables, (2) not simultaneously assessing sleep duration and quality, and (3) only including body mass index (BMI) among the measured body composition variables. A complementary aim of this work was therefore to determine whether sleep duration and quality are associated with obesity and body composition. Methods Research design and participants A total of 137 young healthy adults took part in this cross-sectional study; all were recruited into the ACTIBATE study [37] (ClinicalTrials.gov, ID: NCT02365129) via advertisements in electronic media and leaflets. Supplementary Figure S1 shows a flow-chart explaining how they were enrolled in the present work. The inclusion criteria were as follows: age 18–25 years old, having a sedentary lifestyle (ie, undertaking <20-minute moderate–vigorous physical activity < 3 d/wk at baseline), to not be a smoker or take any medication, having had a stable body weight over the last 3 months (changes <3 kg), to have no cardiometabolic disease (eg, hypertension or diabetes), and to have no first-degree relative history of cancer. Positron emission tomography combined with computed tomography (PET/CT) assessments were completed over eight dates distributed over October, November, and December of 2015 and 2016 (four per year); all assessments were made in Granada (south of Spain). The study was approved by the University of Granada Ethics Committee on Human Research (no. 924) and by that of the Servicio Andaluz de Salud. All work was performed in accordance with the Declaration of Helsinki (2013 revision); all subjects gave their written informed consent to be included. Procedures Sleep duration and quality Sleep duration and other sleep variables were objectively measured by triaxial accelerometry. Subjects wore an ActiGraph GT3X+ accelerometer (Actisleep, Pensacola, FL) on the non-dominant wrist for seven consecutive days, 24 hours per day (thus including sleeping and waking hours) [37]. Subjects were allowed to remove it only during bathing or swimming, etc. Raw accelerations were recorded using an epoch length of 5 s at a frequency of 100 Hz [38]. During the measurement period, the subjects were required to make daily notes of their in-bed time (time between going to bed and waking) in a diary. Accelerometer assessments were usually completed within the 7 days before the PET/CT assessment (see below). The raw acceleration data were exported to csv files using ActiLife v.6.13.3 software (ActiGraph, Pensacola, FL) and processed using the GGIR package (v.1.6-0, https://cran.r-project.org/web/packages/GGIR/index.html) [39] in R (v.3.1.2, https://www.cran.r-project.org/). Previously published methods were used to minimize the sensor calibration error (autocalibration of the data based on local gravity) [40], and accelerations were determined by calculating the Euclidean Norm Minus One (ENMO) value as x2+y2+z2−1G (where 1G ~ 9.8 m/s2) with negative values rounded to zero. The following were then detected and imputed: (1) all nonwear periods, based on the raw acceleration of the three axes, and (2) all sustained, abnormally high accelerations—which are related to the malfunctioning of the accelerometers [39] (see ref. [41] for further information). A previously proposed algorithm (validated via polysomnography) was used to combine data from the accelerometers and the subjects’ diary reports to detect periods of sleep [42, 43]. According to this algorithm, sleep is defined as any period of sustained inactivity in which there is only minimal arm angle change (i.e., <5°) for 5 minutes during a period recorded as sleep in a subject’s diary [42]. Values for the following sleep-related variables were then determined: (1) night onset (time at which the subject fell asleep); (2) wake-up time; (3) in-bed time (time between going to bed and waking up); (4) sleep duration (time between falling asleep and waking up); (5) sleep efficiency (ratio of sleep duration to in-bed time); (6) number and duration of periods spent awake after sleep onset (WASO). Daytime naps were not taken into account. Only data from participants who wore the accelerometers for ≥16 hours per day over at least 4 days (including at least one weekend day) were included in analyses [38]. Sleep quality was determined using the Pittsburgh Sleep Quality Index (PSQI)—a self-rated (via questionnaire), validated, and reliable measurement of this variable that differentiates good from poor sleepers [44]. Subjects responded to 20 items covering seven domains that measure sleep disturbance over the previous month: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction. In the present work, the scoring system was reversed so that higher values indicated better sleep quality (i.e., fewer sleep disturbances). The scores for the seven domains were then summed [44] to obtain an overall PSQI score on an ascending scale from −21 to 0; this eases interpretation and allows comparisons between studies. Good sleepers were deemed to be those with an overall PSQI score of ≥−5, and bad sleepers as those with an overall score of ≤−6 [44]. Sedentary time and physical activity levels The time spent in sedentary behavior and in light or moderate–vigorous physical activity was determined using a procedure similar to the above, applying age-specific cut-offs for the ENMO value as previously described [45, 46]. Personalized cold exposure and 18F-FDG-PET/CT The personalized cold-exposure protocol followed, and the quantification of the BAT volume and activity, were as previously reported [41, 47, 48]. Briefly, subjects sat in a cool room (19.5–20°C) wearing a water-perfused cooling vest (Polar Products Inc., Stow, OH). The water temperature was reduced from 16.6°C to ~1.4°C every 10 minutes until they began shivering (visually detected by evaluators or self-reported). After 48–72 hours had elapsed the subjects went to the Hospital Virgen de las Nieves, where they were again placed in a cool room (19.5–20°C) and wore the same cooling vest but with the water temperature set ~4°C above their earlier shivering threshold test result for 2 hours. After the first hour, the subjects received an injection of 18F-FDG (~185 MBq) and the water temperature was increased by 1°C to avoid visually detectable shivering. One hour later, they were subjected to PET/CT using a Siemens Biograph 16 PET/CT scanner (Siemens, Erlangen, Germany). A low-dose CT scan (120 kV) was first performed for attenuation correction and anatomic localization. Immediately thereafter, one static acquisition of 2 PET bed positions (6 minutes each) was performed from the atlas vertebra to the mid-chest region [48]. All personalized cold exposure treatments and 18F-FDG-PET/CT data acquisitions were performed according to current methodological recommendations [49]. The BAT volume and activity, estimated via the 18F-FDG uptake, were then determined using the Beth Israel plug-in for the FIJI program [48]. This required: (1) outlining regions of interest (ROIs) in the supraclavicular, laterocervical, paravertebral, and mediastinal regions from the atlas vertebra to the fourth thoracic vertebra, using a 3D-axial technique; (2) the determination of the number of pixels in the above ROIs with a radiodensity range of −190 to −10 Hounsfield Units; and (3) the calculation of individualized, standardized threshold 18F-FDG uptake values (SUV) [1.2/(lean body mass/body mass)] [49]. BAT volume was determined as the number of pixels in the above range with an SUV value above the SUV threshold. BAT activity was represented as the mean SUV (SUVmean; the mean quantity of 18FDG in the above same pixels) and peak SUV (SUVpeak; the mean of the three highest 18F-FDG contents in three pixels within a volume of <1 cm3). The mean BAT radiodensity was calculated as the mean HU value for the above mentioned ROIs. The SUVpeak for the descending aorta (reference tissue) at the height of the fourth thoracic vertebra was also determined, using a single ROI from one slice (image). For confirmatory analyses, the BAT SUVmean and SUVpeak with respect to lean body mass (SUVLBM) [50] were calculated. Anthropometry and body composition Subject height and weight were measured using a SECA scale and stadiometer (model 799, Electronic Column Scale, Hamburg, Germany). Lean mass, fat mass, and visceral adipose tissue (VAT) mass were measured using a Discovery Wi dual-energy x-ray absorptiometer (Hologic, Bedford, MA) [51]. The fat mass index was determined as follows: fat mass (kg)/height squared (m2), and the lean mass index (LMI) as follows: lean mass (kg)/height squared (m2). Statistical analysis Descriptive statistics for continuous and categorical variables were used to analyze the subjects’ sociodemographic and clinical characteristics. Pearson correlations were performed to examine the association between the studied sleep variables and BAT volume, activity, and radiodensity. Partial correlations were then performed to examine the previous relationship after adjusting for sex, and for sex and PET/CT date. One-way analyses of variance, as well as one-way analyses of covariance adjusting for sex and PET/CT date, were also performed to examine whether there was any difference in the measured BAT variables based on the number of hours that subjects spent sleeping and on whether they were good or poor sleepers. Pairwise comparisons were performed using Bonferroni post hoc tests when applicable. Pearson and partial correlation tests were also performed to examine whether sleep variables were related to the 18F-FDG uptake in the descending aorta (reference tissue). As complementary analyses, Pearson and partial correlations were also used to examine whether sleep variables were associated with body composition, before and after adjusting for sex. The level of significance was set at p ≤ .05. All statistical analyses were performed using the Statistical Package for the Social Sciences v.24 (SPSS, Inc., Chicago, IL). Results From the initial sample size (participants with complete sleep, 18F-FDG, and body composition data, n = 137), 19 participants were excluded due to problems with data collection or analysis (see Supplementary Figure S1). Hence, a final sample of 118 participants (69% women) was included in the main analyses. Table 1 shows their descriptive characteristics. The participants wore the accelerometers for 6.8 ± 0.5 days, including almost all the night (~99.4% of in-bed time). They slept 6.34 ± 0.73 hours per day, and ~52% were classified as good sleepers (score ≥ −5). Because the interaction of sex with the determined sleep variables did not have any effect on BAT volume, activity or radiodensity or body composition (p > .05), all analyses were performed pooling the data for women and men together. Table 1. Subject characteristics . All (n = 118)a . . Women (n = 81) . . Men (n = 37) . . Age (y) 22 (2) 22 (2) 22 (2) Professional status, n (%)  Student 57 (49) 39 (48) 18 (50)  Unemployed 40 (34) 31 (38) 9 (25)  Other professional activities 20 (17) 11 (14) 9 (25) Body composition  BMI (kg/m2) 24.9 (4.7) 23.7 (3.8) 27.5 (5.4)  LMI (kg/m2) 14.5 (2.4) 13.3 (1.4) 17.2 (2.0)  FMI (kg/m2) 9.0 (3.0) 9.1 (2.7) 8.7 (3.6)  Fat mass (%) 36.2 (7.3) 38.4 (5.9) 31.3 (7.6)  VAT mass (g) 333.8 (177.7) 284.2 (157.6) 442.3 (172.7) Objective sleep measures  Valid days (d) 6.8 (0.5) 6.8 (0.5) 6.7 (0.5)  Nonwear time at night (min/d) 3 (6) 3 (7) 2 (5)  Night onset (hh:mm) 01:16 (01:11) 01:12 (01:11) 01:24 (01:12)  Wake-up time (hh:mm) 08:52 (01:03) 08:47 (00:59) 09:03 (01:10)  In-bed time (min/d) 440 (47) 441 (43) 437 (55)  Sleep duration (min/d) 381 (44) 386 (43) 369 (45)  Sleep efficiency 0.87 (0.05) 0.88 (0.05) 0.85 (0.05)  Time in WASO (min/d) 59 (27) 55 (22) 69 (34)  Blocks in WASO (no/d) 56 (35) 52 (25) 63 (51) Subjective sleep measures (PSQI)  Sleep quality −1.1 (0.7) −1.1 (0.6) −1.3 (0.7)  Sleep latency −1.1 (0.8) −1.1 (0.8) −1.2 (0.8)  Sleep duration −0.8 (0.8) −0.8 (0.8) −0.9 (0.8)  Sleep efficiency −0.5 (0.8) −0.5 (0.8) −0.6 (0.8)  Sleep disturbances −1.1 (0.4) −1.1 (0.4) −1.0 (0.3)  Sleep medication −0.1 (0.5) −0.1 (0.5) −0.2 (0.6)  Daytime dysfunction −0.9 (0.7) −0.9 (0.7) −0.9 (0.7)  Global PSQI score −5.8 (2.6) −5.6 (2.6) −6.1 (2.7) Sedentary behavior and PA  Sedentary time (min/d) 794 (65) 786 (55) 812 (80)  Light PA (min/d) 118 (27) 123 (25) 107 (30)  Moderate–vigorous PA (min/d) 89 (32) 92 (31) 84 (34) PET/CT parameters  SUV threshold 2.06 (0.23) 2.13 (0.21) 1.90 (0.21)  BAT volume (mL) 68.11 (57.89) 63.72 (52.79) 77.70 (67.53)  BAT SUVmean 3.74 (1.97) 3.96 (2.15) 3.26 (1.40)  BAT SUVpeak 11.19 (8.32) 11.71 (8.61) 10.07 (7.66)  BAT radiodensity (HU) −59.03 (11.76) −60.21 (11.55) −56.40 (11.95)  Descending aorta SUVpeak 0.80 (0.20) 0.81 (0.21) 0.77 (0.17) . All (n = 118)a . . Women (n = 81) . . Men (n = 37) . . Age (y) 22 (2) 22 (2) 22 (2) Professional status, n (%)  Student 57 (49) 39 (48) 18 (50)  Unemployed 40 (34) 31 (38) 9 (25)  Other professional activities 20 (17) 11 (14) 9 (25) Body composition  BMI (kg/m2) 24.9 (4.7) 23.7 (3.8) 27.5 (5.4)  LMI (kg/m2) 14.5 (2.4) 13.3 (1.4) 17.2 (2.0)  FMI (kg/m2) 9.0 (3.0) 9.1 (2.7) 8.7 (3.6)  Fat mass (%) 36.2 (7.3) 38.4 (5.9) 31.3 (7.6)  VAT mass (g) 333.8 (177.7) 284.2 (157.6) 442.3 (172.7) Objective sleep measures  Valid days (d) 6.8 (0.5) 6.8 (0.5) 6.7 (0.5)  Nonwear time at night (min/d) 3 (6) 3 (7) 2 (5)  Night onset (hh:mm) 01:16 (01:11) 01:12 (01:11) 01:24 (01:12)  Wake-up time (hh:mm) 08:52 (01:03) 08:47 (00:59) 09:03 (01:10)  In-bed time (min/d) 440 (47) 441 (43) 437 (55)  Sleep duration (min/d) 381 (44) 386 (43) 369 (45)  Sleep efficiency 0.87 (0.05) 0.88 (0.05) 0.85 (0.05)  Time in WASO (min/d) 59 (27) 55 (22) 69 (34)  Blocks in WASO (no/d) 56 (35) 52 (25) 63 (51) Subjective sleep measures (PSQI)  Sleep quality −1.1 (0.7) −1.1 (0.6) −1.3 (0.7)  Sleep latency −1.1 (0.8) −1.1 (0.8) −1.2 (0.8)  Sleep duration −0.8 (0.8) −0.8 (0.8) −0.9 (0.8)  Sleep efficiency −0.5 (0.8) −0.5 (0.8) −0.6 (0.8)  Sleep disturbances −1.1 (0.4) −1.1 (0.4) −1.0 (0.3)  Sleep medication −0.1 (0.5) −0.1 (0.5) −0.2 (0.6)  Daytime dysfunction −0.9 (0.7) −0.9 (0.7) −0.9 (0.7)  Global PSQI score −5.8 (2.6) −5.6 (2.6) −6.1 (2.7) Sedentary behavior and PA  Sedentary time (min/d) 794 (65) 786 (55) 812 (80)  Light PA (min/d) 118 (27) 123 (25) 107 (30)  Moderate–vigorous PA (min/d) 89 (32) 92 (31) 84 (34) PET/CT parameters  SUV threshold 2.06 (0.23) 2.13 (0.21) 1.90 (0.21)  BAT volume (mL) 68.11 (57.89) 63.72 (52.79) 77.70 (67.53)  BAT SUVmean 3.74 (1.97) 3.96 (2.15) 3.26 (1.40)  BAT SUVpeak 11.19 (8.32) 11.71 (8.61) 10.07 (7.66)  BAT radiodensity (HU) −59.03 (11.76) −60.21 (11.55) −56.40 (11.95)  Descending aorta SUVpeak 0.80 (0.20) 0.81 (0.21) 0.77 (0.17) Continuous variables are presented as mean (standard deviation) and categorical variables as number (percentage). BAT = brown adipose tissue, BMI = body mass index, FMI = fat mass index, HU = Hounsfield units, LMI = lean mass index, PA = physical activity, PET/CT = positron emission tomography combined with computed tomography, PSQI = Pittsburgh Sleep Quality Index, SUV = standardized uptake value, VAT = visceral adipose tissue, WASO = awake after sleep onset. aSome data were missing for professional status (remaining cases, n = 117) and BAT radiodensity (remaining cases, n = 116). Open in new tab Table 1. Subject characteristics . All (n = 118)a . . Women (n = 81) . . Men (n = 37) . . Age (y) 22 (2) 22 (2) 22 (2) Professional status, n (%)  Student 57 (49) 39 (48) 18 (50)  Unemployed 40 (34) 31 (38) 9 (25)  Other professional activities 20 (17) 11 (14) 9 (25) Body composition  BMI (kg/m2) 24.9 (4.7) 23.7 (3.8) 27.5 (5.4)  LMI (kg/m2) 14.5 (2.4) 13.3 (1.4) 17.2 (2.0)  FMI (kg/m2) 9.0 (3.0) 9.1 (2.7) 8.7 (3.6)  Fat mass (%) 36.2 (7.3) 38.4 (5.9) 31.3 (7.6)  VAT mass (g) 333.8 (177.7) 284.2 (157.6) 442.3 (172.7) Objective sleep measures  Valid days (d) 6.8 (0.5) 6.8 (0.5) 6.7 (0.5)  Nonwear time at night (min/d) 3 (6) 3 (7) 2 (5)  Night onset (hh:mm) 01:16 (01:11) 01:12 (01:11) 01:24 (01:12)  Wake-up time (hh:mm) 08:52 (01:03) 08:47 (00:59) 09:03 (01:10)  In-bed time (min/d) 440 (47) 441 (43) 437 (55)  Sleep duration (min/d) 381 (44) 386 (43) 369 (45)  Sleep efficiency 0.87 (0.05) 0.88 (0.05) 0.85 (0.05)  Time in WASO (min/d) 59 (27) 55 (22) 69 (34)  Blocks in WASO (no/d) 56 (35) 52 (25) 63 (51) Subjective sleep measures (PSQI)  Sleep quality −1.1 (0.7) −1.1 (0.6) −1.3 (0.7)  Sleep latency −1.1 (0.8) −1.1 (0.8) −1.2 (0.8)  Sleep duration −0.8 (0.8) −0.8 (0.8) −0.9 (0.8)  Sleep efficiency −0.5 (0.8) −0.5 (0.8) −0.6 (0.8)  Sleep disturbances −1.1 (0.4) −1.1 (0.4) −1.0 (0.3)  Sleep medication −0.1 (0.5) −0.1 (0.5) −0.2 (0.6)  Daytime dysfunction −0.9 (0.7) −0.9 (0.7) −0.9 (0.7)  Global PSQI score −5.8 (2.6) −5.6 (2.6) −6.1 (2.7) Sedentary behavior and PA  Sedentary time (min/d) 794 (65) 786 (55) 812 (80)  Light PA (min/d) 118 (27) 123 (25) 107 (30)  Moderate–vigorous PA (min/d) 89 (32) 92 (31) 84 (34) PET/CT parameters  SUV threshold 2.06 (0.23) 2.13 (0.21) 1.90 (0.21)  BAT volume (mL) 68.11 (57.89) 63.72 (52.79) 77.70 (67.53)  BAT SUVmean 3.74 (1.97) 3.96 (2.15) 3.26 (1.40)  BAT SUVpeak 11.19 (8.32) 11.71 (8.61) 10.07 (7.66)  BAT radiodensity (HU) −59.03 (11.76) −60.21 (11.55) −56.40 (11.95)  Descending aorta SUVpeak 0.80 (0.20) 0.81 (0.21) 0.77 (0.17) . All (n = 118)a . . Women (n = 81) . . Men (n = 37) . . Age (y) 22 (2) 22 (2) 22 (2) Professional status, n (%)  Student 57 (49) 39 (48) 18 (50)  Unemployed 40 (34) 31 (38) 9 (25)  Other professional activities 20 (17) 11 (14) 9 (25) Body composition  BMI (kg/m2) 24.9 (4.7) 23.7 (3.8) 27.5 (5.4)  LMI (kg/m2) 14.5 (2.4) 13.3 (1.4) 17.2 (2.0)  FMI (kg/m2) 9.0 (3.0) 9.1 (2.7) 8.7 (3.6)  Fat mass (%) 36.2 (7.3) 38.4 (5.9) 31.3 (7.6)  VAT mass (g) 333.8 (177.7) 284.2 (157.6) 442.3 (172.7) Objective sleep measures  Valid days (d) 6.8 (0.5) 6.8 (0.5) 6.7 (0.5)  Nonwear time at night (min/d) 3 (6) 3 (7) 2 (5)  Night onset (hh:mm) 01:16 (01:11) 01:12 (01:11) 01:24 (01:12)  Wake-up time (hh:mm) 08:52 (01:03) 08:47 (00:59) 09:03 (01:10)  In-bed time (min/d) 440 (47) 441 (43) 437 (55)  Sleep duration (min/d) 381 (44) 386 (43) 369 (45)  Sleep efficiency 0.87 (0.05) 0.88 (0.05) 0.85 (0.05)  Time in WASO (min/d) 59 (27) 55 (22) 69 (34)  Blocks in WASO (no/d) 56 (35) 52 (25) 63 (51) Subjective sleep measures (PSQI)  Sleep quality −1.1 (0.7) −1.1 (0.6) −1.3 (0.7)  Sleep latency −1.1 (0.8) −1.1 (0.8) −1.2 (0.8)  Sleep duration −0.8 (0.8) −0.8 (0.8) −0.9 (0.8)  Sleep efficiency −0.5 (0.8) −0.5 (0.8) −0.6 (0.8)  Sleep disturbances −1.1 (0.4) −1.1 (0.4) −1.0 (0.3)  Sleep medication −0.1 (0.5) −0.1 (0.5) −0.2 (0.6)  Daytime dysfunction −0.9 (0.7) −0.9 (0.7) −0.9 (0.7)  Global PSQI score −5.8 (2.6) −5.6 (2.6) −6.1 (2.7) Sedentary behavior and PA  Sedentary time (min/d) 794 (65) 786 (55) 812 (80)  Light PA (min/d) 118 (27) 123 (25) 107 (30)  Moderate–vigorous PA (min/d) 89 (32) 92 (31) 84 (34) PET/CT parameters  SUV threshold 2.06 (0.23) 2.13 (0.21) 1.90 (0.21)  BAT volume (mL) 68.11 (57.89) 63.72 (52.79) 77.70 (67.53)  BAT SUVmean 3.74 (1.97) 3.96 (2.15) 3.26 (1.40)  BAT SUVpeak 11.19 (8.32) 11.71 (8.61) 10.07 (7.66)  BAT radiodensity (HU) −59.03 (11.76) −60.21 (11.55) −56.40 (11.95)  Descending aorta SUVpeak 0.80 (0.20) 0.81 (0.21) 0.77 (0.17) Continuous variables are presented as mean (standard deviation) and categorical variables as number (percentage). BAT = brown adipose tissue, BMI = body mass index, FMI = fat mass index, HU = Hounsfield units, LMI = lean mass index, PA = physical activity, PET/CT = positron emission tomography combined with computed tomography, PSQI = Pittsburgh Sleep Quality Index, SUV = standardized uptake value, VAT = visceral adipose tissue, WASO = awake after sleep onset. aSome data were missing for professional status (remaining cases, n = 117) and BAT radiodensity (remaining cases, n = 116). Open in new tab Neither sleep duration nor sleep quality was associated with BAT volume, activity, or radiodensity Figure 1 shows that no objective or subjective sleep variable was associated with BAT volume, activity (SUVmean, SUVpeak), or BAT radiodensity (all p > .05). Similarly, partial correlations, after adjusting for sex, and for sex and PET/CT date, revealed no sleep variable to be associated with any measured BAT variable (all p > .05; data not shown). These results remained similar when the analyses were repeated including only those participants with detectable BAT (data not shown). Neither did the results change following additional adjustment for in-bed time, sleep efficiency, nonwear time of the accelerometer during the night, sedentary time, physical activity levels, or any of the body composition variables examined (all p > .05; data not shown). No changes were appreciated when SUV was normalized to lean body mass (SUVLBM), instead of total body mass (SUVBM) for calculating BAT SUVmean and SUVpeak (data not shown). Figure 1. Open in new tabDownload slide Association between sleep variables and brown adipose tissue (BAT) volume and activity (determined via 18F-FDG uptake) (n = 118) and radiodensity (n = 116). Pearson correlations were performed to examine the association between sleep variables and BAT volume (A), mean standardized uptake value (SUVmean) (B), SUVpeak (C), and radiodensity (D). No significant associations were found (p > .05). Higher global Pittsburgh Sleep Quality Index (PSQI) scores are indicative of better sleep quality. Figure 1. Open in new tabDownload slide Association between sleep variables and brown adipose tissue (BAT) volume and activity (determined via 18F-FDG uptake) (n = 118) and radiodensity (n = 116). Pearson correlations were performed to examine the association between sleep variables and BAT volume (A), mean standardized uptake value (SUVmean) (B), SUVpeak (C), and radiodensity (D). No significant associations were found (p > .05). Higher global Pittsburgh Sleep Quality Index (PSQI) scores are indicative of better sleep quality. No differences were found in BAT volume, activity, or radiodensity among subjects who slept 4–5 hours (n = 7), 5–6 hours (n = 23), 6–7 hours (n = 67), 7–8 hours (n = 21) or between good sleepers (n = 61) and poor sleepers (n = 57) (Figure 2) (all p > .05). Neither did any appear following adjustment for sex or for sex and PET/CT date (all p > .05). It is noteworthy that the subjects in these previous categories had similar body composition and cardiometabolic profile and undertook similar levels of physical activity (all p > .05; data not shown). However, the participants who slept for 4–5 hours spent considerably longer in sedentary behavior than those who slept 6–7 hours (879 vs 785 min/d; p < .001) and 7–8 hours (879 vs 746 min/d; p < .001). These results remained after grouping subjects as sleeping for 4–5 or 5–6 hours (data not shown). Figure 2. Open in new tabDownload slide Differences in brown adipose tissue (BAT) volume and activity (determined via 18F-FDG uptake) (n = 118) and radiodensity (n = 116), based on the number of hours spent sleeping and on whether subjects were good or poor sleepers. (A) BAT volume, mean standardized uptake value (SUVmean), SUVpeak, and radiodensity were compared by one-way analysis of variance (ANOVA) based on the average number of hours per night subjects spent sleeping (measured via accelerometry). Subjects were divided into four categories: those who had 4- to 5-h sleep (n = 7), 5- to 6-h sleep (n = 23), 6- to 7-h sleep (n = 67), 7- to 8-h sleep (n = 21). (B) BAT volume, SUVmean, SUVpeak, and radiodensity were compared by ANOVA based on whether subjects were good or bad sleepers. Good sleepers (n = 61) were defined as those who had an overall Pittsburgh Sleep Quality Index (PSQI) score of ≥−5, and bad sleepers (n = 57) as those with a score of ≤−6. Measurements of BAT radiodensity were missing for two subjects (one in the 5- to 6-h sleep time group and one in the 6- to 7-h sleep time group; and one good sleeper and one poor sleeper). HU, Hounsfield units. Figure 2. Open in new tabDownload slide Differences in brown adipose tissue (BAT) volume and activity (determined via 18F-FDG uptake) (n = 118) and radiodensity (n = 116), based on the number of hours spent sleeping and on whether subjects were good or poor sleepers. (A) BAT volume, mean standardized uptake value (SUVmean), SUVpeak, and radiodensity were compared by one-way analysis of variance (ANOVA) based on the average number of hours per night subjects spent sleeping (measured via accelerometry). Subjects were divided into four categories: those who had 4- to 5-h sleep (n = 7), 5- to 6-h sleep (n = 23), 6- to 7-h sleep (n = 67), 7- to 8-h sleep (n = 21). (B) BAT volume, SUVmean, SUVpeak, and radiodensity were compared by ANOVA based on whether subjects were good or bad sleepers. Good sleepers (n = 61) were defined as those who had an overall Pittsburgh Sleep Quality Index (PSQI) score of ≥−5, and bad sleepers (n = 57) as those with a score of ≤−6. Measurements of BAT radiodensity were missing for two subjects (one in the 5- to 6-h sleep time group and one in the 6- to 7-h sleep time group; and one good sleeper and one poor sleeper). HU, Hounsfield units. No association was found between any sleep variable and the descending aorta SUVpeak value, even after adjustment for sex, and for sex and PET/CT date (all p > .05; Figure 3). Figure 3. Open in new tabDownload slide Association between sleep variables and the descending aorta peak standardized uptake value (SUVpeak) (n = 118). Pearson’s correlations were performed. Higher values in the global Pittsburgh Sleep Quality Index (PSQI) score are indicative of better sleep quality. Figure 3. Open in new tabDownload slide Association between sleep variables and the descending aorta peak standardized uptake value (SUVpeak) (n = 118). Pearson’s correlations were performed. Higher values in the global Pittsburgh Sleep Quality Index (PSQI) score are indicative of better sleep quality. Association between sleep duration and quality and body composition In-bed time was inversely associated with BMI and VAT mass (r = −.188, p = .04 and r = −.18, p = .05, respectively), and sleep duration was inversely associated with LMI and VAT mass (r = −.226, p = .014 and r = −.190, p = .04; Table 2). In addition, sleep efficiency and time in WASO were significantly associated with fat mass (r = .229, p = .01 and r = −.269, p = .003). After adjustment for sex, only in-bed time remained significantly associated with BMI and VAT mass, along with time in WASO with body fat mass (r = −.189, p = .041; r = −.185, p = .046 and r = −.186, p = .045; Table 2). Supplementary Table S1 shows the relationships among sleep variables measured objectively (by accelerometry) and subjectively (PSQI); the results agree with those of other studies [52]. Table 2. Association between sleep variables and body composition (n = 118) . Night onset (hh:mm) . Wake-up time (hh:mm) . In-bed time (min/d) . Sleep duration (min/d) . Sleep efficiency . Time in WASO (min/d) . Blocks in WASO (no/d) . Global PSQI score . BMI (kg/m2) 0.068 0.086 −0.188a −0.173 0.003 −0.049 0.070 −0.054 LMI (kg/m2) 0.070 0.095 −0.133 −0.226 −0.156 0.137 0.192 −0.120 FMI (kg/m2) 0.043 0.055 −0.165 −0.067 0.129 −0.180 −0.045 0.029 Fat mass (%) 0.024 0.012 −0.129 0.026 0.229 −0.269a −0.151 0.080 VAT mass (g) 0.093 0.067 −0.183a −0.190 −0.050 −0.011 0.060 -0.117 . Night onset (hh:mm) . Wake-up time (hh:mm) . In-bed time (min/d) . Sleep duration (min/d) . Sleep efficiency . Time in WASO (min/d) . Blocks in WASO (no/d) . Global PSQI score . BMI (kg/m2) 0.068 0.086 −0.188a −0.173 0.003 −0.049 0.070 −0.054 LMI (kg/m2) 0.070 0.095 −0.133 −0.226 −0.156 0.137 0.192 −0.120 FMI (kg/m2) 0.043 0.055 −0.165 −0.067 0.129 −0.180 −0.045 0.029 Fat mass (%) 0.024 0.012 −0.129 0.026 0.229 −0.269a −0.151 0.080 VAT mass (g) 0.093 0.067 −0.183a −0.190 −0.050 −0.011 0.060 -0.117 Pearson’s correlation coefficients are shown. Statistically significant values are shown in bold (p ≤ .05). BMI = body mass index, FMI = fat mass index, LMI = lean mass index, PSQI = Pittsburgh sleep quality index, VAT = visceral adipose tissue, WASO = awake after sleep onset. aAssociations that remained significant (p ≤ .05) after adjusting for sex. Open in new tab Table 2. Association between sleep variables and body composition (n = 118) . Night onset (hh:mm) . Wake-up time (hh:mm) . In-bed time (min/d) . Sleep duration (min/d) . Sleep efficiency . Time in WASO (min/d) . Blocks in WASO (no/d) . Global PSQI score . BMI (kg/m2) 0.068 0.086 −0.188a −0.173 0.003 −0.049 0.070 −0.054 LMI (kg/m2) 0.070 0.095 −0.133 −0.226 −0.156 0.137 0.192 −0.120 FMI (kg/m2) 0.043 0.055 −0.165 −0.067 0.129 −0.180 −0.045 0.029 Fat mass (%) 0.024 0.012 −0.129 0.026 0.229 −0.269a −0.151 0.080 VAT mass (g) 0.093 0.067 −0.183a −0.190 −0.050 −0.011 0.060 -0.117 . Night onset (hh:mm) . Wake-up time (hh:mm) . In-bed time (min/d) . Sleep duration (min/d) . Sleep efficiency . Time in WASO (min/d) . Blocks in WASO (no/d) . Global PSQI score . BMI (kg/m2) 0.068 0.086 −0.188a −0.173 0.003 −0.049 0.070 −0.054 LMI (kg/m2) 0.070 0.095 −0.133 −0.226 −0.156 0.137 0.192 −0.120 FMI (kg/m2) 0.043 0.055 −0.165 −0.067 0.129 −0.180 −0.045 0.029 Fat mass (%) 0.024 0.012 −0.129 0.026 0.229 −0.269a −0.151 0.080 VAT mass (g) 0.093 0.067 −0.183a −0.190 −0.050 −0.011 0.060 -0.117 Pearson’s correlation coefficients are shown. Statistically significant values are shown in bold (p ≤ .05). BMI = body mass index, FMI = fat mass index, LMI = lean mass index, PSQI = Pittsburgh sleep quality index, VAT = visceral adipose tissue, WASO = awake after sleep onset. aAssociations that remained significant (p ≤ .05) after adjusting for sex. Open in new tab Discussion The present results show that sleep duration and quality are not associated with BAT volume or activity (both estimated via 18F-FDG uptake) or BAT radiodensity following cold exposure in young, sedentary adults. These findings persisted after adjusting for sex, PET/CT date, and body composition. Although experiments with rodents indicate sleep homeostatic mechanisms to be closely related to BAT function, evidence for the same in adult humans is scarce. In the single study that exists, Enevoldsen et al. [53] examined whether BAT function was similar in seven patients with narcolepsy type I compared with seven matched healthy controls. Narcolepsy type I is a neurological disorder characterized by the loss of orexinergic neurons; this leads to excessive daytime sleepiness, dysregulated REM sleep, cataplexy, fragmented light sleep, and a higher frequency of sleeping-awake transitions [53, 54]. Thus, it is plausible that patients with narcolepsy, who have largely altered sleep patterns, might also show altered BAT function. However, the latter study found that BAT 18F-FDG uptake and sympathetic outflow upon cold exposure were similar in narcoleptic and control participants, calling into question whether sleep duration and quality have any influence over human BAT recruitment and activation. Given the limited sample size, the latter results cannot be generalized to the healthy population, especially because patients with narcolepsy normally have autonomic dysfunction, including changes in their cardiovascular, sympathetic, and temperature regulation [55]. No previous studies have examined the relationship between sleep and BAT in healthy adults, precluding any comparison with other results. The present results do not concur, however, with observations made in rodent models revealing an intimate relationship between sleep regulation and BAT thermogenic activity. This disagreement might be explained in several ways. First, there are vast differences between species in terms of their morphology and physiology [56]. Rodents have a smaller body volume to surface area ratio and their thermoregulation system is designed to conserve heat, whereas humans have a larger body volume to surface area ratio and thus dissipate more heat. Hence, rodents have a higher reliance on BAT thermogenic activity during cold exposure than humans [57]. Because the systems that regulate energy balance and metabolic homeostasis are often linked to the neural circuit that regulates sleep duration and quality [14–18], it would seem coherent that in rodents, in which BAT thermogenic activity makes important contributions to energy balance and metabolic homeostasis, brown adipocyte function should be intimately related to sleep regulation. The same may not hold in humans, however, because recent evidence indicates that the relative contribution of BAT to energy expenditure is rather low and might be insufficient to affect energy balance [12, 13, 58]. Second, to be translatable, experiments in rodents must be performed under conditions that can accurately reflect the physiology and pathophysiology of the humans (e.g., housing mice within their thermoneutral temperature range) [56]. Third, previous studies in rodent models that examined the relationship between sleep regulation and BAT function were performed under different conditions (e.g., following sleep deprivation, the pharmacological or agonist activation of BAT, or in a scenario of systemic inflammation) to those of the present work. Evidence collected in rodents has shown that their thermoregulatory and sleep mechanisms are closely related [19–21]. Wild-type mice exposed to warm temperatures (35°C) show a robust increase in NREM sleep [28]. In addition, sleep-promoting mechanisms and BAT thermogenesis are both stimulated by sleep loss, being positively correlated [22, 28]. Similarly, in adult humans, an increase in the distal skin temperature during the night (which is phased-opposed to the decrease in core temperature) [23], is associated with shortened sleep latency and increases in sleep duration and depth [24, 25], demonstrating a link between the thermoregulatory and sleep centers. Accordingly, we previously determined in a subcohort of the present subjects (n = 77) that the time at which this increase occurs is weakly related to BAT activity (reflected as SUVmean) (B = −0.2, p = .04; paper submitted) [59]. These findings, together with the present results, suggest that sleep duration and quality might not be directly related to BAT activity following cold exposure. In addition, rodent experiments have shown that BAT may act as a sleep-promoting signaling organ; there are extensive afferent projections running from the BAT into the hypothalamic area, and a population of these is thermosensitive, raising the possibility that BAT thermogenesis may induce sleep independent of changes in core temperature [14, 22, 28, 29]. This hypothesis agrees with the fact that there seems to be a significant time lag between the somnogenic and delayed body temperature effects following pharmacological activation of BAT in rodents [28]. Whether BAT might act as a sleep-prompting signaling organ in humans remains to be seen. Nor can it be ruled out that human BAT might exert its function on sleep regulation via endocrine mechanisms. For instance, BAT secretes adenosine [60], an endogenous factor that promotes sleep by blocking inhibitory inputs to the ventrolateral preoptic area’s sleep-active neurons [61], and can express factors such as interleukin-6, interleukin-1, and tumor necrosis factor, all of which have somnogenic effects [7, 28]. A complementary aim of the present work was to examine whether sleep duration and quality are associated with obesity and body composition. The results show a weak inverse relationship between in-bed time and both BMI and VAT mass (after controlling for sex). Interestingly, in-bed time and sleep duration were moderately and inversely associated, whereas the moment of entering sleep was positively associated with the time spent in sedentary behavior (Supplementary Table S2). This suggests that those subjects who slept less, and who went to sleep later, where those who spent more time in sedentary behavior. This may be explained in that sleep curtailment, or a late chronotype (which is related to misalignment between social rhythms and the circadian clock), may be related to greater drowsiness during the day, and consequently to more sedentary behavior and an increased risk of obesity [62]. Therefore, the in-bed time may be related to increased risk of obesity through indirect mechanisms. Taking everything into account, it is tempting to speculate that the relationship between sleep curtailment and risk of obesity might not be influenced by BAT volume or activity. Other behavioral (e.g., sedentary time) and physiological mechanisms related to homeostatic (e.g., sleep pressure), circadian (e.g., sleeping-awake cycle schedule), and metabolic control (e.g., dysregulated secretion of gastrointestinal peptides, alterations to the appetite regulation centers of the brain) may explain this relationship better. The present results should be interpreted with caution; the study has a cross-sectional design that precludes the establishment of causal relationships. For instance, it might be possible that habitual short sleep and poor sleep quality could alter the function of the BAT metabolism, as it has been previously shown with other metabolic functions (e.g., glucose metabolism) [2]. In contrast, it could be hypothesized that BAT exerts its influence on sleep duration and quality. Anyhow, sleep is a complex phenomenon, which is influenced not only by behavioral, but also by physiological mechanisms related to homeostatic, circadian, and metabolic control under the participant’s natural sleep environment. Therefore, it exits the possibility that these factors could be influencing the relationship between sleep parameters and 18F-FDG uptake and radiodensity. Furthermore, the sample was composed of young adults, most of whom had a healthy cardiometabolic profile (data not shown); this could have masked or weakened the associations between sleep variables and BAT 18F-FDG uptake, BAT radiodensity, or obesity risk. It should also be remembered that the use of the shivering threshold (subjectively assessed) as the end point of the personalized cooling protocol may have introduced variation into the cooling stimulation, which would be reflected in the subjects’ BAT activation [63]. Despite being the most used technique to assess BAT, a single static 18F-FDG-PET/CT scan has several limitations that might not allow for the accurate estimation of cold-induced BAT metabolic activity [64]. Whether the present findings will be replicated when using other radiotracers such as 15O-oxygen [11], C-acetate, or 18F-fluoro-6-thia-heptadecanoic acid to quantify BAT metabolism remains to be seen. It is also necessary to consider that (1) napping time was not included in the analyses, since we do not have information that allows an accurate quantification of it. The timing and duration of napping could have a profound effect on night sleep, and might partially mask the relationship between sleep parameters and BAT 18F-FDG uptake and radiodensity. However, based on the acceleration records and participant reports, it seems that most of our participants did not nap during the day (also probably because many of them were university students and had to attend classes); (2) although accelerometer records (combined with sleep diaries and subjective measures) are a valid and extensively used measure of sleep duration and quality under free-living conditions [43, 65], they are not able to differentiate between REM and NREM sleep, and thus they may provide a limited insight into the real architecture of sleep wake-activity; (3) although we followed the most updated international recommendations [49] to quantify and analyze BAT 18F-FDG uptake, we performed an unique temporal measure after a personalized cold exposure. Therefore, future studies should examine how continuously measured BAT activity is specifically related to REM and NREM sleep using polysomnography records (since these phases are metabolically different [34]), to get a deeper insight into the interaction between BAT function and sleep regulation. This fact will be conditioned by the advance of the current technologies to assess BAT metabolic activity in a noninvasive and nonionizing manner, or by the validation of indirect markers that accurately reflect its activity. Furthermore, experimental studies should manipulate sleep (e.g., sleep deprivation) and/or BAT function (e.g., use of beta-3 adrenergic agonists available) under well-controlled lab conditions to establish a causal relationship. In conclusion, sleep duration and quality appear not to be related to BAT volume or activity (both estimated by 18F-FDG uptake) or BAT radiodensity following cold exposure, in young healthy, sedentary adults. Further studies are needed to fully understand the underlying mechanisms of sleep regulation, and how short sleep duration and poor sleep quality are related to the obesity pandemic and the increase in cardiometabolic disease. Funding This study was supported by the Spanish Ministry of Economy and Competitiveness via the Fondo de Investigación Sanitaria del Instituto de Salud Carlos III (PI13/01393) and PTA-12264, Retos de la Sociedad (DEP2016-79512-R) and European Regional Development Funds (ERDF), the Spanish Ministry of Education (FPU13/04365 and FPU 15/04059), the Fundación Iberoamericana de Nutrición (FINUT), the Redes Temáticas de Investigación Cooperativa RETIC (Red SAMID RD16/0022), the AstraZeneca HealthCare Foundation, the University of Granada Plan Propio de Investigación 2016—Excellence actions: Unit of Excellence on Exercise and Health (UCEES)—and Plan Propio de Investigación 2018—Programa Contratos-Puente, and the Junta de Andalucía, Consejería de Conocimiento, Investigación y Universidades (ERDF: SOMM17/6107/UGR). This study is part of a PhD thesis defended in the Biomedicine Doctoral Studies Programme of the University of Granada, Spain. Acknowledgments F.A.M., G.S.D., B.M.T., J.M.L.L.E., and J.R.R. designed the study; F.A.M., G.S.D., B.M.T., and F.J.A.G. conducted the experiments; F.A.M., G.S.D., B.M.T., and J.H.M. analyzed the data; F.A.M. wrote the manuscript; F.A.M., G.S.D., B.M.T., J.H.M., F.J.A.G., P.C.N.R., J.M.L.L.E., D.P.B., and J.R.R. critically reviewed of the manuscript; J.R.R. was primarily responsible for the article’s final content. References 1. Carley DW , et al. Physiology of sleep . Diabetes Spectr. 2016 ; 29 ( 1 ): 5 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Spiegel K , et al. Effects of poor and short sleep on glucose metabolism and obesity risk . Nat Rev Endocrinol. 2009 ; 5 ( 5 ): 253 – 261 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Ogilvie RP , et al. The epidemiology of sleep and obesity . Sleep Health. 2017 ; 3 ( 5 ): 383 – 388 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Capers PL , et al. A systematic review and meta-analysis of randomized controlled trials of the impact of sleep duration on adiposity and components of energy balance . Obes Rev. 2015 ; 16 ( 9 ): 771 – 782 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Cannon B , et al. Brown adipose tissue: function and physiological significance . Physiol Rev. 2004 ; 84 ( 1 ): 277 – 359 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Ruiz JR , et al. Role of human brown fat in obesity, metabolism and cardiovascular disease: strategies to turn up the heat . Prog Cardiovasc Dis. 2018 ; 61 ( 2 ): 232 – 245 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Villarroya F , et al. Brown adipose tissue as a secretory organ . Nat Rev Endocrinol. 2017 ; 13 ( 1 ): 26 – 35 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Virtanen KAK , et al. Functional brown adipose tissue in healthy adults . N Engl J Med. 2009 ; 360 ( 15 ): 1518 – 1525 . doi: 10.1056/NEJMoa0808949 Google Scholar Crossref Search ADS PubMed WorldCat 9. van Marken Lichtenbelt WD , et al. Cold-activated brown adipose tissue in healthy men . N Engl J Med. 2009 ; 360 ( 15 ): 1500 – 1508 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Cypess AM , et al. Identification and importance of brown adipose tissue in adult humans . N Engl J Med. 2009 ; 360 ( 15 ): 1509 – 1517 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Saito M , et al. High incidence of metabolically active brown adipose tissue in healthy adult humans: effects of cold exposure and adiposity . Diabetes. 2009 ; 58 ( 7 ): 1526 – 1531 . Google Scholar Crossref Search ADS PubMed WorldCat 12. U Din M , et al. Human brown adipose tissue [15O]O2 PET imaging in the presence and absence of cold stimulus . Eur J Nucl Med Mol Imaging. 2016 ; 43 ( 10 ): 1878 – 1886 . doi: 10.1007/s00259-016-3364-y Google Scholar Crossref Search ADS PubMed WorldCat 13. U Din M , et al. Postprandial oxidative metabolism of human brown fat indicates thermogenesis . Cell Metab. 2018 ;28(2):207–216. doi: 10.1016/j.cmet.2018.05.020 Google Scholar OpenURL Placeholder Text WorldCat Crossref 14. Shukla C , et al. Metabolic signals in sleep regulation: recent insights . Nat Sci Sleep. 2016 ; 8 : 9 – 20 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 15. Orozco-Solis R , et al. The circadian clock in the ventromedial hypothalamus controls cyclic energy expenditure . Cell Metab. 2016 ; 23 ( 3 ): 467 – 478 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Morrison SF , et al. Central neural regulation of brown adipose tissue thermogenesis and energy expenditure . Cell Metab. 2014 ; 19 ( 5 ): 741 – 756 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Morrison SF . Central pathways controlling brown adipose tissue thermogenesis . News Physiol Sci. 2004 ; 19 : 67 – 74 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 18. Chou TC , et al. Afferents to the ventrolateral preoptic nucleus . J Neurosci. 2002 ; 22 ( 3 ): 977 – 990 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Kumar VM , Kumar VM. Body Temperature and Sleep: Are They Controlled by the Same Mechanism? Vol. 2 ; 2004 . https://link.springer.com/content/pdf/10.1111/j.1479-8425.2004.00136.x.pdf. Accessed November 6, 2018 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 20. Glotzbach SF , et al. Central nervous regulation of body temperature during sleep . Science. 1976 ; 194 ( 4264 ): 537 – 539 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Szymusiak R , et al. Hypothalamic regulation of sleep and arousal . Ann N Y Acad Sci. 2008 ; 1129 : 275 – 286 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Kapás L , et al. Brown adipose tissue at the intersection of sleep and temperature regulation . Temperature. 2014 ; 1 ( 1 ): 16 – 17 . doi: 10.4161/temp.29120 Google Scholar Crossref Search ADS WorldCat 23. Kräuchi K , et al. Circadian rhythm of heat production, heart rate, and skin and core temperature under unmasking conditions in men . Am J Physiol . 1994 ; 267 ( 3 ): R819 – R829 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 24. Raymann RJEM , et al. Skin temperature and sleep-onset latency: changes with age and insomnia . Physiol Behav. 2007 ; 90 ( 2–3 ): 257 – 266 . doi: 10.1016/J.PHYSBEH.2006.09.008 Google Scholar Crossref Search ADS PubMed WorldCat 25. Kräuchi K , et al. Warm feet promote the rapid onset of sleep . Nature. 1999 ; 401 ( 6748 ): 36 – 37 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Balzano S , Bergmann BM, Gilliland MA, Silva JE, Rechtschaffen A, Refetoff S. Effect of Total Sleep Deprivation on 5’-Deiodinase Activity of Rat Brown Adipose Tissue . Endocrinology. 1990 ; 127 ( 2 ): 882 – 890 . doi: 10.1210/endo-127-2-882 Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Crossref 27. WEBB WB , et al. Sleep deprivation, age, and exhaustion time in the rat . Science. 1962 ; 136 ( 3522 ): 1122 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Szentirmai É , et al. Intact brown adipose tissue thermogenesis is required for restorative sleep responses after sleep loss . Eur J Neurosci. 2014 ; 39 ( 6 ): 984 – 998 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Szentirmai É , et al. The role of the brown adipose tissue in β3-adrenergic receptor activation-induced sleep, metabolic and feeding responses . Sci Rep. 2017 ; 7 ( 1 ): 958 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Dewasmes G , et al. Activation of brown adipose tissue thermogenesis increases slow wave sleep in rat . Neurosci Lett. 2003 ; 339 ( 3 ): 207 – 210 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Koban M , et al. Chronic REM-sleep deprivation of rats elevates metabolic rate and increases UCP1 gene expression in brown adipose tissue . Am J Physiol Endocrinol Metab. 2005 ; 289 ( 1 ): E68 – E74 . Google Scholar Crossref Search ADS PubMed WorldCat 32. Szentirmai É , et al. Brown adipose tissue plays a central role in systemic inflammation-induced sleep responses . PLoS One. 2018 ; 13 ( 5 ): e0197409 . doi: 10.1371/journal.pone.0197409 Google Scholar Crossref Search ADS PubMed WorldCat 33. U Din M , et al. Human brown fat radiodensity indicates underlying tissue composition and systemic metabolic health . J Clin Endocrinol Metab. 2017 ; 102 ( 7 ): 2258 – 2267 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Copinschi G , et al. The important role of sleep in metabolism. In: E. Ghigo ed., How Gut and Brain Control Metabolism . 2014 ;42: 59 – 72 . doi: 10.1159/000358858 Google Scholar Crossref Search ADS Google Preview WorldCat COPAC 35. Spiegel K , et al. Impact of sleep debt on metabolic and endocrine function . Lancet. 1999 ; 354 ( 9188 ): 1435 – 1439 . Google Scholar Crossref Search ADS PubMed WorldCat 36. Knutson KL , et al. The metabolic consequences of sleep deprivation . Sleep Med Rev. 2007 ; 11 ( 3 ): 163 – 178 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Sanchez-Delgado G , et al. Activating brown adipose tissue through exercise (ACTIBATE) in young adults: rationale, design and methodology . Contemp Clin Trials . 2015 ; 45 ( Pt B ): 416 – 425 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Migueles JH , Cadenas-Sanchez C, Ekelund U, et al. Accelerometer Data Collection and Processing Criteria to Assess Physical Activity and Other Outcomes: A Systematic Review and Practical Considerations. Sports Med . 2017 ; 47 ( 9 ): 1821 – 1845 . doi: 10.1007/s40279-017-0716-0 Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Crossref 39. van Hees VT , et al. Separating movement and gravity components in an acceleration signal and implications for the assessment of human daily physical activity . PLoS One. 2013 ; 8 ( 4 ): e61691 . doi: 10.1371/journal.pone.0061691 Google Scholar Crossref Search ADS PubMed WorldCat 40. van Hees VT , et al. Autocalibration of accelerometer data for free-living physical activity assessment using local gravity and temperature: an evaluation on four continents . J Appl Physiol. 2014 ; 117 ( 7 ): 738 – 744 . Google Scholar Crossref Search ADS PubMed WorldCat 41. Acosta FM , et al. Association of objectively measured physical activity with brown adipose tissue volume and activity in young adults . J Clin Endocrinol Metab. 2019 ; 104 ( 2 ): 223 – 233 . doi: 10.1210/jc.2018-01312 Google Scholar PubMed OpenURL Placeholder Text WorldCat 42. van Hees VT , et al. A novel, open access method to assess sleep duration using a wrist-worn accelerometer . PLoS One. 2015 ; 10 ( 11 ): e0142533 . doi: 10.1371/journal.pone.0142533 Google Scholar Crossref Search ADS PubMed WorldCat 43. Sadeh A . The role and validity of actigraphy in sleep medicine: an update . Sleep Med Rev. 2011 ; 15 ( 4 ): 259 – 267 . doi: 10.1016/j.smrv.2010.10.001 Google Scholar Crossref Search ADS PubMed WorldCat 44. Buysse DJ , et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research . Psychiatry Res. 1989 ; 28 ( 2 ): 193 – 213 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Hildebrand M , et al. Age group comparability of raw accelerometer output from wrist- and hip-worn monitors . Med Sci Sports Exerc. 2014 ; 46 ( 9 ): 1816 – 1824 . Google Scholar Crossref Search ADS PubMed WorldCat 46. Hildebrand M , et al. Evaluation of raw acceleration sedentary thresholds in children and adults . Scand J Med Sci Sports. 2017 ; 27 ( 12 ): 1814 – 1823 . Google Scholar Crossref Search ADS PubMed WorldCat 47. Martinez-Tellez B , et al. The impact of using BARCIST 1.0 criteria on quantification of BAT volume and activity in three independent cohorts of adults . Sci Rep . 2018 ; 8 ( 1 ): 8567 . doi: 10.1038/s41598-018-26878-4 Google Scholar Crossref Search ADS PubMed WorldCat 48. Martinez-Tellez B , et al. A new personalized cooling protocol to activate brown adipose tissue in young adults . Front Physiol. 2017 ; 8 : 863 . Google Scholar Crossref Search ADS PubMed WorldCat 49. Chen KY , et al. Brown Adipose Reporting Criteria in Imaging STudies (BARCIST 1.0): recommendations for Standardized FDG-PET/CT experiments in humans . Cell Metab. 2016 ; 24 ( 2 ): 210 – 222 . Google Scholar Crossref Search ADS PubMed WorldCat 50. Leitner BP , et al. Mapping of human brown adipose tissue in lean and obese young men . Proc Natl Acad Sci USA. 2017 ; 114 ( 32 ): 8649 – 8654 . Google Scholar Crossref Search ADS WorldCat 51. Sanchez-Delgado G , et al. Association between brown adipose tissue and bone mineral density in humans . Int J Obes. 2019 ; 43 : 1516 – 1525 . doi: 10.1038/s41366-018-0261-4 . Google Scholar Crossref Search ADS WorldCat 52. Grandner MA , et al. Criterion validity of the Pittsburgh Sleep Quality Index: investigation in a non-clinical sample . Sleep Biol Rhythms. 2006 ; 4 ( 2 ): 129 – 139 . Google Scholar Crossref Search ADS PubMed WorldCat 53. Enevoldsen LH , et al. Functional brown adipose tissue and sympathetic activity after cold exposure in humans with type 1 narcolepsy . Sleep. 2018 . doi: 10.1093/sleep/zsy092 . Google Scholar OpenURL Placeholder Text WorldCat Crossref 54. Thannickal TC , et al. Reduced number of hypocretin neurons in human narcolepsy . Neuron. 2000 ; 27 ( 3 ): 469 – 474 . Google Scholar Crossref Search ADS PubMed WorldCat 55. Plazzi G , et al. Autonomic disturbances in narcolepsy . Sleep Med Rev. 2011 ; 15 ( 3 ): 187 – 196 . Google Scholar Crossref Search ADS PubMed WorldCat 56. Maloney SK , et al. Translating animal model research: does it matter that our rodents are cold? Physiology (Bethesda). 2014 ; 29 ( 6 ): 413 – 420 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 57. Tan CL , et al. Regulation of body temperature by the nervous system . Neuron. 2018 ; 98 ( 1 ): 31 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat 58. Marlatt KL , et al. Is activation of human brown adipose tissue a viable target for weight management? Am J Physiol Regul Integr Comp Physiol. 2018 ; 315 ( 3 ): R479 – R483 . Google Scholar Crossref Search ADS PubMed WorldCat 59. Acosta FM , et al. Relationship between the Daily Rhythm of Distal Skin Temperature and Brown Adipose Tissue 18F-FDG Uptake in Young Sedentary Adults . J Biol Rhyt . 2019 Aug 7 ; 0748730419865400 . Google Scholar OpenURL Placeholder Text WorldCat 60. Gnad T , et al. Adenosine activates brown adipose tissue and recruits beige adipocytes via A2A receptors . Nature. 2014 ; 516 ( 7531 ): 395 – 399 . Google Scholar Crossref Search ADS PubMed WorldCat 61. Morairty S , et al. Disinhibition of ventrolateral preoptic area sleep-active neurons by adenosine: a new mechanism for sleep promotion . Neuroscience. 2004 ; 123 ( 2 ): 451 – 457 . Google Scholar Crossref Search ADS PubMed WorldCat 62. Reutrakul S , et al. Interactions between sleep, circadian function, and glucose metabolism: implications for risk and severity of diabetes . Ann N Y Acad Sci. 2014 ; 1311 ( 1 ): 151 – 173 . Google Scholar Crossref Search ADS PubMed WorldCat 63. Acosta FM , et al. Correction: physiological responses to acute cold exposure in young lean men . PLoS One. 2018 ; 13 ( 7 ): e0200865 . Google Scholar Crossref Search ADS PubMed WorldCat 64. Schilperoort M , et al. Relevance of lipid metabolism for brown fat visualization and quantification . Curr Opin Lipidol. 2016 ; 27 ( 3 ): 242 – 248 . Google Scholar Crossref Search ADS PubMed WorldCat 65. Martin JL , et al. Wrist actigraphy . Chest. 2011 ; 139 ( 6 ): 1514 – 1527 . Google Scholar Crossref Search ADS PubMed WorldCat © Sleep Research Society 2019. Published by Oxford University Press [on behalf of the Sleep Research Society]. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact [email protected] © Sleep Research Society 2019. Published by Oxford University Press [on behalf of the Sleep Research Society].

Maghsoudipour,, Maryam;Moradi,, Ramin;Pouyakian,, Mostafa;Yaseri,, Mehdi

doi: 10.1093/sleep/zsz193pmid: 33320952

The following author name and affiliation were missing from the originally published abstract. This information has now been added to the article. Sara Moghimi Department of Medicine, University of California, San Diego, La Jolla, CA, USA © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail [email protected]. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Streatfeild,, Jared;Hillman,, David;Adams,, Robert;Mitchell,, Scott;Pezzullo,, Lynne

doi: 10.1093/sleep/zsz181pmid: 31403163

Abstract Study Objectives To determine cost-effectiveness of continuous positive airway pressure (CPAP) treatment of obstructive sleep apnea (OSA) in Australia for 2017–2018 to facilitate public health decision-making. Methods Analysis was undertaken of direct per-person costs of CPAP therapy (according to 5-year care pathways), health system and other costs of OSA and its comorbidities averted by CPAP treatment (5-year adherence rate 56.7%) and incremental benefit of therapy (in terms of disability-adjusted life years [DALYs] averted) to determine cost-effectiveness of CPAP. This was expressed as the incremental cost-effectiveness ratio (= dollars per DALY averted). Direct costs of CPAP were estimated from government reimbursements for services and advertised equipment costs. Costs averted were calculated from both the health care system perspective (health system costs only) and societal perspective (health system plus other financial costs including informal care, productivity losses, nonmedical accident costs, deadweight taxation and welfare losses). These estimates of costs (expressed in US dollars) and DALYs averted were based on our recent analyses of costs of untreated OSA. Results From the health care system perspective, estimated cost of CPAP therapy to treat OSA was $12 495 per DALY averted while from a societal perspective the effect was dominant (−$10 688 per DALY averted) meaning it costs more not to treat the problem than to treat it. Conclusions These estimates suggest substantial community investment in measures to more systematically identify and treat OSA is justified. Apart from potential health and well-being benefits, it is financially prudent to do so. cost-effectiveness studies, OSA, OSA–PAP Therapy Statement of Significance Cost-effectiveness studies are vital in decision making, providing means for priority setting based on cost of the intervention relative to others and willingness to pay. Such studies should provide both healthcare system (healthcare costs) and societal (healthcare plus other economic costs) perspectives. These considerations are particularly relevant to continuous positive airway pressure (CPAP) treatment for obstructive sleep apnea (OSA) because: (1) cost of therapy contributes to its suboptimal uptake, despite its efficacy; and (2) the economic cost of OSA is dominated by non-healthcare costs of lost productivity and nonmedical accident expenses. This analysis demonstrates that while the therapy is highly cost-effective in healthcare terms, it is “dominant” in societal terms, costing less to treat than not to. Introduction Assessing the cost-effectiveness of health interventions is an important guide to public health decision-making as it provides a means for priority setting based on their relative costs and to society’s willingness to pay for them [1, 2]. Continuous positive airway pressure (CPAP) therapy for the treatment of obstructive sleep apnea (OSA) invites this scrutiny. OSA is a common, problematic disorder associated with substantial morbidity and economic cost [3, 4]. CPAP is a highly efficacious therapy for it, yet uptake is suboptimal in part because of cost considerations [5–9]. Yet the expense of implementing and maintaining the therapy must be offset against the costs of untreated OSA which are high, both in terms of health expenditure and, particularly, other societal impacts such as productivity losses, nonmedical accident costs, informal care costs and reduced well-being [4, 10–13]. A number of cost-effectiveness studies of CPAP treatment for OSA have been undertaken in the past, but these have been limited in scope, considering health expenditures but not other costs (or a limited array of them) [14–20], or restricting considerations to severe OSA [16, 21, 22], or to OSA in certain contexts (such as diabetes, cardiovascular disease, vehicle accident risk) rather than OSA more generally [21, 23–26], or failing to fully consider equipment and attendant costs or realistically estimate treatment adherence [27]. Basic requirements of cost-effectiveness analysis include referencing the intervention to a standard comparator, generally existing usual care, and adopting a societal perspective [28]. While healthcare planners often restrict their considerations to the health system perspective, reflecting their primary interest in direct health costs, the societal perspective, which also considers nonhealth care costs, better reflects overall impact on the economy. Our previous analyses of the economic cost of OSA, which include consideration of health and other societal costs, provide a basis on which to consider the cost-effectiveness of CPAP treatment in this comprehensive manner [4, 10–13]. Hence the aim of this study was to assess the cost-effectiveness of CPAP therapy for OSA for the 2017–2018 financial year from both a health system and societal perspective. In doing so we undertook to examine a typical CPAP care pathway to allow evaluation of the cost of attendant investigations and follow-up in addition to that of the equipment used. While the specific costs were derived from Australian sources, the principles underlying the analysis are broadly applicable. Methods The cost-effectiveness of CPAP as a treatment for OSA was analyzed using a two-arm Markov model in which outcomes with treatment were referenced to those with no treatment (“usual care”; Figure 1). Such a comparison allows the costs associated with the intervention and its effects to be assessed in relation to the background costs without it [29]. Cost-effectiveness was assessed from two perspectives: (1) a health care system perspective, where net costs of the intervention and associated changes in health care resource utilization were related to the change in quality of life for people with OSA; and (2) a societal perspective, where the net cost of the intervention incorporated health care resource utilization together with productivity losses, informal care costs and other financial costs, which were then also related to the change in quality of life for people with OSA. Figure 1. Open in new tabDownload slide Model structure. The flow diagram illustrates the estimated proportions of individuals adherent and non-adherent to CPAP and the influence this has on the probability of experiencing related health outcomes. Note: HS = health care system perspective. SC = societal perspective. ** the probability only includes the chance of death due to conditions attributed to OSA. No costs were assigned to the outcome death. Cost outcomes do not include the cost of treatment, which is different for people who are adherent or not, and for those who receive specialist care versus primary care. For further explanation, see text. Figure 1. Open in new tabDownload slide Model structure. The flow diagram illustrates the estimated proportions of individuals adherent and non-adherent to CPAP and the influence this has on the probability of experiencing related health outcomes. Note: HS = health care system perspective. SC = societal perspective. ** the probability only includes the chance of death due to conditions attributed to OSA. No costs were assigned to the outcome death. Cost outcomes do not include the cost of treatment, which is different for people who are adherent or not, and for those who receive specialist care versus primary care. For further explanation, see text. The healthcare costs considered included those directly due to OSA and its treatment and to conditions associated with OSA including cardiovascular disease, diabetes, depression, motor vehicle accidents, and workplace accidents. The costs of these associated conditions were determined using a population attributable fraction (PAF) approach [10, 11]. In addition to these healthcare costs productivity losses, nonmedical accident costs, informal care costs and deadweight losses (efficiency losses associated with the need to raise additional taxation to fund the provision of government services) were calculated using sources and methods we have previously described [4, 10–13]. Incremental cost of therapy Direct costs of CPAP therapy. The cost of CPAP therapy was based on a 5-year treatment pathway modeled on existing Australian and US guidelines [30, 31]. Two care pathway options were considered: (1) a specialist pathway that examined costs associated with specialist sleep physician care and laboratory polysomnography (PSG); and (2) an alternate primary care pathway that utilized a primary care physician and a home-based PSG. The costs of intervention included: (1) the diagnostic sleep study; then (2) follow-up consultation with treatment prescription); then (3) treatment initiation (initial technologist/service provider supervised trial of treatment (minimum 1 week, maximum 3 months, median 3–4 weeks) with a physician follow-up during and/or at the end of trial, and equipment issue/purchase for those who demonstrate benefit and a willingness to proceed with long term treatment. For most (90%) of patients it was assumed that sufficient management data (usage, efficacy, pressure requirement, mask/mouth leak would be available from the CPAP trial device to obviate the need for a follow-up sleep study, but that this would be needed in 10% of cases to deal with problems unable to be solved using these simpler means; and finally (4) long term care with a further physician follow-up annually thereafter, supplemented by minor primary care attendances and additional minor attendances by technologist/nurse services with allowance for replacement of minor equipment (masks, straps, tubing, filters) as required. The costs of the medical services provided were based on the remuneration determined by the Australian Government Medicare Benefits Schedule for the relevant item [32]. The main equipment component—the CPAP device itself—was priced according to a simple average of the cost of all CPAP devices advertised by a major equipment retailer and the average useful working life of the device was estimated to be 6 years. The cost of consumables (CPAP masks, straps, tubing, filters) was also based on average advertised costs, assuming yearly replacement of these. Adverse events with CPAP were generally considered of a minor, preventable or reversible nature and were not included in the modeling. It was assumed that those undergoing treatment would all incur the initial costs, although a conservative allowance (10% discontinuation of CPAP therapy in the first month) was made for those undergoing a trial of treatment, but not proceeding to purchase of it [33]. A conservative 5-year adherence rate (56.7%) was used in the model, which was a weighted average of several studies [33–36] using a definition of adherence of at least 4 hours of CPAP use per night on 70% of nights [37]. It is recognized that current adherence rates might be higher based on improvements in equipment utility and comfort in recent years [38]. The effect of reduction in adherence on follow-up medical and consumable costs was factored into calculation of the net cost of treatment, as was the effect of the proportions of patients taking the specialist pathway versus the primary care pathway which was estimated to be 53% versus 47%, based on the volume of laboratory versus home-based PSG studies currently reimbursed through the Australian Government Medicare Benefits Scheme. Financial benefit of CPAP treatment compared to no CPAP treatment: Health System and Other Costs Averted. The financial benefit of CPAP treatment compared to no treatment is in the health care system and other costs averted. While the model assumes the initial costs, given above, were incurred by those initiated on treatment, only those that adhered with therapy, as defined above, were assumed to have accrued health benefits from it. Except for motor vehicle and workplace accidents, the risks of which immediately reduce on implementation of effective treatment [39], and productivity gains, it was assumed that an individual had to adhere to therapy for 5 years before the health and economic benefits of mitigation of risk of conditions associated with OSA were realized (in the fifth year). As the model considers the average annual costs and benefits of CPAP over 5 years, a discount rate of 3% per annum was applied to future costs to reflect the attrition in future dollar values relative to their current worth. The degree to which CPAP was assumed to prevent conditions attributed to OSA was based on the PAFs before and after treatment. Conservatively, and consistent with estimates in a previous systematic review of interventions for the treatment of OSA in adults [40] it was assumed that the average PAFs before treatment would be for OSA in the moderate range (apnea-hypopnea index [AHI] 15–29.99 events/hour), as the majority of patients prescribed CPAP therapy have moderate to severe OSA, and that these would reduce, with treatment, to PAFs associated with mild OSA (AHI 5–14.99 events/hour). The intervention and no-treatment PAFs for common OSA comorbidities, derived using this approach, are provided in Table 1 along with a relative measure of efficacy derived by relating the change in PAF with treatment to the baseline no-treatment PAF. Table 1. Assumed effectiveness of CPAP treatment for OSA on OSA comorbidities Condition . Intervention PAF (%) . No-treatment PAF (%) . Efficacy . Source . Coronary heart disease 1.4 4.8 70.8% Gottlieb et al. [41] Stroke 2.3 4.8 51.7% Redline et al. [42] Congestive heart failure 0.3 1.5 80.4% Marin et al. [43] Depression 1.7 3.6 52.2% Peppard et al. [44] Diabetes 0.6 1.7 63.9% Wang et al. [45] Vehicle accidents 0.2 3.8 94.7% Antonopoulos et al. [46]; Hillman et al. [10] Workplace accidents 0.0 1.3 100.0% Antonopoulos et al. [46]; Hillman et al. [10] Condition . Intervention PAF (%) . No-treatment PAF (%) . Efficacy . Source . Coronary heart disease 1.4 4.8 70.8% Gottlieb et al. [41] Stroke 2.3 4.8 51.7% Redline et al. [42] Congestive heart failure 0.3 1.5 80.4% Marin et al. [43] Depression 1.7 3.6 52.2% Peppard et al. [44] Diabetes 0.6 1.7 63.9% Wang et al. [45] Vehicle accidents 0.2 3.8 94.7% Antonopoulos et al. [46]; Hillman et al. [10] Workplace accidents 0.0 1.3 100.0% Antonopoulos et al. [46]; Hillman et al. [10] Open in new tab Table 1. Assumed effectiveness of CPAP treatment for OSA on OSA comorbidities Condition . Intervention PAF (%) . No-treatment PAF (%) . Efficacy . Source . Coronary heart disease 1.4 4.8 70.8% Gottlieb et al. [41] Stroke 2.3 4.8 51.7% Redline et al. [42] Congestive heart failure 0.3 1.5 80.4% Marin et al. [43] Depression 1.7 3.6 52.2% Peppard et al. [44] Diabetes 0.6 1.7 63.9% Wang et al. [45] Vehicle accidents 0.2 3.8 94.7% Antonopoulos et al. [46]; Hillman et al. [10] Workplace accidents 0.0 1.3 100.0% Antonopoulos et al. [46]; Hillman et al. [10] Condition . Intervention PAF (%) . No-treatment PAF (%) . Efficacy . Source . Coronary heart disease 1.4 4.8 70.8% Gottlieb et al. [41] Stroke 2.3 4.8 51.7% Redline et al. [42] Congestive heart failure 0.3 1.5 80.4% Marin et al. [43] Depression 1.7 3.6 52.2% Peppard et al. [44] Diabetes 0.6 1.7 63.9% Wang et al. [45] Vehicle accidents 0.2 3.8 94.7% Antonopoulos et al. [46]; Hillman et al. [10] Workplace accidents 0.0 1.3 100.0% Antonopoulos et al. [46]; Hillman et al. [10] Open in new tab These PAFs and efficacy values were then applied to the total costs of these various comorbidities of OSA to determine their costs with relative to without treatment and therefore the incremental cost savings associated with treatment of the condition. These total costs had been estimated (but not published at the time) during a recent analysis of the costs of inadequate sleep conducted by our group, using the national databank resources described in that analysis [10]. As these costs were in 2016–2017 dollars they were inflated to 2017–2018 dollars using the Australian Institute of Health and Welfare (AIHW) health price index. Besides these health system expenditures, other (indirect) financial costs of OSA include productivity losses, informal care costs, costs such as aids and modifications costs, legal costs and insurance costs attributed to vehicle and workplace accidents, as well as less obvious efficiency losses that result from reduced taxation revenue and welfare payments. Sources of productivity loss include early retirement, reduced employment prospects, absenteeism, and presenteeism. The derivation of costs associated with these and other indirect financial costs of OSA are also described in our recent analysis of the costs of inadequate sleep [10]. Again, they have been adjusted from 2016–2017 to 2017–2018 dollars using either the Australian Bureau of Statistics’ Consumer Price Index or Wage Price Index. Individual costs of these OSA-associated conditions prior to treatment were derived from their total costs by dividing these by the number of people affected based the proportion of the Australian population aged 20 and over in 2017–2018 (18.98 million) who have moderate to severe OSA (estimated to be 8.3%) [10, 47]. Incremental benefit: changes in well-being The primary quality of life outcome considered in this analysis was the change in disability-adjusted life years (DALYs) due to OSA and its comorbidities following CPAP therapy. DALYs are derived as the sum of the years of healthy life lost due to disability (YLDs) and the years of life lost due to premature deaths (YLLs) and measured on a scale of zero to one, where zero represents a year of perfect health and one represents death. DALYs were calculated for both the individuals with OSA and for cases of other conditions attributable to OSA (including deaths due to attributable conditions) based on PAFs. The approach used follows that used by our group previously [10, 13]. The average YLDs, YLLs and DALYs per case inform the base case for people with OSA who are not receiving CPAP therapy. The incremental effectiveness (adjusted for adherence) is applied to the base case DALY inputs to estimate the proportion of DALYs that may be avoided through CPAP therapy. Estimating the cost-effectiveness of CPAP To estimate the cost-effectiveness of CPAP, the inputs above were combined as follows: The adherence rate was multiplied by the efficacy parameters provided above to estimate the reduction in the probability of the occurrence of each associated condition through use of CPAP therapy. For example, the probability of a person having stroke due to OSA who is initiated on CPAP therapy and adheres to treatment is estimated to be 0.64% (= 0.91 − (0.91 × 0.57 × 0.52)%) compared to 0.91% for a person who is non-adherent; The change in the proportion of people with associated health conditions was then multiplied by the average cost outcomes for each condition—either health system only, or health system and other financial costs—to determine the average incremental cost saving for each person who adheres to CPAP therapy; The change in well-being was estimated by applying the adherence rate and efficacy parameters for each condition to the average number of DALYs per person, which is then multiplied by the proportion of people with OSA with each condition (e.g. 3.10% for coronary heart disease) to derive the average DALYs avoided per person. OSA alone calculations were further adjusted to remove the proportion of people with OSA and a related health condition so that people were not double-counted. For attributed conditions apart from vehicle and workplace accidents, the well-being benefits occur in the fifth year, so these are discounted appropriately. The average DALYs avoided across all attributed conditions was then calculated as the sum of the discounted well-being benefits, noting again that the benefits for vehicle and workplace accidents occur immediately; and The net cost of treatment was then derived as the cost of treatment minus any cost savings from a reduction in associated conditions or health resource utilization by people with OSA alone. To standardize the cost-benefit of CPAP therapy for OSA against other medical interventions, the net cost of treatment was then divided by the change in well-being to derive the incremental cost-effectiveness ratio (ICER), expressed as dollars per DALY averted Sensitivity analysis A probabilistic sensitivity analysis was conducted by assuming the distribution for the model inputs outlined in the previous sections. Probabilistic sensitivity analysis was conducted with regard to adherence rate; effectiveness of CPAP therapy; health system costs and other financial costs due to OSA or its attributed conditions; change in well-being; probability of an associated event occurring; and the cost of treatment. Each input was allowed to vary according to a PERT distribution, which is defined by the minimum, most likely, and maximum values for each variable. In general, the minimum and maximum values for each distribution were assumed to be 20% lower and higher than the base value, respectively. However, where the effect size was 100%, the maximum value for the distribution was also 100%. A PERT distribution was selected for the analysis as we were unable to obtain exact characteristics of the distribution for each modeling input. The sensitivity analysis was then undertaken using a Monte Carlo simulation with 1000 trials. The Monte Carlo simulation simultaneously draws a random number for each input according to its distribution. Simulating results in this way means that all of the model parameters were allowed to vary at the same time according to the underlying distribution. The ICER was then recalculated for each individual trial to provide an estimate of the sensitivity of the results to each individual parameter. The subsequent results from the model reflect the mean value across all 1000 trials. One-way sensitivity analysis was also conducted using the minimum and maximum values underlying the assumed distributions of each model parameter. For example, the adherence rate was assumed to be 45% or 65%, instead of 56.7% as in the base case. Finally, a tiered sensitivity analysis was conducted to determine the relative influence of sleepiness, accidents, and cardiometabolic and other effects on the estimated ICER of CPAP from both the health care system and societal perspectives. The three scenarios examined were: (1) sleepiness, (2) sleepiness plus accident effects, and (3) sleepiness plus accident plus cardiometabolic and other effects, which was the base case. For the sleepiness scenario, all of the intervention PAFs in Table 1 (each of which relates to an effect of CPAP on an OSA comorbidity) were set to their original (no treatment) values so that there was no incremental difference in the model. For the sleepiness plus accident effects scenario, the intervention PAFs for vehicle and workplace accidents were as shown in Table 1, while all other effects were set to their original (no treatment) values. Currency standardization All costs were expressed in US dollars ($), using the 2017 Organization for Economic Cooperation and Development purchasing power parity of 1.444 Australian dollars per US dollar. Results Direct costs of CPAP therapy The estimated treatment costs of CPAP therapy for specialist and primary care pathways are given in Table 2. The table provides a unit cost, a net present value for a 5-year period, and annual average costs (which assume the life of the CPAP device, supplemented by regular replacement of accessories and spares, is 6 years). The estimated annual direct costs of CPAP therapy for patients adherent to it were $579 for the specialist/laboratory PSG pathway and $504 for the primary care/ home-based PSG pathway. These costs were less for those not adherent to CPAP therapy, which were estimated to be $388 and $289, respectively. The overall average annual direct cost of CPAP treatment, allowing for estimated adherence rates and proportion of patients taking the specialist and primary care pathways (Methods) was estimated to be $457 per person per year. Table 2. Estimated per person CPAP treatment costs in specialist and primary care pathways, over 5 years in Australia, 2017–2018 dollars Treatment protocol . Specialist pathway . Primary care pathway . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Overnight sleep study (level 1) 407 447a 89 232 255a 51 First follow-up consultation 106 106 21 51 51 10 Five annual follow-up consultations 53 251b 50b 26 123b 24b Three minor primary care attendances 26 73b 15b 26 73b 15b Six minor attendances by technicians 19 118 24c 19 118 24c Purchase of CPAP machine 1208 1208 202d 1208 1208 202d CPAP mask, accessories, spare parts 895 179 895 179 Total cost (adherent) 3097 579 2722 504 Total cost (non-adherent) 2125 388 1627 289 Treatment protocol . Specialist pathway . Primary care pathway . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Overnight sleep study (level 1) 407 447a 89 232 255a 51 First follow-up consultation 106 106 21 51 51 10 Five annual follow-up consultations 53 251b 50b 26 123b 24b Three minor primary care attendances 26 73b 15b 26 73b 15b Six minor attendances by technicians 19 118 24c 19 118 24c Purchase of CPAP machine 1208 1208 202d 1208 1208 202d CPAP mask, accessories, spare parts 895 179 895 179 Total cost (adherent) 3097 579 2722 504 Total cost (non-adherent) 2125 388 1627 289 Components may not sum to totals due to rounding. aTen percent greater than unit cost to allow for additional overnight sleep study in approximately 10% of cases to facilitate trouble-shooting in case of treatment implementation difficulties (see Methods). bDiscounting effect over 5 years accounts for difference from unit cost. cPredominantly in first year of therapy and so no discounting applied. dEstimated average life of 6 years. Open in new tab Table 2. Estimated per person CPAP treatment costs in specialist and primary care pathways, over 5 years in Australia, 2017–2018 dollars Treatment protocol . Specialist pathway . Primary care pathway . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Overnight sleep study (level 1) 407 447a 89 232 255a 51 First follow-up consultation 106 106 21 51 51 10 Five annual follow-up consultations 53 251b 50b 26 123b 24b Three minor primary care attendances 26 73b 15b 26 73b 15b Six minor attendances by technicians 19 118 24c 19 118 24c Purchase of CPAP machine 1208 1208 202d 1208 1208 202d CPAP mask, accessories, spare parts 895 179 895 179 Total cost (adherent) 3097 579 2722 504 Total cost (non-adherent) 2125 388 1627 289 Treatment protocol . Specialist pathway . Primary care pathway . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Unit cost ($) . Net present value over 5 years ($) . Annual cost ($) . Overnight sleep study (level 1) 407 447a 89 232 255a 51 First follow-up consultation 106 106 21 51 51 10 Five annual follow-up consultations 53 251b 50b 26 123b 24b Three minor primary care attendances 26 73b 15b 26 73b 15b Six minor attendances by technicians 19 118 24c 19 118 24c Purchase of CPAP machine 1208 1208 202d 1208 1208 202d CPAP mask, accessories, spare parts 895 179 895 179 Total cost (adherent) 3097 579 2722 504 Total cost (non-adherent) 2125 388 1627 289 Components may not sum to totals due to rounding. aTen percent greater than unit cost to allow for additional overnight sleep study in approximately 10% of cases to facilitate trouble-shooting in case of treatment implementation difficulties (see Methods). bDiscounting effect over 5 years accounts for difference from unit cost. cPredominantly in first year of therapy and so no discounting applied. dEstimated average life of 6 years. Open in new tab Financial benefit of CPAP treatment compared to no CPAP treatment: Health System and Other Costs Averted. Table 3 provides estimated average annual health system costs per person affected by OSA and its main comorbidities. Table 4 provides estimates of the other (nonhealth system) financial costs (productivity, informal care, nonmedical accident, and other costs, deadweight loss and total) per person affected by OSA and its main comorbidities. The values provided in these tables are the costs averted in patients with OSA and that proportion of patients at risk of OSA-related comorbidities who effectively treat their OSA through adherence to CPAP therapy. Table 3. Average annual health system costs per case affected, 2017–2018 dollars Condition . Cost ($) . PERT distribution (minimum, mode, maximum) ($) . OSA 60 48, 60, 72 Coronary heart disease 1650 1320, 1650, 1980 Stroke 1668 1334, 1668, 2002 Congestive heart failure 1749 1399, 1749, 2099 Depression 1517 1214, 1517, 1820 Vehicle accidents 3726 2981, 3726, 4471 Workplace accidents 6662 5330, 6662, 7994 Diabetes 436 349, 436, 523 Condition . Cost ($) . PERT distribution (minimum, mode, maximum) ($) . OSA 60 48, 60, 72 Coronary heart disease 1650 1320, 1650, 1980 Stroke 1668 1334, 1668, 2002 Congestive heart failure 1749 1399, 1749, 2099 Depression 1517 1214, 1517, 1820 Vehicle accidents 3726 2981, 3726, 4471 Workplace accidents 6662 5330, 6662, 7994 Diabetes 436 349, 436, 523 Source: Deloitte Access Economics (2017) [13] and Hillman et al (2018) [10]. Open in new tab Table 3. Average annual health system costs per case affected, 2017–2018 dollars Condition . Cost ($) . PERT distribution (minimum, mode, maximum) ($) . OSA 60 48, 60, 72 Coronary heart disease 1650 1320, 1650, 1980 Stroke 1668 1334, 1668, 2002 Congestive heart failure 1749 1399, 1749, 2099 Depression 1517 1214, 1517, 1820 Vehicle accidents 3726 2981, 3726, 4471 Workplace accidents 6662 5330, 6662, 7994 Diabetes 436 349, 436, 523 Condition . Cost ($) . PERT distribution (minimum, mode, maximum) ($) . OSA 60 48, 60, 72 Coronary heart disease 1650 1320, 1650, 1980 Stroke 1668 1334, 1668, 2002 Congestive heart failure 1749 1399, 1749, 2099 Depression 1517 1214, 1517, 1820 Vehicle accidents 3726 2981, 3726, 4471 Workplace accidents 6662 5330, 6662, 7994 Diabetes 436 349, 436, 523 Source: Deloitte Access Economics (2017) [13] and Hillman et al (2018) [10]. Open in new tab Table 4. Average annual productivity and other financial costs per case affected, 2017–2018 dollars Condition . Productivity cost ($) . Informal care cost ($) . Other financial cost ($) . Deadweight loss ($) . Total ($) . PERT distribution (min, mode, max) ($) . OSA 883 - - 105 988 790, 988, 1186 Coronary heart disease 5550 742 - 907 7199 5759, 7199, 8639 Stroke 5921 742 - 944 7607 6086, 7607, 9128 Congestive heart failure 3503 742 - 748 4993 3994, 4993, 5992 Depression 5807 - - 879 6686 5349, 6686, 8023 Vehicle accidents 10 090 3005 33 821 1914 48 861 39 089, 48 861, 58 633 Workplace accidents 62 154 2494 4525 6808 75 982 60 786, 75 982, 91 178 Diabetes 907 67 - 186 1161 929, 1161, 1393 Condition . Productivity cost ($) . Informal care cost ($) . Other financial cost ($) . Deadweight loss ($) . Total ($) . PERT distribution (min, mode, max) ($) . OSA 883 - - 105 988 790, 988, 1186 Coronary heart disease 5550 742 - 907 7199 5759, 7199, 8639 Stroke 5921 742 - 944 7607 6086, 7607, 9128 Congestive heart failure 3503 742 - 748 4993 3994, 4993, 5992 Depression 5807 - - 879 6686 5349, 6686, 8023 Vehicle accidents 10 090 3005 33 821 1914 48 861 39 089, 48 861, 58 633 Workplace accidents 62 154 2494 4525 6808 75 982 60 786, 75 982, 91 178 Diabetes 907 67 - 186 1161 929, 1161, 1393 Source: Deloitte Access Economics (2017) [13] and Hillman et al. [10]. Components may not sum to totals due to rounding. Open in new tab Table 4. Average annual productivity and other financial costs per case affected, 2017–2018 dollars Condition . Productivity cost ($) . Informal care cost ($) . Other financial cost ($) . Deadweight loss ($) . Total ($) . PERT distribution (min, mode, max) ($) . OSA 883 - - 105 988 790, 988, 1186 Coronary heart disease 5550 742 - 907 7199 5759, 7199, 8639 Stroke 5921 742 - 944 7607 6086, 7607, 9128 Congestive heart failure 3503 742 - 748 4993 3994, 4993, 5992 Depression 5807 - - 879 6686 5349, 6686, 8023 Vehicle accidents 10 090 3005 33 821 1914 48 861 39 089, 48 861, 58 633 Workplace accidents 62 154 2494 4525 6808 75 982 60 786, 75 982, 91 178 Diabetes 907 67 - 186 1161 929, 1161, 1393 Condition . Productivity cost ($) . Informal care cost ($) . Other financial cost ($) . Deadweight loss ($) . Total ($) . PERT distribution (min, mode, max) ($) . OSA 883 - - 105 988 790, 988, 1186 Coronary heart disease 5550 742 - 907 7199 5759, 7199, 8639 Stroke 5921 742 - 944 7607 6086, 7607, 9128 Congestive heart failure 3503 742 - 748 4993 3994, 4993, 5992 Depression 5807 - - 879 6686 5349, 6686, 8023 Vehicle accidents 10 090 3005 33 821 1914 48 861 39 089, 48 861, 58 633 Workplace accidents 62 154 2494 4525 6808 75 982 60 786, 75 982, 91 178 Diabetes 907 67 - 186 1161 929, 1161, 1393 Source: Deloitte Access Economics (2017) [13] and Hillman et al. [10]. Components may not sum to totals due to rounding. Open in new tab Allowing for adherence rates and efficacy of treatment in reducing costs associated with OSA and its comorbidities (Table 1) the total costs avoided were $110 dollars per person affected per year from the health care perspective and $1130 per person affected per year from the societal perspective. Improvements in well-being with effective CPAP therapy Table 5 provides the estimated average reductions per case affected in YLDs and YLLs and the consequent DALYs lost through ill-health, disability, or early death associated with untreated OSA and its major comorbidities. Table 5. Average years lost due to disability (YLDs), years of life lost due to premature death (YLLs) and DALYs per case, no treatment Condition . YLDs . YLLs . DALYs . DALY PERT distribution (min, mode, max) . OSA 0.07 - 0.07 0.06, 0.07, 0.09 Coronary heart disease 0.18 0.36 0.53 0.43, 0.53, 0.64 Stroke 0.24 0.42 0.66 0.53, 0.66, 0.79 Congestive heart failure 0.16 0.15 0.31 0.25, 0.31, 0.37 Depression 0.26 0.03 0.29 0.23, 0.29, 0.35 Vehicle accidents 0.15 0.10 0.25 0.20, 0.25, 0.30 Workplace accidents 0.18 0.03 0.21 0.17, 0.21, 0.25 Diabetes 0.17 0.03 0.19 0.15, 0.19, 0.23 Condition . YLDs . YLLs . DALYs . DALY PERT distribution (min, mode, max) . OSA 0.07 - 0.07 0.06, 0.07, 0.09 Coronary heart disease 0.18 0.36 0.53 0.43, 0.53, 0.64 Stroke 0.24 0.42 0.66 0.53, 0.66, 0.79 Congestive heart failure 0.16 0.15 0.31 0.25, 0.31, 0.37 Depression 0.26 0.03 0.29 0.23, 0.29, 0.35 Vehicle accidents 0.15 0.10 0.25 0.20, 0.25, 0.30 Workplace accidents 0.18 0.03 0.21 0.17, 0.21, 0.25 Diabetes 0.17 0.03 0.19 0.15, 0.19, 0.23 Source: Deloitte Access Economics (2017) [13] and Hillman et al. [10]. Open in new tab Table 5. Average years lost due to disability (YLDs), years of life lost due to premature death (YLLs) and DALYs per case, no treatment Condition . YLDs . YLLs . DALYs . DALY PERT distribution (min, mode, max) . OSA 0.07 - 0.07 0.06, 0.07, 0.09 Coronary heart disease 0.18 0.36 0.53 0.43, 0.53, 0.64 Stroke 0.24 0.42 0.66 0.53, 0.66, 0.79 Congestive heart failure 0.16 0.15 0.31 0.25, 0.31, 0.37 Depression 0.26 0.03 0.29 0.23, 0.29, 0.35 Vehicle accidents 0.15 0.10 0.25 0.20, 0.25, 0.30 Workplace accidents 0.18 0.03 0.21 0.17, 0.21, 0.25 Diabetes 0.17 0.03 0.19 0.15, 0.19, 0.23 Condition . YLDs . YLLs . DALYs . DALY PERT distribution (min, mode, max) . OSA 0.07 - 0.07 0.06, 0.07, 0.09 Coronary heart disease 0.18 0.36 0.53 0.43, 0.53, 0.64 Stroke 0.24 0.42 0.66 0.53, 0.66, 0.79 Congestive heart failure 0.16 0.15 0.31 0.25, 0.31, 0.37 Depression 0.26 0.03 0.29 0.23, 0.29, 0.35 Vehicle accidents 0.15 0.10 0.25 0.20, 0.25, 0.30 Workplace accidents 0.18 0.03 0.21 0.17, 0.21, 0.25 Diabetes 0.17 0.03 0.19 0.15, 0.19, 0.23 Source: Deloitte Access Economics (2017) [13] and Hillman et al. [10]. Open in new tab Cost-effectiveness of CPAP therapy Table 6 provides a summary of the estimated cost-effectiveness of CPAP therapy from both a health care system and societal perspective. With either perspective, the estimated costs of treatment and the DALYs averted are the same, but in the health care system estimate only health care system costs avoided with treatment are considered. In the societal perspective, both health care system and other financial costs avoided with treatment are considered. The estimated net cost of CPAP treatment of OSA from the health care perspective was $381 per person affected per year and the ICER value for it was estimated to be $12 495 per DALY averted. From the societal perspective, the estimated net cost of treatment was −$326 per person affected per year and the ICER value for it was correspondingly negative (−$10 688), indicative of a dominant effect. Table 6. Cost-effectiveness analysis . Health care system perspective . Societal perspective . Cost of treatment ($ per person per year) 457 457 Total costs avoided due to OSA ($ per person per year) −76 −783 Net cost ($ per person per year) 381 −326 DALYs averted (per person per year) 0.0305 0.0305 ICER ($/DALY averted) 12 495 Dominant (−10 688) . Health care system perspective . Societal perspective . Cost of treatment ($ per person per year) 457 457 Total costs avoided due to OSA ($ per person per year) −76 −783 Net cost ($ per person per year) 381 −326 DALYs averted (per person per year) 0.0305 0.0305 ICER ($/DALY averted) 12 495 Dominant (−10 688) Open in new tab Table 6. Cost-effectiveness analysis . Health care system perspective . Societal perspective . Cost of treatment ($ per person per year) 457 457 Total costs avoided due to OSA ($ per person per year) −76 −783 Net cost ($ per person per year) 381 −326 DALYs averted (per person per year) 0.0305 0.0305 ICER ($/DALY averted) 12 495 Dominant (−10 688) . Health care system perspective . Societal perspective . Cost of treatment ($ per person per year) 457 457 Total costs avoided due to OSA ($ per person per year) −76 −783 Net cost ($ per person per year) 381 −326 DALYs averted (per person per year) 0.0305 0.0305 ICER ($/DALY averted) 12 495 Dominant (−10 688) Open in new tab Sensitivity analysis Figure 2 shows the results of the probabilistic sensitivity analysis, providing distributions of ICER results for each of the 1000 trials from the Monte Carlo simulation from the health care system (panel A) and societal perspectives (panel B). From the perspective of the health care system, the ICER ranged from $8967 per DALY averted to $17 803 per DALY averted. From the perspective of society, the ICER ranged from −$14 672 per DALY averted (dominant) to −$5913 per DALY averted (dominant). Figure 2. Open in new tabDownload slide Probabilistic sensitivity analysis from the perspective of the health care system (panel A) and society (panel B). The figure shows histograms of the trial results with the associated cumulative distribution functions (black curves), providing an indication of the sampling variability. Figure 2. Open in new tabDownload slide Probabilistic sensitivity analysis from the perspective of the health care system (panel A) and society (panel B). The figure shows histograms of the trial results with the associated cumulative distribution functions (black curves), providing an indication of the sampling variability. Table 7 and Figure 3 show the results of the one-way sensitivity analysis. From a health system perspective, the effectiveness of CPAP had the greatest impact on the estimated ICER, followed by the cost of CPAP therapy, with the ICER ranging between $9495 per DALY averted and $16 246 per DALY averted. The underlying PAFs had a relatively small impact on the ICER. However, from a societal perspective, the cost of CPAP therapy and related conditions had the greatest impact on the estimated ICER. Despite this, the ICER remained dominant for all of the one-way sensitivity analyses. Intuitively, the cost-effectiveness of CPAP therapy was strongly related to the likelihood of related conditions and the costs of related conditions. Table 7. One-way sensitivity analysis illustrating effect of changes (±20%) to major input variables on the estimated ICER ($/DALY averted) from both the health care system and societal perspectives . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Base case 12 495 −10 688 Adherence rate  Lower (45%) 15 430 −7753  Higher (65%) 11 067 −12 116 Effectiveness  Lower (20% less) 16 246 −6937  Higher (20% more) 10 147 −11 913 Well-being loss  Lower (20% less) 15 619 −13 359  Higher (20% more) 10 413 −8906 Treatment cost  Lower (20% less) 9495 −13 688  Higher (20% more) 15 496 −7687 Probability of an untreated associated condition  Lower (20% less) 13 823 −7974  Higher (20% more) 11 336 −13 058 Health costs of OSA  Lower (20% less) 12 603 −10 580  Higher (20% more) 12 388 −10 795 Other costs of OSA  Lower (20% less) 12 495* −8910  Higher (20% more) 12 495* −12 465 Health costs of attributed conditions  Lower (20% less) 12 889 −10 294  Higher (20% more) 12 102 −11 081 Other costs of attributed conditions  Lower (20% less) 12 495* −7829  Higher (20% more) 12 495* −13 546 Discount rate  Lower (0%) 10 952 −11 313  Higher (5%) 13 594 −10 242 . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Base case 12 495 −10 688 Adherence rate  Lower (45%) 15 430 −7753  Higher (65%) 11 067 −12 116 Effectiveness  Lower (20% less) 16 246 −6937  Higher (20% more) 10 147 −11 913 Well-being loss  Lower (20% less) 15 619 −13 359  Higher (20% more) 10 413 −8906 Treatment cost  Lower (20% less) 9495 −13 688  Higher (20% more) 15 496 −7687 Probability of an untreated associated condition  Lower (20% less) 13 823 −7974  Higher (20% more) 11 336 −13 058 Health costs of OSA  Lower (20% less) 12 603 −10 580  Higher (20% more) 12 388 −10 795 Other costs of OSA  Lower (20% less) 12 495* −8910  Higher (20% more) 12 495* −12 465 Health costs of attributed conditions  Lower (20% less) 12 889 −10 294  Higher (20% more) 12 102 −11 081 Other costs of attributed conditions  Lower (20% less) 12 495* −7829  Higher (20% more) 12 495* −13 546 Discount rate  Lower (0%) 10 952 −11 313  Higher (5%) 13 594 −10 242 *Unchanged from base case as these costs are contained outside of health care. Open in new tab Table 7. One-way sensitivity analysis illustrating effect of changes (±20%) to major input variables on the estimated ICER ($/DALY averted) from both the health care system and societal perspectives . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Base case 12 495 −10 688 Adherence rate  Lower (45%) 15 430 −7753  Higher (65%) 11 067 −12 116 Effectiveness  Lower (20% less) 16 246 −6937  Higher (20% more) 10 147 −11 913 Well-being loss  Lower (20% less) 15 619 −13 359  Higher (20% more) 10 413 −8906 Treatment cost  Lower (20% less) 9495 −13 688  Higher (20% more) 15 496 −7687 Probability of an untreated associated condition  Lower (20% less) 13 823 −7974  Higher (20% more) 11 336 −13 058 Health costs of OSA  Lower (20% less) 12 603 −10 580  Higher (20% more) 12 388 −10 795 Other costs of OSA  Lower (20% less) 12 495* −8910  Higher (20% more) 12 495* −12 465 Health costs of attributed conditions  Lower (20% less) 12 889 −10 294  Higher (20% more) 12 102 −11 081 Other costs of attributed conditions  Lower (20% less) 12 495* −7829  Higher (20% more) 12 495* −13 546 Discount rate  Lower (0%) 10 952 −11 313  Higher (5%) 13 594 −10 242 . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Base case 12 495 −10 688 Adherence rate  Lower (45%) 15 430 −7753  Higher (65%) 11 067 −12 116 Effectiveness  Lower (20% less) 16 246 −6937  Higher (20% more) 10 147 −11 913 Well-being loss  Lower (20% less) 15 619 −13 359  Higher (20% more) 10 413 −8906 Treatment cost  Lower (20% less) 9495 −13 688  Higher (20% more) 15 496 −7687 Probability of an untreated associated condition  Lower (20% less) 13 823 −7974  Higher (20% more) 11 336 −13 058 Health costs of OSA  Lower (20% less) 12 603 −10 580  Higher (20% more) 12 388 −10 795 Other costs of OSA  Lower (20% less) 12 495* −8910  Higher (20% more) 12 495* −12 465 Health costs of attributed conditions  Lower (20% less) 12 889 −10 294  Higher (20% more) 12 102 −11 081 Other costs of attributed conditions  Lower (20% less) 12 495* −7829  Higher (20% more) 12 495* −13 546 Discount rate  Lower (0%) 10 952 −11 313  Higher (5%) 13 594 −10 242 *Unchanged from base case as these costs are contained outside of health care. Open in new tab Figure 3. Open in new tabDownload slide Tornado diagram. The figure shows the results of the one-way sensitivity analysis as a tornado diagram from the perspective of the health system (Panel A) and society (Panel B). Figure 3. Open in new tabDownload slide Tornado diagram. The figure shows the results of the one-way sensitivity analysis as a tornado diagram from the perspective of the health system (Panel A) and society (Panel B). Table 8 shows the results of the tiered sensitivity analysis. From a societal perspective, CPAP was no longer a dominant strategy when accidents and cardiometabolic effects were excluded from the analysis. However, CPAP remained highly cost-effective because the treatment costs are offset by productivity gains as the effects of sleepiness are reduced in the target population. Table 8. Tiered sensitivity analysis illustrating the relative influence of sleepiness, sleepiness plus accident effects and sleepiness plus accident plus cardiometabolic effects (the base case) on the estimated ICER ($/DALY averted) from both the health care system and societal perspectives . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Sleepiness effects only 24 203 9329 Sleepiness plus accident effects 21 564 −6690 Sleepiness plus accident plus cardiometabolic effects (base case) 12 495 −10 688 . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Sleepiness effects only 24 203 9329 Sleepiness plus accident effects 21 564 −6690 Sleepiness plus accident plus cardiometabolic effects (base case) 12 495 −10 688 Open in new tab Table 8. Tiered sensitivity analysis illustrating the relative influence of sleepiness, sleepiness plus accident effects and sleepiness plus accident plus cardiometabolic effects (the base case) on the estimated ICER ($/DALY averted) from both the health care system and societal perspectives . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Sleepiness effects only 24 203 9329 Sleepiness plus accident effects 21 564 −6690 Sleepiness plus accident plus cardiometabolic effects (base case) 12 495 −10 688 . Health care system perspective ($/DALY averted) . Societal perspective ($/DALY averted) . Sleepiness effects only 24 203 9329 Sleepiness plus accident effects 21 564 −6690 Sleepiness plus accident plus cardiometabolic effects (base case) 12 495 −10 688 Open in new tab Discussion This study examined the cost-effectiveness of CPAP therapy for OSA and found that from the health care system perspective its estimated net cost was $381 per person affected per year, which equated to an ICER of $12 495 per DALY averted. From a societal perspective (where other financial costs associated with the untreated condition, such as productivity losses and nonmedical accident costs, are also taken into account) the estimated net cost of treatment was −$326 per person affected per year and the ICER value for it was correspondingly negative (−$10 688), indicative of a dominant effect. A dominant effect means that it costs society more not to treat the problem than to treat it. The analysis considers the economic costs broadly and does not attempt to assign the economic benefits of treatment between individuals (such as avoidance of out-of-pocket expenses, increased earnings and greater well-being) and government (such as reduced health system payments and greater taxation revenue). We used “no treatment” as a comparator as this permitted costs associated with the intervention and its effects to be assessed in relation to the background costs without it [29]. Furthermore, while there are other treatments for OSA, which are effective in subgroups of patients, OSA is often under-recognized and either not treated or treated inadequately [48]. The cost-effectiveness model described here differs from earlier work in this field. The first, most obvious difference is that our model considers the effect of CPAP therapy from a societal perspective rather than just the perspective of a health payer or health system more broadly [14–20]. Secondly, our model also considered the effect of CPAP therapy on a broader range of conditions, including effects on cardiovascular health, depression and anxiety, diabetes and workplace accidents, rather than particular contexts (such as diabetes, vascular disease and/or vehicle accident risk) [21, 23–26]. Thirdly, we considered moderate to severe OSA rather than restricting the analysis to severe OSA [16, 21, 22]. Fourthly, we used a shorter time horizon (5 years) than some previous studies [16] to account for limitations in long term adherence data. This is a conservative stance, as CPAP therapy becomes more cost-effective when longer time horizons are considered [16]. In the present study, we modeled the impact of CPAP therapy on the average person within the moderate to severe OSA range. Specifically, the individual costs before and after treatment were derived from estimated total costs averaged across the number of people affected based on the proportion of the adult Australian population with moderate to severe OSA (Methods). This is likely to underestimate the impact of therapy on individuals with more severe OSA, where the effectiveness of CPAP therapy is likely to be greater [49]. It is recognized that there will be considerable variability in costs between individuals around these averages, influenced by factors such as age, gender, other health issues, and socioeconomic circumstances [50]. The incremental costs in the present study are comparable to those reported by the Canadian Agency for Drugs and Technologies in Health (CADTH), which were approximately CAD250 to CAD400 higher per patient per year over the modeled lifetime of the patient [51]. A case-control analysis in the UK found that use of CPAP increased the total NHS cost of patient management by £4141 over 5 years, which is higher than the $1900 estimated in the present study [24]. However, all of the patients in the UK study had both OSA and type 2 diabetes, which is likely to confound differences between the studies [24]. Further, this OSA-diabetes association raises the possibility that the UK study may have involved patients with more severe OSA, although severity was not specified in that study. The costs in the present study are derived from Australian sources, including health system costs. All Australian citizens and permanent residents are covered by a universal health insurance system which provides basic cover for the cost of medical services. This is financed by a compulsory taxation levy. The costs for OSA-related consultations provided here are based on this basic cover. While costs may vary between countries, the principles underlying the analysis are broadly applicable [25, 50, 52, 53]. The present study demonstrates that CPAP therapy offers substantial societal level gains through its mediation of the effect of OSA on sleepiness, accident risk, and related conditions. The value of expressing cost-effectiveness in $ per DALY averted (or its equivalent, $ per quality-adjusted life year (QALY) gained) is that this allows the relative cost of different treatments to be judged against each other and against society’s willingness to pay for them [1]. The threshold within which there is a “willingness to pay” varies between countries according to their wealth and values. In Australian healthcare terms an ICER of $19 413 (95% CI $14 375–$26 085) per QALY gained (equivalent to DALY averted) has been suggested as a base case reference ICER, although higher thresholds ($27 700) are commonly used [2]. At $12 495 per DALY averted CPAP treatment for OSA sits comfortably below the lower bound of this base case ICER reference value. When a societal perspective is considered, the effect is dominant which makes the case for encouraging diagnosis and treatment of the disorder an unarguable economic proposition. While it is understandable that health care planners may try to concentrate on health care rather than the societal perspective, given that their primary concern is in apportioning a health care budget between competing needs, this is particularly inappropriate for chronic conditions such as OSA where the financial costs to the community are very strongly weighted towards potentially readily recoverable nonhealth costs such as productivity loss and nonmedical accident costs [10]. The dominant ICER from a societal perspective directly reflects this. At present, while its impacts on health and well-being are keenly appreciated by its sufferers and their professional and other carers, OSA remains underappreciated for its impacts on communal well-being. It is very common and has very tangible effects on performance and well-being. Productivity losses and accident risk involve the sufferer and those around him or her, at home, in the workplace, and on the roads. Our group have previously investigated the economic cost of sleep disorders and of inadequate sleep [4, 10, 11]. In our investigation of the costs of inadequate sleep the economic costs of moderate to severe OSA were modeled and used in relationship to the issue of symptomatic inadequate sleep, but not published in their own right [10]. However, they are worth considering here. Taking into account the Australian population aged 20 years or older of 18.98 million in 2017–2018, the estimated prevalence of OSA of (8.3%) (as used here), the PAFs for the various comorbidities (Table 1) and the per-person costs and weightings, as outlined in Tables 2–5, the total financial cost of OSA was $3.5 billion. This comprised direct health costs of $0.35 billion, productivity losses of $2.35 billion, informal care costs of $0.1 billion, nonmedical accident costs of $0.39 billion, and deadweight losses of $0.31 billion. OSA also caused a loss of 174 204 DALYs in 2017–18, which represents a nonfinancial cost of $23.62 billion. It is these large economic impacts that make the case for more active pursuit of OSA by health professionals and healthcare planners who have a collective responsibility to find and treat the problem. These consequences and costs are avoidable with effective treatment. CPAP remains a gold standard therapy for OSA given that the problem is eliminated with its use. However, CPAP is an intrusive therapy and is not accepted by all that are exposed to it, particularly those with relatively mild symptoms. Our analysis accounts for this, and while the acceptance rate appears to have improved in recent years, our modeling has adopted a relatively conservative adherence rate at 5 years of 56.7%. Relevant to this, many CPAP intolerant patients can be identified during a supervised trial of treatment prior to purchase of it, avoiding that cost aspect and the costs that would follow. Alternative OSA treatments exist, but CPAP remains the standard against which they are compared [20, 54]. While this analysis was based on Australian costs, the results are likely to be applicable to similar economies worldwide and make an economic case for vigorous communal measures to more systematically identify and treat OSA. Funding This work was supported by the Sleep Health Foundation, an Australian not-for-profit organization devoted to improving sleep health, and an unrestricted grant from the ResMed Foundation which had no part in conception, planning, execution or write-up of it. Conflict of interest statement. None declared. References 1. Brouwer W , et al. When is it too expensive? Cost-effectiveness thresholds and health care decision-making . Eur J Health Econ. 2019 ; 20 ( 2 ): 175 – 180 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Edney LC , et al. Estimating the reference incremental cost-effectiveness ratio for the Australian Health System . PharmacoEcon. 2018 ; 36 ( 2 ): 239 – 252 . Google Scholar Crossref Search ADS WorldCat 3. Franklin KA , et al. Obstructive sleep apnea is a common disorder in the population-a review on the epidemiology of sleep apnea . J Thorac Dis. 2015 ; 7 ( 8 ): 1311 – 1322 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 4. Hillman DR , et al. Public health implications of sleep loss: the community burden . Med J Aust. 2013 ; 199 ( 8 ): S7 – S10 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 5. Bakker JP , et al. Ethnic disparities in CPAP adherence in New Zealand: effects of socioeconomic status, health literacy and self-efficacy . Sleep. 2011 ; 34 ( 11 ): 1595 – 1603 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Catcheside PG . Predictors of continuous positive airway pressure adherence . F1000 Medicine Reports. 2010 ; 2 : 70 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 7. Engleman HM , et al. Improving CPAP use by patients with the Sleep Apnoea/Hypopnoea Syndrome (SAHS) . Sleep Med Rev. 2003 ; 7 ( 1 ): 81 – 99 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Kendzerska T , et al. The effect of patient neighborhood income level on the purchase of continuous positive airway pressure treatment among patients with sleep apnea . Ann Am Thorac Soc. 2016 ; 13 ( 1 ): 93 – 100 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Tarasiuk A , et al. Financial incentive increases CPAP acceptance in patients from low socioeconomic background . PLoS One. 2012 ; 7 ( 3 ): e33178 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Hillman D , et al. The economic cost of inadequate sleep . Sleep. 2018 ; 41 ( 8 ). OpenURL Placeholder Text WorldCat 11. Hillman DR , et al. The economic cost of sleep disorders . Sleep. 2006 ; 29 ( 3 ): 299 – 305 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Deloitte Access Economics . Re-awakening Australia . 2011 . https://www.sleephealthfoundation.org.au/pdfs/news/Reawakening%20Australia.pdf. Accessed May 5, 2019 . 13. Deloitte Access Economics . Asleep on the Job: Costs of Inadequate Sleep in Australia . 2017 . https://www.sleephealthfoundation.org.au/files/Asleep_on_the_job/Asleep_on_the_Job_SHF_report-WEB_small.pdf. Accessed May 5, 2019 . 14. Billings ME , et al. Medicare long-term CPAP coverage policy: a cost-utility analysis . J Clin Sleep Med. 2013 ; 9 ( 10 ): 1023 – 1029 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 15. Català R , et al. Cost-effectiveness of continuous positive airway pressure treatment in moderate-severe obstructive sleep apnea syndrome . Arch Bronconeumol. 2016 ; 52 ( 9 ): 461 – 469 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Guest JF , et al. Cost-effectiveness of using continuous positive airway pressure in the treatment of severe obstructive sleep apnoea/hypopnoea syndrome in the UK . Thorax. 2008 ; 63 ( 10 ): 860 – 865 . Google Scholar Crossref Search ADS PubMed WorldCat 17. McDaid C , et al. Continuous positive airway pressure devices for the treatment of obstructive sleep apnoea-hypopnoea syndrome: a systematic review and economic analysis . Health Technol Assess. 2009 ; 13 ( 4 ): iii – iv , xi–xiv, 1–119, 143–274. Google Scholar Crossref Search ADS WorldCat 18. Pietzsch JB , et al. An integrated health-economic analysis of diagnostic and therapeutic strategies in the treatment of moderate-to-severe obstructive sleep apnea . Sleep. 2011 ; 34 ( 6 ): 695 – 709 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 19. Poullié AI , et al. COST-effectiveness of treatments for mild-to-moderate obstructive sleep apnea in France . Int J Technol Assess Health Care. 2016 ; 32 ( 1-2 ): 37 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Weatherly HL , et al. An economic analysis of continuous positive airway pressure for the treatment of obstructive sleep apnea-hypopnea syndrome . Int J Technol Assess Health Care. 2009 ; 25 ( 1 ): 26 – 34 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Mermigkis C , et al. Cost-effectiveness of CPAP treatment related to cardiovascular disease in patients with severe OSAS: the Greek experience . Sleep Breath. 2015 ; 19 ( 2 ): 737 – 739 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Tan KB , et al. A cost-effectiveness analysis of surgery for middle-aged men with severe obstructive sleep apnea intolerant of CPAP . J Clin Sleep Med. 2015 ; 11 ( 5 ): 525 – 535 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 23. Ayas NT , et al. Cost-effectiveness of continuous positive airway pressure therapy for moderate to severe obstructive sleep apnea/hypopnea . Arch Intern Med. 2006 ; 166 ( 9 ): 977 – 984 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Guest JF , et al. Clinical outcomes and cost-effectiveness of continuous positive airway pressure to manage obstructive sleep apnea in patients with type 2 diabetes in the U.K . Diabetes Care. 2014 ; 37 ( 5 ): 1263 – 1271 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Tan MC , et al. Cost-effectiveness of continuous positive airway pressure therapy in patients with obstructive sleep apnea-hypopnea in British Columbia . Can Respir J. 2008 ; 15 ( 3 ): 159 – 165 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Trakada G , et al. A health-economic analysis of diagnosis and treatment of obstructive sleep apnea with continuous positive airway pressure in relation to cardiovascular disease. The Greek experience . Sleep Breath. 2015 ; 19 ( 2 ): 467 – 472 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Gander P , et al. Societal costs of obstructive sleep apnoea syndrome . N Z Med J. 2010 ; 123 ( 1321 ): 13 – 23 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 28. Shepard DS . Cost‐effectiveness in health and medicine. By MR Gold, JE Siegel, LB Russell, and MC Weinstein (eds). New York: Oxford University Press, 1996 . J Mental Health Policy Econ. 1999 ; 2 ( 2 ): 91 – 92 . Google Scholar Crossref Search ADS WorldCat 29. Owen L , et al. The cost-effectiveness of public health interventions . J Public Health (Oxf). 2012 ; 34 ( 1 ): 37 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Epstein LJ , et al. ; Adult Obstructive Sleep Apnea Task Force of the American Academy of Sleep Medicine . Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults . J Clin Sleep Med. 2009 ; 5 ( 3 ): 263 – 276 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 31. Australasian Sleep Association . Position Paper: Best Practice Guidelines for Provision of CPAP Therapy. Version 2.2 . 2009 . https://www.sleep.org.au/documents/item/66. Accessed May 5, 2019 . 32. Australian Government Department of Health . Medicare Benefits Schedule Book 2018 . http://www.mbsonline.gov.au/internet/mbsonline/publishing.nsf/Content/591044099DA7621ACA2582B2007AF605/$File/201807-MBS%205%20Jul.pdf. Accessed May 5, 2019 . 33. Schoch OD , et al. Baseline predictors of adherence to positive airway pressure therapy for sleep apnea: a 10-year single-center observational cohort study . Respiration. 2014 ; 87 ( 2 ): 121 – 128 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Kohler M , et al. Predictors of long-term compliance with continuous positive airway pressure . Thorax. 2010 ; 65 ( 9 ): 829 – 832 . Google Scholar Crossref Search ADS PubMed WorldCat 35. McArdle N , et al. Long-term use of CPAP therapy for sleep apnea/hypopnea syndrome . Am J Respir Crit Care Med. 1999 ; 159 ( 4 Pt 1 ): 1108 – 1114 . Google Scholar Crossref Search ADS PubMed WorldCat 36. Salepci B , et al. CPAP adherence of patients with obstructive sleep apnea . Respir Care. 2013 ; 58 ( 9 ): 1467 – 1473 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Park JG , et al. Updates on definition, consequences, and management of obstructive sleep apnea . Mayo Clin Proc. 2011 ; 86 ( 6 ): 549 – 554 ; quiz 554. Google Scholar Crossref Search ADS PubMed WorldCat 38. Cistulli P , et al. Real world PAP adherence: results from a big data approach in more than two million patients . In: B109. Treatment Options In Sleep Disordered Breathing: Adherence and Health Outcomes . American Thoracic Society ; 2018 : A4391 – A4391 . https://www.atsjournals.org/doi/abs/10.1164/ajrccm-conference.2018.197.1_MeetingAbstracts.A4391. Accessed May 5, 2019 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 39. Rodenstein D . Sleep apnea: traffic and occupational accidents–individual risks, socioeconomic and legal implications . Respiration. 2009 ; 78 ( 3 ): 241 – 248 . Google Scholar Crossref Search ADS PubMed WorldCat 40. Laratta CR , et al. Diagnosis and treatment of obstructive sleep apnea in adults . CMAJ. 2017 ; 189 ( 48 ): E1481 – E1488 . Google Scholar Crossref Search ADS PubMed WorldCat 41. Gottlieb DJ , et al. Prospective study of obstructive sleep apnea and incident coronary heart disease and heart failure: the sleep heart health study . Circulation. 2010 ; 122 ( 4 ): 352 – 360 . Google Scholar Crossref Search ADS PubMed WorldCat 42. Redline S , et al. Obstructive sleep apnea-hypopnea and incident stroke: the sleep heart health study . Am J Respir Crit Care Med. 2010 ; 182 ( 2 ): 269 – 277 . Google Scholar Crossref Search ADS PubMed WorldCat 43. Marin JM , et al. Long-term cardiovascular outcomes in men with obstructive sleep apnoea-hypopnoea with or without treatment with continuous positive airway pressure: an observational study . Lancet. 2005 ; 365 ( 9464 ): 1046 – 1053 . Google Scholar Crossref Search ADS PubMed WorldCat 44. Peppard PE , et al. Longitudinal association of sleep-related breathing disorder and depression . Arch Intern Med. 2006 ; 166 ( 16 ): 1709 – 1715 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Wang X , et al. Obstructive sleep apnoea and the risk of type 2 diabetes: a meta-analysis of prospective cohort studies . Respirology. 2013 ; 18 ( 1 ): 140 – 146 . Google Scholar Crossref Search ADS PubMed WorldCat 46. Antonopoulos CN , et al. Nasal continuous positive airway pressure (nCPAP) treatment for obstructive sleep apnea, road traffic accidents and driving simulator performance: a meta-analysis . Sleep Med Rev. 2011 ; 15 ( 5 ): 301 – 310 . Google Scholar Crossref Search ADS PubMed WorldCat 47. Adams RJ , et al. Sleep health of Australian adults in 2016: results of the 2016 Sleep Health Foundation national survey . Sleep Health. 2017 ; 3 ( 1 ): 35 – 42 . Google Scholar Crossref Search ADS PubMed WorldCat 48. Kapur V , et al. Underdiagnosis of sleep apnea syndrome in U.S. communities . Sleep Breath. 2002 ; 6 ( 2 ): 49 – 54 . Google Scholar Crossref Search ADS PubMed WorldCat 49. Batool-Anwar S , et al. Impact of Continuous Positive Airway Pressure (CPAP) on quality of life in patients with Obstructive Sleep Apnea (OSA) . J Sleep Res. 2016 ; 25 ( 6 ): 731 – 738 . Google Scholar Crossref Search ADS PubMed WorldCat 50. Tarasiuk A , et al. The economic impact of obstructive sleep apnea . Curr Opin Pulm Med. 2013 ; 19 ( 6 ): 639 – 644 . Google Scholar Crossref Search ADS PubMed WorldCat 51. Kim J , et al. Interventions for the Treatment of Obstructive Sleep Apnea in Adults: A Health Technology Assessment . Ottawa, ON : Canadian Agency for Drugs and Technologies in Health ; 2017 . Google Scholar PubMed OpenURL Placeholder Text Google Scholar Google Preview WorldCat COPAC 52. Albarrak M , et al. Utilization of healthcare resources in obstructive sleep apnea syndrome: a 5-year follow-up study in men using CPAP . Sleep. 2005 ; 28 ( 10 ): 1306 – 1311 . Google Scholar Crossref Search ADS PubMed WorldCat 53. Kapur V , et al. The medical cost of undiagnosed sleep apnea . Sleep. 1999 ; 22 ( 6 ): 749 – 755 . Google Scholar Crossref Search ADS PubMed WorldCat 54. Giles TL , et al. Continuous positive airways pressure for obstructive sleep apnoea in adults . Cochrane Database Syst Rev. 2006 ( 3 ): Cd001106 . OpenURL Placeholder Text WorldCat © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail [email protected]. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Bandarabadi,, Mojtaba;Boyce,, Richard;Gutierrez Herrera,, Carolina;Bassetti, Claudio, L;Williams,, Sylvain;Schindler,, Kaspar;Adamantidis,, Antoine

doi: 10.1093/sleep/zsz182pmid: 31410477

Abstract Theta phase modulates gamma amplitude in hippocampal networks during spatial navigation and rapid eye movement (REM) sleep. This cross-frequency coupling has been linked to working memory and spatial memory consolidation; however, its spatial and temporal dynamics remains unclear. Here, we first investigate the dynamics of theta–gamma interactions using multiple frequency and temporal scales in simultaneous recordings from hippocampal CA3, CA1, subiculum, and parietal cortex in freely moving mice. We found that theta phase dynamically modulates distinct gamma bands during REM sleep. Interestingly, we further show that theta–gamma coupling switches between recorded brain structures during REM sleep and progressively increases over a single REM sleep episode. Finally, we show that optogenetic silencing of septohippocampal GABAergic projections significantly impedes both theta–gamma coupling and theta phase coherence. Collectively, our study shows that phase-space (i.e. cross-frequency coupling) coding of information during REM sleep is orchestrated across time and space consistent with region-specific processing of information during REM sleep including learning and memory. REM sleep, active wake, theta–gamma coupling, optogenetics, hippocampus Significance Classical descriptions of sleep-related brain activity using low spatial resolution electroencephalography led to the distinction between non-rapid eye movement (NREM) and REM (or paradoxical) sleep. Yet, local brain activity during each of these states is characterized by circuit-specific oscillations (slow waves, spindles, and theta). In this study, we show that theta–gamma cross-frequency coupling (CFC) in the hippocampus is dynamically modulated in space and time during single REM sleep episode, distinct from neocortical CFC. These findings extend the multiple criteria used to define REM sleep state in health and disease. Introduction Both rapid eye movement (REM) and non-REM (NREM) sleep are associated with consolidation of some aspects of memory following task acquisition in rodents and humans [1–7]. Classically, the reactivation of hippocampal place cells during NREM sleep provides one possible mechanism for the long-term encoding of newly acquired spatial [1, 8] but also procedural and emotional information [9–11]. However, the underlying mechanisms of memory consolidation during REM sleep are unclear. Recent work involving in-depth analysis of local field potentials (LFPs) recorded from memory-associated brain structures, which includes the hippocampus or neocortex, has provided a possible neural mechanism for memory consolidation [12–14]. One of these includes phase-amplitude cross-frequency coupling (CFC), a phenomenon that describes the increased modulation of fast oscillation amplitude by slow oscillation phase during cognitive and perceptual tasks [15–18]. Specifically, hippocampal and cortical phase-amplitude theta–gamma coupling is prevalent during locomotion and REM sleep in rodents [19, 20] and increases during behavioral and learning tasks in rodents [21, 22], monkeys [23, 24], and humans [25–27]. In addition, dynamic phase entrainment of low-frequency oscillations (<15 Hz) during cognitive tasks and sensory inputs has also been described [28, 29]. Furthermore, impaired phase–amplitude coupling has been reported in Alzheimer’s disease [30, 31], schizophrenia [32], and Parkinson’s disease [33]. Thus, the existence of phase–amplitude coupling and phase entrainment of low frequencies suggests a broad implication of CFC of network oscillations in hierarchical information processing essential to cognitive functions during wakefulness or sleep (for a review see [34]). The hippocampal formation is essential for the learning and storage of information related to fundamental behaviors [35]. Theta and gamma are predominant oscillations during active wake and REM sleep in rodent hippocampal structures and are increased by spatial navigation and learning tasks [36]. Theta rhythm in the hippocampus is generated through extrinsic [3, 37] and intrinsic mechanisms [38, 39]. We recently showed that hippocampal theta rhythm during REM sleep is essential to contextual memory consolidation [3]. Indeed, optogenetic silencing of medial septal GABAergic (MSGABA) neurons during REM sleep (extrinsic inputs) impaired contextual memory of a previously acquired fear-conditioned task in mice. Here, we characterize the phase–amplitude coupling between theta and fast oscillations in hippocampal LFP and cortical electrocorticographic (ECoG) recordings of freely moving mice during spontaneous sleep–wake states and upon optogenetic perturbation of MSGABA neurons during REM sleep. We found that episodes of both active wake and REM sleep showed significant theta–middle gamma coupling, with the highest level of coupling observed during REM sleep. Theta–high gamma coupling was also present during REM sleep, but not during active wake. We further found that theta–gamma coupling is a discrete phenomenon that switches between hippocampal and cortical structures and increases over a single REM sleep episode. Finally, we extended our analysis to previously acquired optogenetic dataset and found that optogenetic silencing of MSGABA neurons impeded phase–amplitude coupling and theta phase coherence in the hippocampus and neocortex during REM sleep. Materials and Methods Animals We used male VGAT-ires-Cre (VGAT::Cre) transgenic mice (Jackson Laboratory). All mice were housed individually in polycarbonate cages at constant temperature (22 ± 1°C), humidity (30%–50%), and light cycle (12 h:12 h, light:dark; lights on at 0800). Mice were allowed unlimited access to food and water. Animals were treated according to protocols and guidelines approved by McGill University and the Canadian Council of Animal Care. Data included in this study were collected from new experimental procedures or a prior study (Figure 2 and Supplementary Figure 4) [3]. Figure 1. Open in new tabDownload slide General patterns of phase-amplitude coupling during active wake and REM sleep. (A) Representative recordings from pyramidal layers of CA3 and subiculum, CA1 radiatum, PCx, and EMG signal during active wake and REM sleep. Power spectral densities obtained from 24-h recordings. Note the distinguishable peaks within fast oscillations during REM sleep. (B) Comodulogram graphs show MI for a wide range of frequency pairs, obtained from 24-h recordings of one animal. (C) Corresponding phase–amplitude histograms for theta–gammaM and either theta–gammaH or theta–gammaUH coupling during active wake and REM sleep for the same animal as in panel (B). The distributions are consistent across animals. Zero and 180 degrees correspond to trough and peak of local theta cycle, respectively. (D) Quantification of theta–gammaM coupling during active wake and REM sleep (REM sleep vs. active wake: P < 0.001; n = 7 animals; two-way ANOVA). Coupling during REM sleep is significantly stronger than active wake within the subiculum and PCx, but not within CA3 and CA1. (E) Example of gammaUH in the pyramidal layer of subiculum during REM sleep. Figure 1. Open in new tabDownload slide General patterns of phase-amplitude coupling during active wake and REM sleep. (A) Representative recordings from pyramidal layers of CA3 and subiculum, CA1 radiatum, PCx, and EMG signal during active wake and REM sleep. Power spectral densities obtained from 24-h recordings. Note the distinguishable peaks within fast oscillations during REM sleep. (B) Comodulogram graphs show MI for a wide range of frequency pairs, obtained from 24-h recordings of one animal. (C) Corresponding phase–amplitude histograms for theta–gammaM and either theta–gammaH or theta–gammaUH coupling during active wake and REM sleep for the same animal as in panel (B). The distributions are consistent across animals. Zero and 180 degrees correspond to trough and peak of local theta cycle, respectively. (D) Quantification of theta–gammaM coupling during active wake and REM sleep (REM sleep vs. active wake: P < 0.001; n = 7 animals; two-way ANOVA). Coupling during REM sleep is significantly stronger than active wake within the subiculum and PCx, but not within CA3 and CA1. (E) Example of gammaUH in the pyramidal layer of subiculum during REM sleep. Virus-mediated targeting of opsin and eYFP expression For optogenetics experiments, transgenic mice were anesthetized with isoflurane (5% induction, 1%–2% maintenance) and placed in a stereotaxic frame at approximately 10 weeks age. Recombinant AAVdj-EF1alpha-DIO-ArchT-eYFP virus (eYFP = enhanced yellow fluorescent protein) (0.6 µL) was stereotactically injected into the medial septum (relative to bregma, in mm; Anterior–Posterior [AP]: +0.86; Medial–Lateral [ML]: 0.0; Dorsal-Ventral [DV] –4.5) of anesthetized mice. An in-depth evaluation of the precision and efficacy of the genetic targeting strategy and ArchT-expression silencing of MSGABA neurons was reported in a prior study [3]. Electrode and optic fiber implantation Tetrodes were made by twisting four individual 17.5 µm diameter platinum–iridium (platinum:iridium 90%:10%) wires into a single strand. For targeting dorsal hippocampal CA3, CA1, and subiculum, leads from 3 tetrodes were then soldered to an electrode interface board and fixed in the correct orientation using epoxy to facilitate implantation as a single array. Tetrode ends were cut to the appropriate length in relation to one another and subsequently cleaned immediately before surgery, giving a measured impedance of approximately 1 MΩ. At 18 weeks age, mice were anesthetized with isoflurane (5% induction, 0.5%–2% maintenance) and subsequently placed in a stereotaxic frame. After clearing the skull of connective tissue and drying with alcohol, holes were drilled in the bone above the dorsal hippocampus (relative to bregma, in mm; CA3: AP: –1.8, ML: +2.1; CA1: AP: –2.45, ML: +1.8; subiculum: AP: –3.10, ML +1.5). The dura was gently cut, and the electrode array was slowly lowered until the tips of the tetrodes in the prearranged array were at the correct depth (CA3: DV: –2.25; CA1: DV: –1.3; subiculum: DV: –1.5; Supplementary Figure 1). A similar procedure was used for implantation of silicon probes (NeuroNexus) used to record laminar LFPs at 50 µm increments oriented vertically through hippocampal CA1, although slightly different coordinates were used for optimal orientation of the probe perpendicular to the CA1 layers (AP: –2.45, ML: +1.5, probe tip DV: –1.5). In all experiments, a screw implanted in the skull above the parietal cortex (AP: –2.3; ML: –1.35), which contacted the surface of brain tissue and served as ECoG, and two stranded tungsten wires inserted into the neck musculature used as EMG signals for assessing postural tone. Screws placed in the bone above the frontal cortex and cerebellum served as ground and reference, respectively. For optogenetics experiments, an optic fiber implant that used to deliver laser light to the MS virus-transfection zone was implanted just above the MS (AP: +0.86; ML: –0.2; DV: –3.82; Supplementary Figure 1). As a final step, dental cement was applied to the skull to permanently secure all components of the implant. In vivo electrophysiological recording Mice were allowed to recover from surgery for at least 1 week. Afterward, a headstage preamplifier tether was attached to a connector on the top of the implanted electrode interface board. Mice were subsequently returned to their home cage where they were allowed to habituate to the tether system for 1 week. Once mice were habituated to the chronic tethering system, a consecutive 24-h recording was completed for all mice. For optogenetics experiments, a 2-h recording session was conducted to assess how disrupting MSGABA activity would influence CFC and theta phase coherence during REM sleep. Mice were continuously monitored until they entered into at least 10 s of stable REM sleep (see criteria in “Determination of vigilance states”). Continuous square pulses of orange light, approximately 7 s in duration, were delivered to the MS area via the optic fiber implant at a frequency of approximately 2 pulses/min. This protocol was continued until mice transitioned to wakefulness, at which point MSGABA silencing ceased until subsequent REM sleep occurred. The intensity of light delivered to the MS was calibrated to approximately 20 mW, a value that, in the absence of ArchT expression, was previously shown to have no effect on baseline electroencephalogram (EEG) characteristics and animal behavior [3]. The precise onset and offset of light delivery were timestamped to neurophysiological recordings. For each session, all recorded signals from the implanted electrodes were amplified by the headstage preamplifier tether before being digitized and acquired at 16 kHz sampling rate using a digital recording system (Neuralynx). To obtain LFP/ECoG/EMG signals, we downsampled raw traces into 1,000 Hz using the “decimate” function of MATLAB, which applies a low pass filter (400 Hz, 8th order Chebyshev Type I) before downsampling to prevent aliasing and spike contamination. Histological confirmation After completion of experiments, mice were deeply anesthetized by intraperitoneal injection of ketamine/xylazine/acepromazide (100, 16, and 3 mg/kg, respectively) and electrode sites were subsequently marked by passing a 10 µA current for approximately 10 s through each tetrode. After lesioning, mice were perfused transcardially with 1× PBS-heparin 0.1%, pH 7.4, followed by 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). Brains were removed from the skull and placed in PFA at 4ºC overnight before being cryoprotected in 30% sucrose dissolved in PBS at 4ºC for an additional 24 h. Brains were then sectioned (50 µm) using a cryostat; half of sections collected were further processed to evaluate accuracy of ArchT-eYFP construct expression, whereas the remaining sections were mounted on gelatin-coated glass slides, stained with cresyl violet, and coverslipped. Final electrode and optic fiber locations were determined by viewing these stained sections with a light microscope. Only mice with confirmed placement of electrodes in the CA3 pyramidal layer, CA1 stratum radiatum (all layers for silicon probe-implanted mice), and the subiculum pyramidal layer, as well as proper optic fiber tip placement immediately above the MS transfection zone for optogenetics experiments, were further analyzed. To confirm accurate targeting of the ArchT-eYFP construct to MSGABA, sections not used for electrode/fiber localization were initially washed in PBS 1×-Triton (0.3%) (PBST) and then incubated for 60 min at room temperature in a blocking solution, composed of 4% bovine serum albumin (BSA) dissolved in PBST, before being incubated in rabbit anti-green fluorescent protein (GFP) (diluted 1:5000 in BSA) overnight at 4°C. To detect the primary antibody, sections were incubated the following day in Alexa Fluor 488 ex anti-rabbit IgG (H + L) (diluted 1:1,000 in PBST) for 1 h at room temperature, before being mounted onto glass slides and coverslipped with Fluoromount-G. Once slides were sufficiently dry, fluorescent images of immunolabelled sections were acquired using a fluorescent microscope. Only mice with confirmed construct expression restricted to neurons localized within the MS and diagonal band regions were further analyzed. Determination of vigilance states Vigilance state was manually scored in 5-s epochs for all recordings through the concurrent evaluation of ECoG signals, EMG-derived muscle activity, and video monitoring of behavior. Quiet wake was defined as awake periods in which mice were immobile with the absence of theta band ECoG activity and tonic muscle activity. Active wake was defined as periods of theta band ECoG activity and EMG bursts of movement-related activity. NREM sleep was identified as periods with a relative high-amplitude, low-frequency EcoG, and reduced muscle tone relative to quiet wake. REM sleep was defined as sustained periods of theta band ECoG activity and behavioral quiescence associated with muscle atonia, save for brief phasic muscle twitches. To avoid mixed states (or transitions), we excluded the ± 5 s epochs during transition periods from our analysis. Polysomnographic scorings were performed double-blinded and were statistically confirmed to lie within a 95% confidence interval (Supplementary Figure 2) [40]. Data analysis All simulations and analyses were performed in MATLAB (R2016a, Natick, MA) environment using custom scripts and built-in functions available on request. Modulation Index We used the Modulation Index (MI) to measure phase–amplitude coupling [22]. We first bandpass-filtered LFP/ECoG signals into low- and high-frequency bands, i.e. theta (6–10 Hz) and middle gamma (50–90 Hz), using finite impulse response (FIR) filters in both forward and reverse directions to eliminate phase distortion (“filtfilt” function, MATLAB). We designed FIR filters using the window-based approach (“fir1” function, MATLAB) with an order equal to three cycles of the low cutoff frequency. We then estimated instantaneous phase of low frequency and the envelope of high frequency oscillations using the Hilbert transform. Then, phase of the low frequency was discretized into 18 equal bins (N = 18, each 20°) and the average value of fast oscillations’ envelope inside each bin was calculated. The resulting phase-amplitude histogram (P) was compared with a uniform distribution (U) using the Kullback–Leibler distance, DKL(P,U)=∑Nj=1P(j)∗log[P(j)/U(j)] ⁠, which was normalized by log(N) to obtain MI, MI=DKL/log(N) ⁠. Comodulogram analysis MI-based comodulogram analysis is a powerful tool to assess phase–amplitude coupling for a wide range of frequencies, as MI is independent of power of fast frequencies and comodulogram graph is not biased for bands with a higher power. To explore coupling between different pairs of frequency bands, we considered 18 frequency bands for phase (0.5–18.5 Hz, 1-Hz increments, 2-Hz bandwidth), and 28 frequency bands for amplitude (20–310 Hz, 10-Hz increments, 20-Hz bandwidth). MI values were then calculated for all these pairs to obtain the comodulogram graph. For each animal, we first filtered continuous 24-h recordings into the mentioned frequency bands and estimated the Hilbert transform. Then, we concatenated episodes of each vigilance state to derive the stage-specific comodulogram graphs. To avoid power line interferences, frequency bands in the vicinity of 60 Hz and its harmonics were reassigned with a 2-Hz safe margin from these interfering frequencies. Time-resolved phase–amplitude coupling We used a time-resolved approach to track changes of phase–amplitude coupling in both time and frequency domains [41]. We first bandpass-filtered the LFP/ECoG signal into 24 sub-bands for amplitude frequency, as described in “Comodulogram Analysis” section, and obtained envelopes of filtered signals using the Hilbert transform. We segmented the resulting envelopes into 4-s windows having 75% overlap and calculated the fast Fourier transform (FFT) for each segment. To minimize the side effects of filtering on signal edges, we applied segmentation after filtering. The frequency band for the phase was obtained from the envelope of the filtered signal for the amplitude frequency band. We estimated the peak frequency of the envelope using the FFT, and then bandpass-filtered LFP/ECoG signals at this peak frequency using a 2-Hz bandwidth. MI was then calculated between the phase of this frequency band and the corresponding high-frequency band. The logic behind this approach is that if there is phase–amplitude coupling between low and high-frequency bands, the modulating low frequency and dominant frequency of envelope of the modulated fast frequency are similar. Therefore, we can track coupling dynamics over time and frequency by replacing phase-frequency axis in the comodulogram (horizontal axis) with the time axis. Theta phase coherence We estimated the phase coherence in the theta range between two signals using mean phase coherence [42], which is the most prominent measure of phase synchronization. To obtain theta phase coherence, we first filtered LFP/ECoG traces in the theta range (6–10 Hz) using a 500th-order FIR filter (“fir1” function, MATLAB) in both the forward and reverse directions. Afterward, we extracted the instantaneous phases of filtered signals using the Hilbert transform and obtained theta phase coherence by averaging the instantaneous phase differences of two filtered signals projected onto a unit circle in the complex plane. This measure has a value between [0 1], where zero and one indicate completely incoherent and coherent theta rhythms, respectively. Spectral analysis We estimated power spectral density (PSD) using Welch’s method (“pwelch” function, MATLAB) with 4-s segments having 75% overlap. Simply, the Welch’s method applies a Hamming window on each segment, calculates the FFT of segments, and then averages over all resulting FFTs. We obtained time–frequency representations using the multitaper method from Chronux signal processing toolbox (window size = 3 s, step size = 0.1 s, tapers [3 5]). Statistical analysis The reported values are mean with the standard error of the mean (mean ± SEM). Two-tailed Wilcoxon signed-rank test and two-way repeated-measures ANOVA were used for the reported statistics, unless otherwise mentioned. To evaluate phase–amplitude coupling results, surrogate data test was performed to test significance for MI values. Surrogate data were generated by shuffling the phase of low-frequency bands 200 times, and MI values were calculated between these shuffled phase time series and the envelope of the amplitude frequency. Reported P values were calculated by dividing the number of surrogate MI values greater than the original MI value by 200. Results LFPs were simultaneously recorded in the hippocampal pyramidal layers of CA3 and subiculum, hippocampal CA1 stratum radiatum, as well as ECoG in the parietal cortex (PCx) across vigilance states in freely moving mice (Figure 1A). We calculated phase–amplitude coupling using the MI [22] and assessed this CFC for a wide range of slow and fast oscillations using the comodulogram analysis. To investigate dynamics of coupling across multiple frequency and temporal scales, we computed the MI using the time-resolved approach. Theta modulates distinct gamma bands during REM sleep To assess general patterns of phase–amplitude coupling during vigilance states for each recorded site, we calculated state-specific comodulograms for active wake, REM sleep, and NREM sleep from continuous 24-h recordings (Figure 1A and B and Supplementary Figure 3). We found that theta phase (6–10 Hz) significantly modulates amplitude of middle gamma band (gammaM: 50–90 Hz) during active wake and REM sleep within the CA3, CA1, subiculum, and PCx recordings (P < 0.001 for all sites, surrogate test; Figure 1B and C). Theta–gammaM coupling significantly increased for subiculum and PCx, but not for CA3 and CA1, recordings during REM sleep compared to active wake (REM sleep vs. active wake: P < 0.001; n = 7 animals; two-way ANOVA; Figure 1D and Supplementary Table 1). During REM sleep, in addition to theta–gammaM coupling, theta phase significantly modulated high gamma (gammaH: 110–160 Hz) in the PCx (P < 0.001, surrogate test), and the ultrahigh gamma (gammaUH: 160–250 Hz) in pyramidal layers of CA3 and subiculum (P < 0.001 for both, surrogate test; Figure 1B–E). Theta–gammaM coupling was stronger than theta–gammaH and theta–gammaUH coupling during REM sleep for all sites (Figure 1B). During NREM sleep, ripples (150–250 Hz) that are generated from synchronous bursts of neural activity in the hippocampal formation, exhibited coupling to sharp waves within the CA3 pyramidal layer and CA1 stratum radiatum (Supplementary Figure 3). Note that NREM ripples of subicular pyramidal layer were phase-coupled to both slow waves (0.5–1.5 Hz) and spindles (Supplementary Figure 3B). To find the preferred theta phases for gamma during active wake and REM sleep, we used phase-amplitude histograms obtained for calculation of the MI. We analyzed CA3, subiculum and PCx recordings that showed distinct coupling patterns both in gammaM and either gammaH or gammaUH during REM sleep (Figure 1C). We found that the gammaUH modulation phase lagged behind gammaM modulation by approximately 90° (~32 ms time delay) in CA3 pyramidal layer, whereas no difference was observed for the subicular pyramidal layer during REM sleep. For the PCx, a slight difference of approximately 20° (~8 ms time delay) was found between preferential theta phases of gammaM and gammaH. Furthermore, comparing preferred theta phase for gammaM during active wake and REM sleep revealed no significant difference (Figure 1C). Collectively, these results showed a state- and region-specific phase–amplitude coupling in the hippocampus and neocortex of mice. Optogenetic silencing of MSGABA neurons suppresses theta–gamma coupling and theta phase coherence We previously showed that optogenetic silencing of MSGABA neurons during REM sleep impairs both object recognition and fear-conditioned contextual memory in mice [3]. To investigate possible underlying mechanisms, we computed theta–gamma coupling, theta phase coherence, and PSD of recordings before, during, and after silencing of MSGABA neurons during REM sleep (Figure 2A and approximately Table 2) from animals recorded in a previous study [3]. We found that optogenetic silencing of MSGABA neurons significantly decreases theta–gamma coupling during REM sleep (silencing vs. baseline; P < 0.001 for all sites and gamma bands; n = 4 animals, 10 silencing trials each; one-way ANOVA with multiple comparison tests; Figure 2B). Furthermore, there was a strong phase coherence in the theta range between recorded sites before and after optogenetic silencing, while significantly decreased during MSGABA silencing (silencing vs. baseline; P < 0.001 for all pairs; n = 4 animals; 10 trials each; Figure 2). Consistent with the previous study [3], we also found that although theta power was significantly reduced during silencing, gamma power showed no significant changes (theta: P < 0.001 for all sites; gamma: P > 0.05 for all sites and gamma bands; n = 4 animals, 10 trials each; one-way ANOVA with multiple comparison tests; Figure 2A and F). Dynamic modulation of theta–gamma coupling during REM sleep To investigate dynamics of the phase–amplitude coupling during REM sleep, we used a time-resolved approach (Figure 3A). We found that theta–gamma coupling “waxes and wanes” over time during a single REM sleep episode that corresponds to a discrete and transient phenomenon lasting for only few seconds with no detectable periodicity. We also used the time-resolved phase-locking value (PLV) as a control measure of phase–amplitude coupling (Figure 3B). Interestingly, this dynamic regulation of CFC is independent of the existence of gamma activity over the whole course of a REM sleep episode (Figure 3C). Figure 3D shows the dominant frequency of fast oscillation envelope, which was used as phase frequency to obtain time-resolved MI and PLV measures in Figure 3A and B. To assess the dependency of theta–gamma coupling strength on theta and gamma power, we calculated the normalized cross-correlation between theta–gammaM coupling and theta/gamma power time series, which were estimated using a 4-s moving window with 75% overlap (Figure 3E and F; Supplementary Figure 5 and Table 3). We found that the normalized cross-correlation between theta–gammaM coupling and theta/gamma power time series is very weak (CA3: MI-theta: 0.19 ± 0.03, MI-gamma: 0.18 ± 0.05; Supplementary Table 3), indicating that an increase in coupling is not a reflection of a higher theta and/or gamma power. We further quantified the normalized cross-correlation between theta power and gamma power (Figure 3G), and we found that their power is independent of each other (CA3: theta–gamma: 0.06 ± 0.03; Supplementary Table 3). These results show that phase–amplitude coupling in hippocampo-cortical structures is dynamically regulated over time, as evidence for other network oscillations, e.g. thalamocortical spindles, during sleep [36, 43]. Theta–gamma coupling is spatially orchestrated during REM sleep The discrete nature of theta–gamma coupling raises the question as to whether CFC is locally modulated. To assess this, we studied dynamics of theta–gammaM coupling from simultaneous LFPs recordings from electrodes located in CA3, CA1, subiculum, and PCx during active wake and REM sleep. Interestingly, we found that theta–gammaM coupling switches between recorded sites, while it coexists during short periods of time in pairs of recordings (Figure 4A). Importantly, theta–gammaM coupling switches did not occur in between CA1 layers as assessed by linear silicon probe recordings (Figure 4B), suggesting uniform modulatory input. To quantify concurrent coupling between hippocampal and neocortical recording sites, we estimated the overlap ratio and cross-correlation between pairs of channels during active wake and REM sleep (Figure 4C and D). Overlap ratio was restricted to periods where at least one region showed high levels of theta–gammaM coupling (1.5 SD + mean) and indicates portion of these times that two sites exhibited theta–gammaM coupling higher than 1.5 SD above the mean, simultaneously. We found that overlap ratio significantly increased during REM sleep compared to active wake for all the studied pairs, except for CA3-PCx (REM sleep vs. active wake: P = 0.002; n = 7 animals; two-way repeated-measures ANOVA; Figure 4C and Supplementary Table 4). Cross-correlation value was obtained considering whole periods during active wake or REM sleep and increased significantly during REM sleep compared to active wake for all the recording pairs (REM sleep vs. active wake: P < 0.001; n = 4 animals; two-way repeated-measures ANOVA; Figure 4D and Supplementary Table 4). To examine the existence of periodicity in coupling, we measured distance between theta–gammaM coupling events (intercoupling interval) within individual active wake and REM episodes. We found that intercoupling interval varies from few seconds to tens of seconds (ECoG: 16.9 ± 0.4 s); however, the average and distribution of intercoupling interval is quite consistent across active wake and REM sleep (Figure 4E; Supplementary Figure 6 and Table 1). To assess the correlation between eye movements and theta–gamma coupling during REM sleep, we recorded simultaneous EEG, EMG, and electrooculogram signals from 4 mice. We estimated the theta–gamma coupling during REM sleep periods with and without eye movement activity, but no significant difference was found between the coupling values of two groups (Supplementary Figure 7). Thus, our results suggest a region-specific modulation of theta–gamma coupling during REM sleep that may support information transfer during sleep-dependent consolidation of information storage [44]. Theta–gamma coupling is strengthened during a single REM episode Temporal representations of theta–gamma coupling during individual REM sleep episodes suggested that it progressively increases over time within a single episode of REM sleep (Figure 3A and Figure 4A). To further verify this, we divided each REM episode into three equal “early/middle/late” segments and calculated theta–gammaM coupling and theta/gammaM power for each segment and normalized middle and late values to the early segment. We included only REM episodes lasting more than 30 s for this analysis to provide segments with at least 10 s length. We found that the theta–gammaM coupling significantly increased across REM sleep for all recorded sites (late vs. early REM segment: P < 0.001 for all sites; n = 7 animals; Wilcoxon test; Figure 5A; Supplementary Table 5). No significant change in theta power across REM sleep was observed (late vs. early REM segment: P > 0.05 for all sites except Sub; n = 7 animals; Wilcoxon test; Figure 5B), whereas gammaM power significantly increased (late vs. early REM segment: P < 0.001 for all sites; n = 7 animals; Wilcoxon test; Figure 5C). We did not find any linear correlation between duration of REM episode and variations in MI. However, shorter REM episodes showed less variations in coupling across episodes (Figure 5D). In CA3 and CA1, average relative MI showed the highest increase for REM episodes with approximately 70 s duration, with decreasing values for longer REM sleep episodes in CA3, whereas remained almost stable in CA1 after approximately 70 s. The MI variations in the subiculum and PCx were less dependent on duration of REM sleep, consistent with the overall modulation of theta–gamma coupling in both space and time. Discussion Our findings show that theta–gamma coupling in the hippocampus and neocortex is dynamically modulated in time and space during REM sleep. Importantly, our results emphasize that this modulation strongly fluctuates within a single REM sleep episode, a previously unappreciated feature of the neural circuits’ activity during this sleep state. Although the mechanism underlying this dynamic modulation of brain activity during REM sleep remains unknown, this modulation likely reflects local coordination of their inputs/outputs at a given time that supports communication between regional networks while minimizing interference [45]. Indeed, coupling predominantly switches between CA3, CA1, subiculum, and PCx networks, although they co-occur in these structures for relatively short periods of time. The overlap ratio and cross-correlation analysis between theta–gammaM coupling events of LFPs recorded from pairs of electrodes exhibited greater synchronized coupling during REM sleep compared to active wake for all pairs (Figure 4C and D). Furthermore, the intercoupling interval histograms had similar distributions for active wake and REM sleep (Supplementary Figure 6), indicating that the increase in concurrent coupling between brain structures during REM sleep, compared to active wake, is due to the way in which theta–gammaM coupling is orchestrated, and not simply the result of increased number of coupling events. These observations can be interpreted as a co-activation of two, or more, circuits during REM sleep, which may provide windows of consolidation of previously acquired information. The increase of theta–gammaM coupling in subicular and cortical, but not CA3 and CA1, structures during REM sleep compared to active wakefulness may reflect a possible role for subiculum and cortex as anatomical and functional outputs in the hippocampal–neocortical dialog supporting long-term memory [44]. One can speculate that considering phase–amplitude coupling as a mechanism for the processing and transmission of information [31, 45, 46], low level of theta–gammaUH coupling during active wake may indicate a low information transfer to neocortex, whereas high level of coupling during REM sleep may be indicative of an information flow out of the hippocampus to the neocortex for long-term storage. Alternatively, one can hypothesize that low gamma coupling favors interregional synchronization, whereas high gamma coupling supports more local synchronization [47, 48]. In partial agreement with this, theta nested gamma oscillations has been hypothesized to provide the basic mechanism for memory retrieval and encoding [49, 50], whereas the occurrence of distinct gamma bands on different theta phases may help to minimize the interference between retrieval and encoding streams [50, 51]. In this context, subicular theta–gammaUH coupling significantly decreased in a mouse model of Alzheimer’s disease [31], which further refers to theta–gammaUH coupling as an indicator of new memory formation. Further investigations are required to experimentally test this hypothesis. Consistent with a possible role of phase–amplitude coupling in memory processing during wakefulness [14, 16, 22], we found that theta–gamma coupling, theta phase coherence, and theta power are strongly decreased during optogenetic silencing of MSGABA neurons (~75%, ~40%, ~55%, respectively), whereas gamma power remained nearly unaltered (~10% reduction). Interestingly, our previous study showed that this optogenetic perturbation completely blocks the consolidation of contextual memory during REM sleep [3]. Thus, a possible mechanistic interpretation of this latter finding is that MSGABA neuron projections to the hippocampus contribute to the modulation of theta–gamma coupling and theta phase coherence during REM sleep, both of which have been shown to be essential for proper short-term [52] or working [53, 54] memory formation and long-term memory formation (e.g. spike-timing-dependent plasticity) [47, 48, 53, 55, 56]. However, this link remains correlative at this point and awaits further investigations as other extrinsic, or intrinsic, inputs modulating theta–gamma coupling in the hippocampus may exist. Figure 2. Open in new tabDownload slide Optogenetic silencing of extrinsic theta projections disrupts theta–gamma coupling and theta phase coherence. (A) Representative recordings before, during, and after silencing of MSGABA neurons during REM sleep. Baseline was considered as 5 s of REM sleep immediately before silencing. (B) Theta phase coherence between CA3, CA1, subiculum, and PCx sites during baseline, silencing, and post silencing. Graphs show average results from four animals, 10 silencing trials each. (C) Difference in theta phase coherence between baseline and silencing. The highest decrease in phase coherence was between CA3 and PCx sites, which dropped from approximately 0.7 to approximately 0.3. (D) Quantification of theta phase coherence. There is a strong theta phase coherence between recorded sites before and after silencing, whereas coherence significantly decreased during MSGABA silencing (silencing vs. baseline; P < 0.001 for all pairs; n = 4 animals, 10 trials each). (E) Comodulogram graphs before, during and after MSGABA silencing estimated from 10 trials of one animal. (F) Quantification of theta–gamma coupling and theta/gamma power during and after silencing compared to the baseline (silencing vs. baseline; P < 0.001 for all sites and gamma bands; n = 4 animals, 10 silencing trials each; one-way ANOVA with multiple comparison tests). Vertical and horizontal asterisks are same. Figure 2. Open in new tabDownload slide Optogenetic silencing of extrinsic theta projections disrupts theta–gamma coupling and theta phase coherence. (A) Representative recordings before, during, and after silencing of MSGABA neurons during REM sleep. Baseline was considered as 5 s of REM sleep immediately before silencing. (B) Theta phase coherence between CA3, CA1, subiculum, and PCx sites during baseline, silencing, and post silencing. Graphs show average results from four animals, 10 silencing trials each. (C) Difference in theta phase coherence between baseline and silencing. The highest decrease in phase coherence was between CA3 and PCx sites, which dropped from approximately 0.7 to approximately 0.3. (D) Quantification of theta phase coherence. There is a strong theta phase coherence between recorded sites before and after silencing, whereas coherence significantly decreased during MSGABA silencing (silencing vs. baseline; P < 0.001 for all pairs; n = 4 animals, 10 trials each). (E) Comodulogram graphs before, during and after MSGABA silencing estimated from 10 trials of one animal. (F) Quantification of theta–gamma coupling and theta/gamma power during and after silencing compared to the baseline (silencing vs. baseline; P < 0.001 for all sites and gamma bands; n = 4 animals, 10 silencing trials each; one-way ANOVA with multiple comparison tests). Vertical and horizontal asterisks are same. Figure 3. Open in new tabDownload slide Dynamics of phase–amplitude coupling during REM sleep. (A, B) Time-resolved MI and PLV measures estimated from PCx across a REM sleep episode. Horizontal and vertical axes are time and modulated frequencies (i.e. gamma range), respectively. Blue horizontal line in the top of graphs indicates REM sleep. (C) Normalized power of fast oscillations during REM sleep. Power of amplitude-frequency bands was normalized to their first 10 s average power to provide meaningful comparison between frequency bands. (D) Dominant frequency of the envelope of fast oscillations, which is considered as phase frequency or modulating frequency, in the time-resolved approach. All measures were estimated using a moving window of 4 s having 75% overlap during a REM sleep episode. (E, F) Normalized cross-correlation between theta–gammaM coupling and theta/gamma power time series, which were estimated using a 4-s moving window with 75% overlap for all the recording sites. (G) Same as panel (E), except for theta and gamma power. The normalized cross-correlation between theta–gammaM coupling and theta/gamma power time series, as well as between theta and gamma power time series are very weak (n = 7 animals). Figure 3. Open in new tabDownload slide Dynamics of phase–amplitude coupling during REM sleep. (A, B) Time-resolved MI and PLV measures estimated from PCx across a REM sleep episode. Horizontal and vertical axes are time and modulated frequencies (i.e. gamma range), respectively. Blue horizontal line in the top of graphs indicates REM sleep. (C) Normalized power of fast oscillations during REM sleep. Power of amplitude-frequency bands was normalized to their first 10 s average power to provide meaningful comparison between frequency bands. (D) Dominant frequency of the envelope of fast oscillations, which is considered as phase frequency or modulating frequency, in the time-resolved approach. All measures were estimated using a moving window of 4 s having 75% overlap during a REM sleep episode. (E, F) Normalized cross-correlation between theta–gammaM coupling and theta/gamma power time series, which were estimated using a 4-s moving window with 75% overlap for all the recording sites. (G) Same as panel (E), except for theta and gamma power. The normalized cross-correlation between theta–gammaM coupling and theta/gamma power time series, as well as between theta and gamma power time series are very weak (n = 7 animals). Figure 4. Open in new tabDownload slide Theta–gamma coupling swings between hippocampal and cortical circuitries during REM sleep. (A) Top graph shows dynamics of theta-gammaM coupling over a REM sleep episode for CA3, CA1, subiculum, and PCx recordings. Coupling mainly switches between networks, but co-occurs during short periods of time. The MI measure was estimated using a moving window of 4 s having 75% overlap, and then smoothed by a 10-point Hanning window. Bottom graph illustrates corresponding time-frequency representation. (B) Same as panel (A), except for different layers of CA1. Theta–gammaM coupling is highly synchronized across the layers of CA1. (C) Quantification of coupling concurrency between recording pairs during active wake and REM sleep. Overlap ratio indicates the proportion of times that two regions simultaneously show theta–gammaM coupling higher than 1.5 SD above the mean. Overlap ratio significantly increased during REM sleep compared to active wake for all the studied pairs, except for CA3-PCx (REM sleep vs. active wake: P = 0.002; n = 7 animals; two-way repeated measures ANOVA). (D) Normalized cross-correlation between theta–gammaM coupling time series of different recording pairs that is obtained considering whole periods during active wake or REM sleep. Normalized cross-correlation increased significantly during REM sleep compared to active wake for all the recording pairs (REM sleep vs. active wake: P < 0.001; n = 4 animals; two-way repeated measures ANOVA). (E) Average distance between theta–gammaM coupling episodes. Intercoupling interval indicates silent periods between two coupling events. The average intercoupling interval is quite consistent across active wake and REM sleep. Figure 4. Open in new tabDownload slide Theta–gamma coupling swings between hippocampal and cortical circuitries during REM sleep. (A) Top graph shows dynamics of theta-gammaM coupling over a REM sleep episode for CA3, CA1, subiculum, and PCx recordings. Coupling mainly switches between networks, but co-occurs during short periods of time. The MI measure was estimated using a moving window of 4 s having 75% overlap, and then smoothed by a 10-point Hanning window. Bottom graph illustrates corresponding time-frequency representation. (B) Same as panel (A), except for different layers of CA1. Theta–gammaM coupling is highly synchronized across the layers of CA1. (C) Quantification of coupling concurrency between recording pairs during active wake and REM sleep. Overlap ratio indicates the proportion of times that two regions simultaneously show theta–gammaM coupling higher than 1.5 SD above the mean. Overlap ratio significantly increased during REM sleep compared to active wake for all the studied pairs, except for CA3-PCx (REM sleep vs. active wake: P = 0.002; n = 7 animals; two-way repeated measures ANOVA). (D) Normalized cross-correlation between theta–gammaM coupling time series of different recording pairs that is obtained considering whole periods during active wake or REM sleep. Normalized cross-correlation increased significantly during REM sleep compared to active wake for all the recording pairs (REM sleep vs. active wake: P < 0.001; n = 4 animals; two-way repeated measures ANOVA). (E) Average distance between theta–gammaM coupling episodes. Intercoupling interval indicates silent periods between two coupling events. The average intercoupling interval is quite consistent across active wake and REM sleep. Figure 5. Open in new tabDownload slide Theta–gamma coupling strength increases over REM sleep episodes. (A) Normalized theta–gammaM coupling during middle and late phases of REM sleep episodes for CA3, CA1, subiculum, and PCx recordings. For each REM episode, middle and late coupling values were normalized to the early segment. Theta–gammaM coupling significantly increased for all sites (late vs. early REM segment: P < 0.001 for all sites; n = 7 animals; Wilcoxon test). (B, C) Same graphs as panel (A) but for theta and gammaM power, respectively. Theta power showed no significant changes across REM sleep, except within the subiculum, whereas gammaM power significantly increased (late vs. early REM segment: theta: P > 0.05; gammaM: P < 0.001 for all sites; n = 7 animals; Wilcoxon test). (D) Relation between duration of REM episodes and variations in theta–gammaM coupling. Each point represents one REM sleep episode and lines indicate mean ± SEM of points. Figure 5. Open in new tabDownload slide Theta–gamma coupling strength increases over REM sleep episodes. (A) Normalized theta–gammaM coupling during middle and late phases of REM sleep episodes for CA3, CA1, subiculum, and PCx recordings. For each REM episode, middle and late coupling values were normalized to the early segment. Theta–gammaM coupling significantly increased for all sites (late vs. early REM segment: P < 0.001 for all sites; n = 7 animals; Wilcoxon test). (B, C) Same graphs as panel (A) but for theta and gammaM power, respectively. Theta power showed no significant changes across REM sleep, except within the subiculum, whereas gammaM power significantly increased (late vs. early REM segment: theta: P > 0.05; gammaM: P < 0.001 for all sites; n = 7 animals; Wilcoxon test). (D) Relation between duration of REM episodes and variations in theta–gammaM coupling. Each point represents one REM sleep episode and lines indicate mean ± SEM of points. Finally, the synchronization of circuit oscillations extended beyond theta–gamma coupling during REM sleep [57, 58]. During NREM sleep, slow waves, spindles, and hippocampal sharp wave-ripples (SWRs) are temporally coupled, a mechanism thought to facilitate hippocampal–neocortical transfer during sleep-dependent memory consolidation [57–61]. Hippocampal SWRs involve cycles of 50–150 ms bursting spikes, which propagate along the CA3-CA1-subiculum axis [62]. They are associated with the replay of place cells and synaptic activities critical for the encoding of space/context/time [63–65]. Coupling between ripples and spindles in human has been previously reported in hippocampal [66] and para-hippocampal sites [57]. Consistent with previous observations in rodents [67], monkeys [68], and humans [66], our findings show that hippocampal ripples are also phase-locked to both spindles and slow oscillations within the subiculum in mice, although for CA3 and CA1 recording sites ripples were only coupled to sharp waves (Supplementary Figure 3). The coordination of spindles with ripple oscillations in the subiculum is consistent with a possible mechanism mediating hippocampal–neocortical dialog during NREM sleep [44, 58]. How these phenomena during NREM sleep relate to theta–gamma coupling, or theta coherence, during REM sleep and the overall memory consolidation remains unclear. They may represent distinct, but complementary, pathways for the consolidation of different types of memories, although this awaits further investigation. Transition between wake, NREM and REM sleep are relatively slow (i.e. several seconds) and implicate many different cell types including neurons and glia. Further insights in the temporal and spatial dynamics of brain-wide distributed network controlling those states will refine the transitions between states. In this context, MI represents an additional criterion to define REM sleep from either its onset from NREM or its termination by wakefulness in rodents, as well as NREM in human. Consistent with the region-specific regulation of sleep oscillations during NREM [9, 69] and REM [70] sleep, our findings further extend the identification of region-specific coupling of oscillatory networks during REM sleep. Our results also emphasize the spatiotemporal dynamic modulation of neural network activity within single REM sleep episode and its possible implication in sleep-dependent functions. Funding M.B. was supported by the Inselspital University Hospital, the University of Bern. A.R.A. was supported by the Human Frontier Science Program (RGY0076/2012), Inselspital University Hospital, the University of Bern, Swiss National Science Foundation (156156), and the European Research Council (ERC-2016-COG-725850). Acknowledgments We thank the Tidis laboratory members for insightful comments and discussions on a previous version of this manuscript. Disclosure The authors declare no competing financial interests. References 1. Wilson MA , et al. Reactivation of hippocampal ensemble memories during sleep . Science. 1994 ; 265 ( 5172 ): 676 – 679 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Siegel JM . The REM sleep-memory consolidation hypothesis . Science. 2001 ; 294 ( 5544 ): 1058 – 1063 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Boyce R , et al. Causal evidence for the role of REM sleep theta rhythm in contextual memory consolidation . Science. 2016 ; 352 ( 6287 ): 812 – 816 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Tononi G , et al. Sleep and the price of plasticity: from synaptic and cellular homeostasis to memory consolidation and integration . Neuron. 2014 ; 81 ( 1 ): 12 – 34 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Lee ML , et al. Fragmentation of rapid eye movement and nonrapid eye movement sleep without total sleep loss impairs hippocampus-dependent fear memory consolidation . Sleep. 2016 ; 39 ( 11 ): 2021 – 2031 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Stickgold R , et al. Sleep, learning, and dreams: off-line memory reprocessing . Science. 2001 ; 294 ( 5544 ): 1052 – 1057 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Fraize N , et al. Levels of interference in long and short-term memory differentially modulate non-REM and REM sleep . Sleep. 2016 ; 39 ( 12 ): 2173 – 2188 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Nadasdy Z , et al. Replay and time compression of recurring spike sequences in the hippocampus . J Neurosci. 1999 ; 19 ( 21 ): 9497 – 9507 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Huber R , et al. Local sleep and learning . Nature. 2004 ; 430 ( 6995 ): 78 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Aeschbach D , et al. A role for non-rapid-eye-movement sleep homeostasis in perceptual learning . J Neurosci. 2008 ; 28 ( 11 ): 2766 – 2772 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Leong RLF , Koh SYJ , Chee MWL , Lo JC . Slow wave sleep facilitates spontaneous retrieval in prospective memory . Sleep. 2019 ; 42 ( 4 ), zsz003 . doi: 10.1093/sleep/zsz003 . Google Scholar Crossref Search ADS PubMed WorldCat Crossref 12. Scheffer-Teixeira R , et al. Theta phase modulates multiple layer-specific oscillations in the CA1 region . Cereb Cortex. 2012 ; 22 ( 10 ): 2404 – 2414 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Jensen O , et al. Cross-frequency coupling between neuronal oscillations . Trends Cogn Sci. 2007 ; 11 ( 7 ): 267 – 269 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Tort AB , et al. Theta-gamma coupling increases during the learning of item-context associations . Proc Natl Acad Sci USA. 2009 ; 106 ( 49 ): 20942 – 20947 . Google Scholar Crossref Search ADS WorldCat 15. van Wingerden M , et al. Phase-amplitude coupling in rat orbitofrontal cortex discriminates between correct and incorrect decisions during associative learning . J Neurosci. 2014 ; 34 ( 2 ): 493 – 505 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Axmacher N , et al. Cross-frequency coupling supports multi-item working memory in the human hippocampus . Proc Natl Acad Sci USA. 2010 ; 107 ( 7 ): 3228 – 3233 . Google Scholar Crossref Search ADS WorldCat 17. Hyafil A , et al. Speech encoding by coupled cortical theta and gamma oscillations . Elife. 2015 ; 4 : e06213 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Sweeney-Reed CM , et al. Corticothalamic phase synchrony and cross-frequency coupling predict human memory formation . Elife. 2014 ; 3 : e05352 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Scheffzük C , et al. Selective coupling between theta phase and neocortical fast gamma oscillations during REM-sleep in mice . PLoS One. 2011 ; 6 ( 12 ): e28489 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Schomburg EW , et al. Theta phase segregation of input-specific gamma patterns in entorhinal-hippocampal networks . Neuron. 2014 ; 84 ( 2 ): 470 – 485 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Belluscio MA , et al. Cross-frequency phase-phase coupling between θ and γ oscillations in the hippocampus . J Neurosci. 2012 ; 32 ( 2 ): 423 – 435 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Tort AB , et al. Dynamic cross-frequency couplings of local field potential oscillations in rat striatum and hippocampus during performance of a T-maze task . Proc Natl Acad Sci USA. 2008 ; 105 ( 51 ): 20517 – 20522 . Google Scholar Crossref Search ADS WorldCat 23. Voloh B , et al. Theta-gamma coordination between anterior cingulate and prefrontal cortex indexes correct attention shifts . Proc Natl Acad Sci USA. 2015 ; 112 ( 27 ): 8457 – 8462 . Google Scholar Crossref Search ADS WorldCat 24. Takeuchi S , et al. Gamma oscillations and their cross-frequency coupling in the primate hippocampus during sleep . Sleep. 2015 ; 38 ( 7 ): 1085 – 1091 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Cohen MX , et al. Oscillatory activity and phase-amplitude coupling in the human medial frontal cortex during decision making . J Cogn Neurosci. 2009 ; 21 ( 2 ): 390 – 402 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Canolty RT , et al. High gamma power is phase-locked to theta oscillations in human neocortex . Science. 2006 ; 313 ( 5793 ): 1626 – 1628 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Watrous AJ , et al. Phase-amplitude coupling supports phase coding in human ECoG . Elife. 2015 ; 4 : e07886 . Google Scholar Crossref Search ADS WorldCat 28. Schroeder CE , et al. Low-frequency neuronal oscillations as instruments of sensory selection . Trends Neurosci. 2009 ; 32 ( 1 ): 9 – 18 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Stefanics G , et al. Phase entrainment of human delta oscillations can mediate the effects of expectation on reaction speed . J Neurosci. 2010 ; 30 ( 41 ): 13578 – 13585 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Zhang X , et al. Impaired theta-gamma coupling in APP-deficient mice . Sci Rep. 2016 ; 6 : 21948 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Goutagny R , et al. Alterations in hippocampal network oscillations and theta-gamma coupling arise before Aβ overproduction in a mouse model of Alzheimer’s disease . Eur J Neurosci. 2013 ; 37 ( 12 ): 1896 – 1902 . Google Scholar Crossref Search ADS PubMed WorldCat 32. Barr MS , et al. Impaired theta-gamma coupling during working memory performance in schizophrenia . Schizophr Res. 2017 ; 189 : 104 – 110 . Google Scholar Crossref Search ADS PubMed WorldCat 33. de Hemptinne C , et al. Exaggerated phase-amplitude coupling in the primary motor cortex in Parkinson disease . Proc Natl Acad Sci USA. 2013 ; 110 ( 12 ): 4780 – 4785 . Google Scholar Crossref Search ADS WorldCat 34. Canolty RT , et al. The functional role of cross-frequency coupling . Trends Cogn Sci. 2010 ; 14 ( 11 ): 506 – 515 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Jarrard LE . On the role of the hippocampus in learning and memory in the rat . Behav Neural Biol. 1993 ; 60 ( 1 ): 9 – 26 . Google Scholar Crossref Search ADS PubMed WorldCat 36. Montgomery SM , et al. Theta and gamma coordination of hippocampal networks during waking and rapid eye movement sleep . J Neurosci. 2008 ; 28 ( 26 ): 6731 – 6741 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Buzsáki G . Theta oscillations in the hippocampus . Neuron. 2002 ; 33 ( 3 ): 325 – 340 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Goutagny R , et al. Self-generated theta oscillations in the hippocampus . Nat Neurosci. 2009 ; 12 ( 12 ): 1491 – 1493 . Google Scholar Crossref Search ADS PubMed WorldCat 39. Amilhon B , et al. Parvalbumin interneurons of hippocampus tune population activity at theta frequency . Neuron. 2015 ; 86 ( 5 ): 1277 – 1289 . Google Scholar Crossref Search ADS PubMed WorldCat 40. Adamantidis AR , et al. Neural substrates of awakening probed with optogenetic control of hypocretin neurons . Nature. 2007 ; 450 ( 7168 ): 420 – 424 . Google Scholar Crossref Search ADS PubMed WorldCat 41. Cohen MX . Assessing transient cross-frequency coupling in EEG data . J Neurosci Methods. 2008 ; 168 ( 2 ): 494 – 499 . Google Scholar Crossref Search ADS PubMed WorldCat 42. Mormann F , et al. Mean phase coherence as a measure for phase synchronization and its application to the EEG of epilepsy patients . Physica D. 2000 ; 144 ( 3–4 ): 358 – 369 . Google Scholar Crossref Search ADS WorldCat 43. Helfrich RF , et al. Old brains come uncoupled in sleep: slow wave-spindle synchrony, brain atrophy, and forgetting . Neuron. 2018 ; 97 ( 1 ): 221 – 230.e4 . Google Scholar Crossref Search ADS PubMed WorldCat 44. Diekelmann S , et al. The memory function of sleep . Nat Rev Neurosci. 2010 ; 11 ( 2 ): 114 – 126 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Buzsáki G , et al. Mechanisms of gamma oscillations . Annu Rev Neurosci. 2012 ; 35 : 203 – 225 . Google Scholar Crossref Search ADS PubMed WorldCat 46. Sirota A , et al. Entrainment of neocortical neurons and gamma oscillations by the hippocampal theta rhythm . Neuron. 2008 ; 60 ( 4 ): 683 – 697 . Google Scholar Crossref Search ADS PubMed WorldCat 47. Fries P . A mechanism for cognitive dynamics: neuronal communication through neuronal coherence . Trends Cogn Sci. 2005 ; 9 ( 10 ): 474 – 480 . Google Scholar Crossref Search ADS PubMed WorldCat 48. Axmacher N , et al. Memory formation by neuronal synchronization . Brain Res Rev. 2006 ; 52 ( 1 ): 170 – 182 . Google Scholar Crossref Search ADS PubMed WorldCat 49. Colgin LL . Theta-gamma coupling in the entorhinal-hippocampal system . Curr Opin Neurobiol. 2015 ; 31 : 45 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat 50. Colgin LL , et al. Frequency of gamma oscillations routes flow of information in the hippocampus . Nature. 2009 ; 462 ( 7271 ): 353 – 357 . Google Scholar Crossref Search ADS PubMed WorldCat 51. Hasselmo ME , et al. A proposed function for hippocampal theta rhythm: separate phases of encoding and retrieval enhance reversal of prior learning . Neural Comput. 2002 ; 14 ( 4 ): 793 – 817 . Google Scholar Crossref Search ADS PubMed WorldCat 52. Lisman JE , et al. Storage of 7 +/- 2 short-term memories in oscillatory subcycles . Science. 1995 ; 267 ( 5203 ): 1512 – 1515 . Google Scholar Crossref Search ADS PubMed WorldCat 53. Fell J , et al. The role of phase synchronization in memory processes . Nat Rev Neurosci. 2011 ; 12 ( 2 ): 105 – 118 . Google Scholar Crossref Search ADS PubMed WorldCat 54. Lisman JE , et al. The θ-γ neural code . Neuron. 2013 ; 77 ( 6 ): 1002 – 1016 . Google Scholar Crossref Search ADS PubMed WorldCat 55. Benchenane K , et al. Coherent theta oscillations and reorganization of spike timing in the hippocampal- prefrontal network upon learning . Neuron. 2010 ; 66 ( 6 ): 921 – 936 . Google Scholar Crossref Search ADS PubMed WorldCat 56. Fries P . Rhythms for cognition: communication through coherence . Neuron. 2015 ; 88 ( 1 ): 220 – 235 . Google Scholar Crossref Search ADS PubMed WorldCat 57. Clemens Z , et al. Temporal coupling of parahippocampal ripples, sleep spindles and slow oscillations in humans . Brain. 2007 ; 130 ( Pt 11 ): 2868 – 2878 . Google Scholar Crossref Search ADS PubMed WorldCat 58. Maingret N , et al. Hippocampo-cortical coupling mediates memory consolidation during sleep . Nat Neurosci. 2016 ; 19 ( 7 ): 959 – 964 . Google Scholar Crossref Search ADS PubMed WorldCat 59. Laventure S , et al. Beyond spindles: interactions between sleep spindles and boundary frequencies during cued reactivation of motor memory representations . Sleep. 2018 ; 41 ( 9 ). OpenURL Placeholder Text WorldCat 60. Siapas AG , et al. Coordinated interactions between hippocampal ripples and cortical spindles during slow-wave sleep . Neuron. 1998 ; 21 ( 5 ): 1123 – 1128 . Google Scholar Crossref Search ADS PubMed WorldCat 61. Sirota A , et al. Communication between neocortex and hippocampus during sleep in rodents . Proc Natl Acad Sci USA. 2003 ; 100 ( 4 ): 2065 – 2069 . Google Scholar Crossref Search ADS WorldCat 62. Chrobak JJ , et al. High-frequency oscillations in the output networks of the hippocampal-entorhinal axis of the freely behaving rat . J Neurosci. 1996 ; 16 ( 9 ): 3056 – 3066 . Google Scholar Crossref Search ADS PubMed WorldCat 63. Buzsáki G , et al. Cellular bases of hippocampal EEG in the behaving rat . Brain Res. 1983 ; 287 ( 2 ): 139 – 171 . Google Scholar Crossref Search ADS PubMed WorldCat 64. Girardeau G , et al. Selective suppression of hippocampal ripples impairs spatial memory . Nat Neurosci. 2009 ; 12 ( 10 ): 1222 – 1223 . Google Scholar Crossref Search ADS PubMed WorldCat 65. Axmacher N , et al. Ripples in the medial temporal lobe are relevant for human memory consolidation . Brain. 2008 ; 131 ( Pt 7 ): 1806 – 1817 . Google Scholar Crossref Search ADS PubMed WorldCat 66. Staresina BP , et al. Hierarchical nesting of slow oscillations, spindles and ripples in the human hippocampus during sleep . Nat Neurosci. 2015 ; 18 ( 11 ): 1679 – 1686 . Google Scholar Crossref Search ADS PubMed WorldCat 67. Latchoumane CV , et al. Thalamic spindles promote memory formation during sleep through triple phase-locking of cortical, Thalamic, and Hippocampal rhythms . Neuron. 2017 ; 95 ( 2 ): 424 – 435.e6 . Google Scholar Crossref Search ADS PubMed WorldCat 68. Takeuchi S , et al. Spatiotemporal organization and cross-frequency coupling of sleep spindles in primate cerebral cortex . Sleep. 2016 ; 39 ( 9 ): 1719 – 1735 . Google Scholar Crossref Search ADS PubMed WorldCat 69. Vyazovskiy VV , et al. Local sleep in awake rats . Nature. 2011 ; 472 ( 7344 ): 443 – 447 . Google Scholar Crossref Search ADS PubMed WorldCat 70. Funk CM , et al. Local slow waves in superficial layers of primary cortical areas during REM sleep . Curr Biol. 2016 ; 26 ( 3 ): 396 – 403 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes Co-last authors. © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please email: [email protected] This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Assadi, Mohammad, H;Segev,, Yael;Tarasiuk,, Ariel

doi: 10.1093/sleep/zsz176pmid: 31353408

Abstract Study Objectives Treatment of obstructive sleep apnea increases obesity risk by an unclear mechanism. Here, we explored the effects of upper airway obstruction and its removal on respiratory homeostasis, energy expenditure, and feeding hormones during the sleep/wake cycle from weaning to adulthood. Methods The tracheas of 22-day-old rats were narrowed, and obstruction removal was performed on post-surgery day 14. Energy expenditure, ventilation, and hormone-regulated feeding were analyzed during 49 days before and after obstruction. Results Energy expenditure increased and body temperature decreased in upper airway obstruction and was only partially recovered in obstruction removal despite normalization of airway resistance. Increased energy expenditure was associated with upregulation of ventilation. Decreased body temperature was associated with decreased brown adipose tissue uncoupling protein 1 level, suppressed energy expenditure response to norepinephrine, and decreased leptin level. Upper airway obstructed animals added less body weight, in spite of an increase in food intake, due to elevated hypothalamic orexin and neuropeptide Y and plasma ghrelin. Animals who underwent obstruction removal fed more due to an increase in hypothalamic neuropeptide Y and plasma ghrelin. Conclusions The need to maintain respiratory homeostasis is associated with persistent abnormal energy metabolism and hormonal regulation of feeding. Surgical treatment per se may not be sufficient to correct energy homeostasis, and endocrine regulation of feeding may have a larger effect on weight change. pediatric OSA, energy expenditure, thermogenesis, feeding hormones Statement of Significance The mechanisms linking obstructive sleep apnea and its treatment with whole-body energy balance and hormonal regulation of feeding are not well understood. Here, we used an integrative approach to explore the effects of upper airway obstruction and its removal on respiratory homeostasis, whole-body energy balance, and hormone-regulated feeding in a rat model. Our evidence indicates that upper airway obstruction causes persistent elevation of energy expenditure and feeding hormones, which last long after airway obstruction removal and normalization of airway resistance. Additional therapeutic measures to restore adequate neurohumoral control of feeding are essential in order to prevent adverse weight changes following surgery. Introduction Pediatric obstructive sleep apnea (OSA) is a syndrome of upper airway obstruction (AO) and fragmented sleep, which leads to either growth retardation or metabolic syndrome [1, 2]. Treatment by adenotonsillectomy in children or positive airway pressure in adults predisposes accelerated weight gain in children [2, 3] and adults [4–6]. Few studies have explored the effects of OSA treatment on whole-body energy balance and hormonal regulation of feeding, and the mechanisms underlying their derangements have not been well characterized [3–7]. Regulation of energy balance depends on factors such as the metabolic rate, physical activity, hormonal regulation of feeding, thermic effect of food intake, and eating behavior [2, 4, 5, 8–10]. OSA may elevate energy expenditure by increasing sympathetic activity and/or increased work of breathing [4, 5, 7]. The juvenile upper AO in rats by trachea narrowing mimics many of the clinical characteristics of childhood OSA including sleep fragmentation, growth retardation, and increased feeding [11–13]. The sympathetic nervous system activation stimulates brown adipose tissue by upregulating uncoupling protein 1 expression in order to dissipate chemical energy to body heat [14, 15], and leptin plays an important role in this process [16]. Intracerebroventricular administration of orexin-A has been shown to increase the sympathetic activity, which causes brown adipose tissue activation via β-adrenergic receptors, leading to an increase in body temperature (Tb) [17]. This response was abolished in leptin-deficient mice [16]. We previously reported Tb decreases in AO animals despite increased energy intake [12, 13, 18] although the mechanisms of this hypothermia remain elusive. Here, we hypothesized that the need to maintain respiratory homeostasis in AO leads to an elevation of energy expenditure, increased feeding behavior, and loss of thermoregulation, and that these metabolic AO consequences persist following successful obstruction removal (OR). We used an integrative approach to explore the effects of upper AO and its removal on respiratory homeostasis, whole-body energy balance, and hormone-regulated feeding. Methods Upper airway obstruction Rats were anesthetized using tribromoethanol (200 mg/kg i.p). Upper AO or sham control surgery was performed in 22-day-old male Sprague-Dawley rats [11, 18–20]. On day 14, OR surgery was performed on half the AO animals [13]. During the surgical procedures, the mortality rates in the AO and OR group were less than 10%, and an additional 5% mortality was observed 2–5 days after surgery. Figure 1 illustrates the experimental design and the time-line of data collection. Animals were kept on a 12–12 light-dark cycle with lights on at 09:00 at 23±1.0°C. Food (3272 kcal/kg) and water were given ad libitum. Figure 1. Open in new tabDownload slide Flow diagram of study groups and time data. Figure 1. Open in new tabDownload slide Flow diagram of study groups and time data. Experimental schedule AO or sham surgery was performed on 22-day-old rats (day = 0) and animals were returned to their home cages (Figure 1). On day 14, the AO group was randomized; OR of the silicon band was performed on half of the animals. Measurements of respiration (and arterial blood gases) were performed on day 2, 13, and 45 after surgery by whole-body plethysmograph. A telemetric transmitter was implanted for recording body Tb; locomotion activity (MA) and data were continuously collected on day 45 for 24 hours. Metabolic and activity profiles on conscious rats’ data were performed on day 46–48. On day 49, animals were sacrificed and tissue was extracted after death. Respiratory and energy metabolism Trachea diameter was measured by magnetic resonance imaging (MRI), as previously described [21]. Respiratory activity at room air and following administration of 7% CO2 in 93% O2 was recorded by whole-body plethysmography (Buxco, DSI, St. Paul, MN). Inspiratory swings in esophageal pressure (∆Pes) were measured in anesthetized animals [11, 12, 18, 20]. Esophageal pressure was measured from a fine saline-filled catheter placed in the lower third of the esophagus and connected to a pressure transducer. Ten breaths were used to analyze esophageal pressure. Inspiratory flow was determined from whole-body plethysmography. Airway resistance was calculated as ∆Pes/flow. Metabolic and activity profiles were measured using the high-definition behavioral phenotyping system (Sable Instruments, Las Vegas, NV). Respiratory gases were measured by a pull-mode negative-pressure system. Airflow was measured and controlled, water vapor was continuously measured, and its dilution effect on O2 and CO2 was calculated. Animals were allowed a 24-hour acclimation period followed by a 48-hour sampling duration. Effective body mass was calculated by analysis of covariance (ANCOVA) analysis (Supplementary Figure S1) [22]. Energy expenditure was calculated as VO2 × (3.815 + 1.232 × respiratory quotient), and was normalized to effective body mass. The respiratory quotient was calculated as the ratio of CO2 produced by O2 consumed by the body. Temperature-dependent changes in metabolism were performed by a modified Scholander test [14, 23]. Brown adipose tissue recruitment was performed following norepinephrine injection (1 mg/kg) in pentobarbital-anesthetized (45 mg/kg) animals [14]. The interscapular and tail surface temperatures were measured using an infrared camera [14, 16]. Tb and locomotion activity (MA) were analyzed by a telemetry system [11, 18]. Arterial blood gases were determined under anesthesia at day 2, 14, and 48 after surgery. Trachea histology was performed, as previously described [13]. Protein analysis Plasma ghrelin and leptin protein levels, and hypothalamic orexin and neuropeptide Y (NPY) levels were determined 2–3 hours after lights on using specific enzyme-linked immunosorbent assay kits [11–13]. Data analysis Significance was analyzed by an unpaired t-test. Two-way analysis of variance for repeated measures was used to determine significance between time and group, temperature and group, or norepinephrine and group for post hoc comparisons by Student–Newman–Keuls test. The null hypotheses were rejected at the 5% level. Results During the observation period, behavior in both the AO and OR groups was similar to that of controls. Animals explored their cage and engaged in normal social activity, such as grooming. Both light and dark phase energy expenditure increased by 45% and 15% in the AO and OR groups, respectively (p < 0.001; Figure 2, A; Table 1; n = 22, 20, 21 for the control, AO, and OR, respectively). The elevation of energy expenditure was associated with increased O2 consumption and CO2 production (p < 0.001; Figure 2, B and C; Table 1) without any effect in respiratory quotient. Locomotion activity on day 45 during the dark phase decreased by 20% in the AO group and increased by 30% in the OR group (p < 0.001; Figure 2, E). Increased CO2 production in AO was associated with 35%, 68%, and 260% elevation of ventilation during room air breathing at day 2, 13, and 45, respectively (p < 0.01; Figure 3, A–C; Table 2; n = 9 for all groups). CO2 sensitivity in the AO group decreased by 46% and 72% at day 2 and 45, respectively (p < 0.01; Figure 3, D; Table 2). Supplementary Figure S2 in the summarizes the effects of AO and OR on respiratory rate and tidal volume. Interestingly, the OR group had a 14.7% higher resting energy expenditure (Figure 2, D) despite restoration of trachea diameter (Figure 1, G and H; n = 8 for all groups) and airway resistance (Table 2; n = 11, 10, and 10 for the control, AO, and OR, respectively). Minute ventilation remained 40% higher in the OR group (p < 0.01; Figure 2, C). During the observation period, all groups maintained their normal arterial blood gases (Supplementary Table S1). Table 1. Energy metabolism . Control . AO . OR . O2 consumption (mL/min/kg) 18.43 ± 0.32 25.41 ± 0.38* 21.08 ± 0.39* ,# CO2 production (mL/min/kg) 16.63 ± 0.3 22.51 ± 0.4* 18.88 ± 0.36* ,# EE (kcal/h/kg) 5.46 ± 0.09 7.5 ± 0.11* 6.23 ± 0.11* ,# R_EE (kcal/h/kg) 3.99 ± 0.08 5.49 ± 0.09* 4.58 ± 0.08* ,# Body weight (g) 364 ± 5.8 254 ± 14.1* 316 ± 10.6* ,# Food intake (g/kg) 55.22 ± 1.6 73.37 ± 4.0* 62.87 ± 1.97* ,# Water intake (mL/kg) 32.34 ± 2.5 45.89 ± 5.3* 32.97 ± 3.8 . Control . AO . OR . O2 consumption (mL/min/kg) 18.43 ± 0.32 25.41 ± 0.38* 21.08 ± 0.39* ,# CO2 production (mL/min/kg) 16.63 ± 0.3 22.51 ± 0.4* 18.88 ± 0.36* ,# EE (kcal/h/kg) 5.46 ± 0.09 7.5 ± 0.11* 6.23 ± 0.11* ,# R_EE (kcal/h/kg) 3.99 ± 0.08 5.49 ± 0.09* 4.58 ± 0.08* ,# Body weight (g) 364 ± 5.8 254 ± 14.1* 316 ± 10.6* ,# Food intake (g/kg) 55.22 ± 1.6 73.37 ± 4.0* 62.87 ± 1.97* ,# Water intake (mL/kg) 32.34 ± 2.5 45.89 ± 5.3* 32.97 ± 3.8 Mean values are over 24 hours. Body weight was corrected for effective body mass. Values are mean ± SEM. Control (n = 22); AO, obstructive (n = 20); OR, obstruction removal (n = 21). EE, energy expenditure; R_EE, resting energy expenditure. *p < 0.01, C vs. AO or OR. #p < 0.05, AO vs. OR. Open in new tab Table 1. Energy metabolism . Control . AO . OR . O2 consumption (mL/min/kg) 18.43 ± 0.32 25.41 ± 0.38* 21.08 ± 0.39* ,# CO2 production (mL/min/kg) 16.63 ± 0.3 22.51 ± 0.4* 18.88 ± 0.36* ,# EE (kcal/h/kg) 5.46 ± 0.09 7.5 ± 0.11* 6.23 ± 0.11* ,# R_EE (kcal/h/kg) 3.99 ± 0.08 5.49 ± 0.09* 4.58 ± 0.08* ,# Body weight (g) 364 ± 5.8 254 ± 14.1* 316 ± 10.6* ,# Food intake (g/kg) 55.22 ± 1.6 73.37 ± 4.0* 62.87 ± 1.97* ,# Water intake (mL/kg) 32.34 ± 2.5 45.89 ± 5.3* 32.97 ± 3.8 . Control . AO . OR . O2 consumption (mL/min/kg) 18.43 ± 0.32 25.41 ± 0.38* 21.08 ± 0.39* ,# CO2 production (mL/min/kg) 16.63 ± 0.3 22.51 ± 0.4* 18.88 ± 0.36* ,# EE (kcal/h/kg) 5.46 ± 0.09 7.5 ± 0.11* 6.23 ± 0.11* ,# R_EE (kcal/h/kg) 3.99 ± 0.08 5.49 ± 0.09* 4.58 ± 0.08* ,# Body weight (g) 364 ± 5.8 254 ± 14.1* 316 ± 10.6* ,# Food intake (g/kg) 55.22 ± 1.6 73.37 ± 4.0* 62.87 ± 1.97* ,# Water intake (mL/kg) 32.34 ± 2.5 45.89 ± 5.3* 32.97 ± 3.8 Mean values are over 24 hours. Body weight was corrected for effective body mass. Values are mean ± SEM. Control (n = 22); AO, obstructive (n = 20); OR, obstruction removal (n = 21). EE, energy expenditure; R_EE, resting energy expenditure. *p < 0.01, C vs. AO or OR. #p < 0.05, AO vs. OR. Open in new tab Table 2. Respiratory data . Group . Days after surgery . 2 . 13 . 49 . ∆Pes (cmH2O) Control −8.8 ± 0.8 −10.8 ± 1.9 AO −18.2 ± 3.2* −20.2 ± 7.6* OR −10.8 ± 2.1 Raw (cmH2O/L/minute) Control 22.2 ± 1.9 20.7 ± 3.1 AO 56.9 ± 3.9* 45.3 ± 12.7* OR 22.9 ± 3.7 Room air ventilation (mL/minute/100 g) Control 141 ± 7.2 95 ± 4.4 51 ± 3.5 AO 192 ± 22* 160 ± 23* 141 ± 19* OR 72 ± 6.7* 7% CO2 ventilation (mL/minute/100 g) Control 411 ± 47+ 244 ± 9.4+ 135 ± 7.9+ AO 365 ± 27+ 314 ± 18* ,+ 154 ± 7.2* OR 172 ± 10.2* ,+ CO2 response (%) Control 193 ± 16.3 160 ± 13.6 163 ± 11.5 AO 104 ± 15.5* 139 ± 27 45.5 ± 22.5* OR 146.5 ± 14.8 . Group . Days after surgery . 2 . 13 . 49 . ∆Pes (cmH2O) Control −8.8 ± 0.8 −10.8 ± 1.9 AO −18.2 ± 3.2* −20.2 ± 7.6* OR −10.8 ± 2.1 Raw (cmH2O/L/minute) Control 22.2 ± 1.9 20.7 ± 3.1 AO 56.9 ± 3.9* 45.3 ± 12.7* OR 22.9 ± 3.7 Room air ventilation (mL/minute/100 g) Control 141 ± 7.2 95 ± 4.4 51 ± 3.5 AO 192 ± 22* 160 ± 23* 141 ± 19* OR 72 ± 6.7* 7% CO2 ventilation (mL/minute/100 g) Control 411 ± 47+ 244 ± 9.4+ 135 ± 7.9+ AO 365 ± 27+ 314 ± 18* ,+ 154 ± 7.2* OR 172 ± 10.2* ,+ CO2 response (%) Control 193 ± 16.3 160 ± 13.6 163 ± 11.5 AO 104 ± 15.5* 139 ± 27 45.5 ± 22.5* OR 146.5 ± 14.8 Values are mean ± SEM. AO, obstructive; OR, obstruction removal; ∆Pes, inspiratory swings in esophageal pressure; Raw, airway resistance. *p < 0.01, C vs. AO or OR. +p < 0.01, room air vs. CO2. Open in new tab Table 2. Respiratory data . Group . Days after surgery . 2 . 13 . 49 . ∆Pes (cmH2O) Control −8.8 ± 0.8 −10.8 ± 1.9 AO −18.2 ± 3.2* −20.2 ± 7.6* OR −10.8 ± 2.1 Raw (cmH2O/L/minute) Control 22.2 ± 1.9 20.7 ± 3.1 AO 56.9 ± 3.9* 45.3 ± 12.7* OR 22.9 ± 3.7 Room air ventilation (mL/minute/100 g) Control 141 ± 7.2 95 ± 4.4 51 ± 3.5 AO 192 ± 22* 160 ± 23* 141 ± 19* OR 72 ± 6.7* 7% CO2 ventilation (mL/minute/100 g) Control 411 ± 47+ 244 ± 9.4+ 135 ± 7.9+ AO 365 ± 27+ 314 ± 18* ,+ 154 ± 7.2* OR 172 ± 10.2* ,+ CO2 response (%) Control 193 ± 16.3 160 ± 13.6 163 ± 11.5 AO 104 ± 15.5* 139 ± 27 45.5 ± 22.5* OR 146.5 ± 14.8 . Group . Days after surgery . 2 . 13 . 49 . ∆Pes (cmH2O) Control −8.8 ± 0.8 −10.8 ± 1.9 AO −18.2 ± 3.2* −20.2 ± 7.6* OR −10.8 ± 2.1 Raw (cmH2O/L/minute) Control 22.2 ± 1.9 20.7 ± 3.1 AO 56.9 ± 3.9* 45.3 ± 12.7* OR 22.9 ± 3.7 Room air ventilation (mL/minute/100 g) Control 141 ± 7.2 95 ± 4.4 51 ± 3.5 AO 192 ± 22* 160 ± 23* 141 ± 19* OR 72 ± 6.7* 7% CO2 ventilation (mL/minute/100 g) Control 411 ± 47+ 244 ± 9.4+ 135 ± 7.9+ AO 365 ± 27+ 314 ± 18* ,+ 154 ± 7.2* OR 172 ± 10.2* ,+ CO2 response (%) Control 193 ± 16.3 160 ± 13.6 163 ± 11.5 AO 104 ± 15.5* 139 ± 27 45.5 ± 22.5* OR 146.5 ± 14.8 Values are mean ± SEM. AO, obstructive; OR, obstruction removal; ∆Pes, inspiratory swings in esophageal pressure; Raw, airway resistance. *p < 0.01, C vs. AO or OR. +p < 0.01, room air vs. CO2. Open in new tab Figure 2. Open in new tabDownload slide Energy metabolism analysis: (A) EE; (B) O2 consumption; (C) CO2 production; (D) resting EE; (E) MA; (F) trachea diameter; (G) representative image of trachea in control (C), obstruction (AO), and obstruction removal (OR) rats; (H) representative image of trachea 3D MRI. Gray area in A–C represents lights of phase (active period, 21:00–09:00) on a 12:12-hour cycle. Values in A–D were adjusted to effective body mass by ANCOVA analysis. EE, energy expenditure; R_EE, resting EE calculated as mean value for 30-minute period with lowest EE; MA, locomotion activity; MRI, magnetic resonance imaging. Blue: control (n = 22); green: obstructive (n = 20); red: obstruction removal (n = 21). Values are mean ± SEM. *p < 0.001, C vs. AO group. #p < 0.01, AO vs. OR. In A–C, statistical differences were determined by a two-way analysis of variance (ANOVA) followed by a post hoc Student–Newman–Keuls test. In D and F, differences were determined by a two-tailed t-test. ***p < 0.002. Figure 2. Open in new tabDownload slide Energy metabolism analysis: (A) EE; (B) O2 consumption; (C) CO2 production; (D) resting EE; (E) MA; (F) trachea diameter; (G) representative image of trachea in control (C), obstruction (AO), and obstruction removal (OR) rats; (H) representative image of trachea 3D MRI. Gray area in A–C represents lights of phase (active period, 21:00–09:00) on a 12:12-hour cycle. Values in A–D were adjusted to effective body mass by ANCOVA analysis. EE, energy expenditure; R_EE, resting EE calculated as mean value for 30-minute period with lowest EE; MA, locomotion activity; MRI, magnetic resonance imaging. Blue: control (n = 22); green: obstructive (n = 20); red: obstruction removal (n = 21). Values are mean ± SEM. *p < 0.001, C vs. AO group. #p < 0.01, AO vs. OR. In A–C, statistical differences were determined by a two-way analysis of variance (ANOVA) followed by a post hoc Student–Newman–Keuls test. In D and F, differences were determined by a two-tailed t-test. ***p < 0.002. Figure 3. Open in new tabDownload slide Respiratory parameters: (A) minute ventilation on day 2; (B) minute ventilation on day 13; (C) minute ventilation on day 45; (D) CO2 response calculated as the percent change in minute ventilation from room air breathing to 7% CO2 stimulation. RA, room air; blue: control; green: obstructive; red: obstruction removal. Values are mean ± SEM. n = 9 for all groups. *p < 0.01, C vs. AO. +p < 0.01, RA vs. 7% CO2. Differences were determined by a two-tailed t-test. Figure 3. Open in new tabDownload slide Respiratory parameters: (A) minute ventilation on day 2; (B) minute ventilation on day 13; (C) minute ventilation on day 45; (D) CO2 response calculated as the percent change in minute ventilation from room air breathing to 7% CO2 stimulation. RA, room air; blue: control; green: obstructive; red: obstruction removal. Values are mean ± SEM. n = 9 for all groups. *p < 0.01, C vs. AO. +p < 0.01, RA vs. 7% CO2. Differences were determined by a two-tailed t-test. Tb of AO decreased by 0.8°C (p < 0.01; Figure 4, A) compared to controls. We assessed the thermogenic capacity of brown adipose tissue using norepinephrine (Figure 4, B–D; n = 9 for all groups). Following administration of norepinephrine, energy expenditure increased significantly in the control and OR groups and did not change significantly in the AO group. The maximal energy expenditure response in OR was close to 70% of controls response (p < 0.01). Total upregulating uncoupling protein 1 level in AO decreased by 28% (p < 0.01; Figure 4, E). To measure temperature-dependent changes in metabolism, modified Scholander tests [27] were performed. Figure 4, F illustrates the average energy expenditure at each temperature. Controls displayed a typical thermoneutral temperature range of 23–27°C (p < 0.01). Changing ambient temperature above or below this temperature range caused a significant increase in energy expenditure (p < 0.01). Interestingly, the energy expenditure of AO animals was unaffected by changes in ambient temperature (Figure 4, F). The OR thermoneutrality range was similar to that of controls (Figure 4, F). Infrared thermography of skin surface temperature was measured at 27°C. The interscapular and tail surface temperatures of AO and OR rats were similar to those of controls (Figure 5, B and C; n = 9 for all groups), and similar temperature differences between interscapular and tail temperature were found in all groups (Figure 5, D). Figure 4. Open in new tabDownload slide Thermogenic capacity: (A) body temperature; (B–D) functional analysis of brown adipose tissue thermogenic capacity using the norepinephrine (NE) injection (1 mg/kg, arrow) test in animals anesthetized with pentobarbital (45 g/kg) at day 45; (B) control; (C) obstructive; (D) obstruction removal; (E) upregulating uncoupling protein 1 by western blot analysis of 100 μg brown fat samples; (F) Scholander plot. Gray area in A represents light phase (09:00–21:00) on a 12:12-hour cycle. Blue: control (C); green: obstructive (AO); red: obstruction removal (OR). n = 9 for all groups. *p < 0.01, C vs. AO. #p < 0.05, AO vs. OR. In A–D, F, statistical difference between groups was determined by a two-way analysis of variance (ANOVA), followed by a post hoc Student–Newman–Keuls test. In E, statistical differences were determined by an unpaired two-tailed t-test. Figure 4. Open in new tabDownload slide Thermogenic capacity: (A) body temperature; (B–D) functional analysis of brown adipose tissue thermogenic capacity using the norepinephrine (NE) injection (1 mg/kg, arrow) test in animals anesthetized with pentobarbital (45 g/kg) at day 45; (B) control; (C) obstructive; (D) obstruction removal; (E) upregulating uncoupling protein 1 by western blot analysis of 100 μg brown fat samples; (F) Scholander plot. Gray area in A represents light phase (09:00–21:00) on a 12:12-hour cycle. Blue: control (C); green: obstructive (AO); red: obstruction removal (OR). n = 9 for all groups. *p < 0.01, C vs. AO. #p < 0.05, AO vs. OR. In A–D, F, statistical difference between groups was determined by a two-way analysis of variance (ANOVA), followed by a post hoc Student–Newman–Keuls test. In E, statistical differences were determined by an unpaired two-tailed t-test. Figure 5. Open in new tabDownload slide Thermography analysis: (A) representative infrared images; (B) interscapular temperature; (C) tail temperature measured 0.5 cm from the tail base; (D) the differences between interscapular and tail temperatures; blue: control (C); green: obstructive (AO); red: obstruction removal (OR). n = 9 for all groups. Figure 5. Open in new tabDownload slide Thermography analysis: (A) representative infrared images; (B) interscapular temperature; (C) tail temperature measured 0.5 cm from the tail base; (D) the differences between interscapular and tail temperatures; blue: control (C); green: obstructive (AO); red: obstruction removal (OR). n = 9 for all groups. Over the observation period, body weight gain was 30% and 13% less than control rats in the AO and OR groups, respectively (p < 0.01; Figure 6, A). Decreased body weight in AO was associated with a large decrease of retroperitoneal adiposity volume (Supplementary Figure S3A), cell diameter by 29% (p <0.01; Figure 6, B and C; n = 20 for all groups), and adipocyte distribution (Supplementary Figure S3D). Food intake was 33% and 14% higher in AO and OR groups, respectively (p < 0.01; Figure 6, B), and water intake during the light phase was 80% higher in the AO group (p < 0.01; Figure 6, C). Total body energy balance (i.e. energy consumption vs. energy expenditure; Figure 6, D) was 49% less (p < 0.01) in the AO group. Increased feeding was associated with upregulation of hypothalamic orexigenic activity. This activity was 20% higher in the AO group (p < 0.01; Figure 6, E; n = 8 for all groups). Hypothalamic NPY was 16% and 13% higher in the AO and OR groups, respectively (p < 0.01; Figure 6, F; n = 9 for all groups). Plasma ghrelin increased by 38% and 35% in the AO and OR groups, respectively (p < 0.01; Figure 6, G; n = 10 for all groups). Plasma leptin decreased by 50% in the AO group (p < 0.01; Figure 6, H). Figure 6. Open in new tabDownload slide Energy intake and hormones: (A) body weight; (B) food intake; (C) water intake; (D) total body energy balance (energy consumption vs. energy expenditure); (E) hypothalamic orexin protein level; (F) hypothalamic NPY protein level; (G) serum ghrelin level; (H) serum leptin level. Specific ELISA determined protein levels. Blue: control; green: obstructive; red: obstruction removal; values in B and C were adjusted to effective body mass calculated by ANCOVA analysis. Values are mean ± SEM. *p < 0.01, C vs. AO. #p < 0.05, AO vs. OR. In A, the statistical difference between groups was determined by a two-way analysis of variance (ANOVA), followed by a post hoc Student–Newman–Keuls test. In B–I, statistical differences were determined by an unpaired two-tailed t-test. Figure 6. Open in new tabDownload slide Energy intake and hormones: (A) body weight; (B) food intake; (C) water intake; (D) total body energy balance (energy consumption vs. energy expenditure); (E) hypothalamic orexin protein level; (F) hypothalamic NPY protein level; (G) serum ghrelin level; (H) serum leptin level. Specific ELISA determined protein levels. Blue: control; green: obstructive; red: obstruction removal; values in B and C were adjusted to effective body mass calculated by ANCOVA analysis. Values are mean ± SEM. *p < 0.01, C vs. AO. #p < 0.05, AO vs. OR. In A, the statistical difference between groups was determined by a two-way analysis of variance (ANOVA), followed by a post hoc Student–Newman–Keuls test. In B–I, statistical differences were determined by an unpaired two-tailed t-test. Discussion Using the rat model, we found that AO causes an increase in energy expenditure that was only partially recovered following the OR, despite near complete normalization of the trachea diameter. Increased energy expenditure was associated with increased CO2 production and with upregulation of ventilation. In parallel, the AO animals’ Tb was significantly lower than in the control, most probably due to decreased expression of brown adipose tissue uncoupling protein 1, lower responsiveness to norepinephrine, and decreased leptin level. Our evidence indicates that the AO animals gained less body weight, in spite of an increase in food intake, due to elevated hypothalamic orexinergic activity (orexin and NPY) and plasma ghrelin. This effect was only partially reversed upon OR. Several animal models are available to explore in-depth the effects of OSA on regulation of upper airway physiology, sleep, energy metabolism, and functional consequences [24, 25]. To the best of our knowledge, the effects of AO and its removal on respiratory and energy metabolism maintenance have not been explored. Our model has implications for OSA; similar to sleep apnea, AO animals exhibit abnormal sleep and growth retardation [25]. However, in this model, we introduced both inspiratory and expiratory upper airway loading, which may be representative of OSA patients with persistently high upper airway resistance that is not exclusively sleep related, as with patients with upper airway anatomic abnormalities, i.e. increased nasal resistance, subglottic or tracheal stenosis, retrognathia, or macroglossia. In OSA, upper AO is intermittent and present only during sleep with opportunity for recovery during the day [1], and it is continuously present in our model. Thus, there may be physiological adaptations to chronic and persistent AO in this model that do not occur in OSA. Our model mimics one of the defining features of OSA, i.e. mechanical loading that is not coupled with hypoxia [20, 25]. In this study, sleep was not measured, and it is possible that ventilation and arterial PO2 will decrease during sleep. Previously, we measured arterial blood gases in unrestrained animals and found normal arterial blood gases during the sleep period [21]. We found a considerable increase in energy expenditure during both light and dark phases in AO animals. This elevation could be related to the increased work of breathing. In our study, trachea diameter was reduced by 50%, and airway resistance increased by 217%. Earlier studies found tracheal resistance increased by 46%–350% [11, 18, 19], and increased diaphragmatic force [20]. Estimates of metabolic rate from its gaseous content (i.e. exchange of O2 for CO2) with calorimetry are valid only when metabolic energy is derived from aerobic sources [14, 22]. The energetic cost of breathing is considerably greater during resistive breathing, relative to isocapnic hyperventilation. The effect of OSA on the resting metabolic rate in adults was investigated the morning following an overnight fast in cross-sectional studies. Resting metabolic rate was significantly higher in OSA patients following adjustment for age and body mass index [7, 26, 27]. OSA may lead to increased 24-hour energy expenditure; however, results are less consistent. A significant positive association was found between resting metabolic rate and OSA severity [28, 29]. As far as we are aware, there is little information available on the effect of upper AO on energy expenditure in children and juvenile animals. This study is the first to show the effects of AO on energy expenditure from weaning to adulthood. In our study, respiratory loading was not sleep related and this may explain the higher energy expenditure found in AO. In OSA, on the other hand, the upper AO is mostly inspiratory and sleep related. Nevertheless, similarities of this model to OSA are remarkable [29]. Moreover, partial sleep loss [11, 12, 30] may also contribute to elevated energy expenditure in AO. Energy expenditure increases in sleep-deprived rats [30] and during insufficient sleep in humans [8]. Decreased sleep in rodents regardless of the methodology, including AO in rats, leads to hypothermia despite increased energy expenditure [31, 32]. Brown adipose tissue is responsible for heat generation via uncoupling of electron transport from ATP synthesis by upregulating uncoupling protein 1 [14, 16]. We assessed the thermogenic capacity of brown adipose tissue using norepinephrine [16]. An impressive decrease of brown adipose tissue’s thermogenic capacity in AO and almost complete recovery in OR was observed. Moreover, leptin is an adipokine that plays a key role in regulation of both energy homeostasis and thermoregulation by stimulating the sympathetic nervous system to brown adipose tissue. Thermogenesis of brown adipose tissue in ob/ob (−/−) mice that do not have functional leptin is abolished compared with the wild type [16]. The decreased Tb in AO and its normalization in OR was associated with decreased uncoupling protein 1 and leptin levels and normalization of their levels in OR. We performed modified Scholander experiments [23] to measure temperature-dependent effects on energy expenditure [18]. AO animals displayed high-energy expenditure at all ambient temperatures tested and did not exhibit a thermoneutral range. This finding supports the possibility that AO leads to non-functioning brown adipose tissue. Finally, we used infrared thermography to measure changes in tail surface and interscapular temperature differences to explore changes in heat loss [14]. The tail of rodents has been used in the study of peripheral vasomotor control of thermoregulation [14, 33]. No changes in temperature differences were found, and heat loss cannot explain the hypothermia seen in AO. Collectively, our findings support the possibility that AO-induced hypothermia is related to non-functioning brown adipose tissue due to the decreased source of fuel available to generate body heat and decreased leptin. Interestingly, energy expenditure was higher in the OR group, despite restoration of airway resistance. Further studies are needed to explore the possibility of functional abnormalities in upper airway patency [12] during sleep-wake cycle that may contribute to increased energy expenditure following OR. Our study also suggests that increased energy expenditure could be related to the energetic costs of increased food intake and increased locomotion activity. Respiratory activity In the current study, trachea narrowing was associated with changes in ventilation. The increased ventilation by 260% in AO at day 45 was a physiological response to increased CO2 production. The decreased response to CO2 at day 2 could be related to enhancement of endogenous opioids [19, 34]. As we reported previously [19], respiratory rate is significantly decreased at day 2 following the AO. This decrease probably reflects a reflectory increase in respiratory duty cycle in order to minimize work of breathing. Chronic ventilatory loads induce an adaptive hypercapnic response that ultimately diminishes ventilatory sensitivity to CO2. The depression of the respiratory response to CO2 accompanies many diseases that impose chronic loads on the ventilatory system. It is unlikely that alterations in diaphragmatic contractility and/or changes in pulmonary mechanics are responsible for the suppression of CO2 sensitivity. The expiratory resistance produced by this model may lead to hyperinflation with an adverse effect on diaphragm function, which could account for some of the ventilatory findings. The chronic increase in functional residual volume in rats may lead to an adaptive decrease in optimal diaphragm length, diaphragmatic mass, and contractility [20]. We found, however, that diaphragm mass and contractility increased after 6 weeks of AO with no significant change in optimal length or functional residual capacity [20], whereas other investigators found increased endurance after 24 weeks of loading [35]. Although slow body weight gain was observed in AO, diaphragmatic atrophy that may contribute to ventilatory dysfunction is unlikely to have occurred in our obstructed animals. Nutritional factors may also contribute to ventilatory dysfunction [36, 37]. Although AO animals had lower body weight gain, they showed no evidence of malnutrition: serum protein level and tissue (liver, soleus muscle, and diaphragm) compositions for water, protein, and fat were similar to those of controls [12]. Inadequate sleep can lead to adverse health outcomes including increased appetite hormones (orexin, ghrelin) and food intake. Increased feeding in AO was associated with upregulation of hypothalamic orexin [11, 12, 21] that directly interacts with NPY neurons [38] and by elevation of circulating ghrelin that is activated via AMP-activated protein kinase hypothalamic NPY and other feeding mediators [39]. Although AO animals increased their caloric intake, the slow body weight gain strongly indicates the high metabolic cost of AO. It is possible that AO animals could not consume enough calories to meet additional energy requirements to maintain energy metabolism homeostasis. In humans, although adenotonsillectomy and positive airway pressure restore normal breathing and sleep, these treatments can lead to increased body weight gain and obesity by an unclear mechanism [2, 3, 10]. It is possible that sedentary lifestyles and selection of high calorie foods may have greater impact on weight change post-treatment in humans. Our findings also suggest that increased feeding behavior, at least in animals, is also associated with a persistent increase in feeding homeostasis long after the removal of the upper AO. Conclusion The upper airway model provides mechanistic insight concerning functional interactions between breathing, energy expenditure, and hormonal regulation of feeding. The need to maintain respiratory homeostasis is associated with persistent abnormal energy metabolism and hormonal regulation of feeding. Surgical intervention alone may not be sufficient to correct energy homeostasis, and endocrine regulation of feeding may have greater impacts on weight change. Funding This research was supported by the Israel Science Foundation, grant no. 164/2018. Acknowledgments M.H.A. conceived, designed, and performed the experiments and analyzed the data; Y.S., PhD, principal investigator, acquired molecular endocrinology data and analyzed it, recruited funds, and helped write the manuscript. A.T., PhD, principal investigator, recruited funds, analyzed the data, and helped write the manuscript. A.T. and Y.S. co-wrote the final version of the manuscript. Conflict of interest statement. None declared. References 1. Marcus CL , et al. ; American Academy of Pediatrics . Diagnosis and management of childhood obstructive sleep apnea syndrome . Pediatrics. 2012 ; 130 ( 3 ): e714 – e755 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Katz ES , et al. Growth after adenotonsillectomy for obstructive sleep apnea: an RCT . Pediatrics. 2014 ; 134 ( 2 ): 282 – 289 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Amin R , et al. Growth velocity predicts recurrence of sleep-disordered breathing 1 year after adenotonsillectomy . Am J Respir Crit Care Med . 2008 ; 177 ( 6 ): 654 – 659 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Tachikawa R , et al. Changes in energy metabolism after continuous positive airway pressure for obstructive sleep apnea . Am J Respir Crit Care Med . 2016 ; 194 ( 6 ): 729 – 738 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Shechter A . Obstructive sleep apnea and energy balance regulation: a systematic review . Sleep Med Rev . 2017 ; 34 : 59 – 69 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Drager LF , et al. Effects of CPAP on body weight in patients with obstructive sleep apnoea: a meta-analysis of randomised trials . Thorax . 2015 ; 70 ( 3 ): 258 – 264 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Stenlöf K , et al. Energy expenditure in obstructive sleep apnea: effects of treatment with continuous positive airway pressure . Am J Physiol . 1996 ; 271 ( 6 Pt 1 ): E1036 – E1043 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 8. Markwald RR , et al. Impact of insufficient sleep on total daily energy expenditure, food intake, and weight gain . Proc Natl Acad Sci USA . 2013 ; 110 ( 14 ): 5695 – 5700 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Hill JO , et al. Energy balance and obesity . Circulation . 2012 ; 126 ( 1 ): 126 – 132 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Beebe DW , et al. Dietary intake following experimentally restricted sleep in adolescents . Sleep . 2013 ; 36 ( 6 ): 827 – 834 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Tarasiuk A , et al. Role of growth hormone-releasing hormone in sleep and growth impairments induced by upper airway obstruction in rats . Eur Respir J . 2011 ; 38 ( 4 ): 870 – 877 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Tarasiuk A , et al. Role of orexin in respiratory and sleep homeostasis during upper airway obstruction in rats . Sleep . 2014 ; 37 ( 5 ): 987 – 998 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Assadi MH , et al. Abnormal growth and feeding behavior persist after removal of upper airway obstruction in juvenile rats . Sci Rep. 2017 ; 7 ( 1 ): 2730 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Fischer AW , et al. Leptin raises defended body temperature without activating thermogenesis . Cell Rep . 2016 ; 14 ( 7 ): 1621 – 1631 . Google Scholar Crossref Search ADS PubMed WorldCat 15. Cannon B , et al. Brown adipose tissue: function and physiological significance . Physiol Rev . 2004 ; 84 ( 1 ): 277 – 359 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Martins FF , et al. Thermogenesis, fatty acid synthesis with oxidation, and inflammation in the brown adipose tissue of ob/ob (-/-) mice . Ann Anat . 2017 ; 210 : 44 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Messina G , et al. Orexin-A controls sympathetic activity and eating behavior . Front Psychol . 2014 ; 5 : 997 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Segev Y , et al. Upper airway loading induces growth retardation and change in local chondrocyte IGF-I expression is reversed by stimulation of GH release in juvenile rats . J Appl Physiol (1985) . 2008 ; 105 ( 5 ): 1602 – 1609 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Greenberg HE , et al. Effect of chronic resistive loading on ventilatory control in a rat model . Am J Respir Crit Care Med . 1995 ; 152 ( 2 ): 666 – 676 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Tarasiuk A , et al. Effect of chronic resistive loading on inspiratory muscles in rats . J Appl Physiol (1985) . 1991 ; 70 ( 1 ): 216 – 222 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Tarasiuk A , et al. Orexin plays a role in growth impediment induced by obstructive sleep breathing in rats . Sleep . 2016 ; 39 ( 4 ): 887 – 897 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Tschöp MH , et al. A guide to analysis of mouse energy metabolism . Nat Methods . 2011 ; 9 ( 1 ): 57 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Scholander PF , et al. Heat regulation in some arctic and tropical mammals and birds . Biol Bull . 1950 ; 99 ( 2 ): 237 – 258 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Davis EM , et al. Rodent models of sleep apnea . Respir Physiol Neurobiol . 2013 ; 188 ( 3 ): 355 – 361 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Tarasiuk A , et al. Abnormal growth and feeding behavior in upper airway obstruction in rats . Front Endocrinol (Lausanne). 2018 ; 9 : 298 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Fekete K , et al. Resting energy expenditure in OSAS: the impact of a single CPAP application . Sleep Breath . 2016 ; 20 ( 1 ): 121 – 128 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Ryan CF , et al. Energy expenditure in obstructive sleep apnea . Sleep . 1995 ; 18 ( 3 ): 180 – 187 . Google Scholar Crossref Search ADS PubMed WorldCat 28. de Jonge L , et al. ; NIDDK Sleep Extension Study Group . Poor sleep quality and sleep apnea are associated with higher resting energy expenditure in obese individuals with short sleep duration . J Clin Endocrinol Metab . 2012 ; 97 ( 8 ): 2881 – 2889 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Kezirian EJ , et al. Resting energy expenditure in adults with sleep disordered breathing . Arch Otolaryngol Head Neck Surg . 2008 ; 134 ( 12 ): 1270 – 1275 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Gradwohl G , et al. Chronic upper airway obstruction induces abnormal sleep/wake dynamics in juvenile rats . PLoS One . 2014 ; 9 ( 5 ): e97111 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Bergmann BM , et al. Sleep deprivation in the rat: V. Energy use and mediation . Sleep . 1989 ; 12 ( 1 ): 31 – 41 . Google Scholar Crossref Search ADS PubMed WorldCat 32. Everson CA , et al. Recurrent restriction of sleep and inadequate recuperation induce both adaptive changes and pathological outcomes . Am J Physiol Regul Integr Comp Physiol . 2009 ; 297 ( 5 ): R1430 – R1440 . Google Scholar Crossref Search ADS PubMed WorldCat 33. Tattersall GJ . Infrared thermography: a non-invasive window into thermal physiology . Comp Biochem Physiol A Mol Integr Physiol . 2016 ; 202 : 78 – 98 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Scardella AT , et al. The role of endogenous opioids in the ventilatory response to acute flow-resistive loads . Am Rev Respir Dis . 1986 ; 133 ( 1 ): 26 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Prezant DJ , et al. Effects of long-term continuous respiratory resistive loading on rat diaphragm function and structure . J Appl Physiol (1985) . 1993 ; 74 ( 3 ): 1212 – 1219 . Google Scholar Crossref Search ADS PubMed WorldCat 36. Roberts BM , et al. Diaphragm and ventilatory dysfunction during cancer cachexia . FASEB J . 2013 ; 27 ( 7 ): 2600 – 2610 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Wheatley JR , et al. Influence of sleep on genioglossus muscle activation by negative pressure in normal men . Am Rev Respir Dis . 1993 ; 148 ( 3 ): 597 – 605 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Kola B . Role of AMP-activated protein kinase in the control of appetite . J Neuroendocrinol . 2008 ; 20 ( 7 ): 942 – 951 . Google Scholar Crossref Search ADS PubMed WorldCat 39. López M , et al. Estradiol effects on hypothalamic AMPK and BAT thermogenesis: a gateway for obesity treatment? Pharmacol Ther . 2017 ; 178 : 109 – 122 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes These authors contributed equally to the work. © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail [email protected]. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Dash, Michael, B

doi: 10.1093/sleep/zsz170pmid: 31353415

Abstract Slow wave activity (SWA; the EEG power between 0.5 and 4 Hz during non-rapid eye movement sleep [NREM]) is the best electrophysiological marker of sleep need; SWA dissipates across the night and increases following sleep deprivation. In addition to these well-documented homeostatic SWA trends, SWA exhibits extensive variability across shorter timescales (seconds to minutes) and between local cortical regions. The physiological underpinnings of SWA variability, however, remain poorly characterized. In male Sprague-Dawley rats, we observed that SWA exhibits pronounced infraslow fluctuations (~40- to 120-s periods) that are coordinated across disparate cortical locations. Peaks in SWA across infraslow cycles were associated with increased slope, amplitude, and duration of individual slow waves and a reduction in the total number of waves and proportion of multipeak waves. Using a freely available data set comprised of extracellular unit recordings during consolidated NREM episodes in male Long-Evans rats, we further show that infraslow SWA does not appear to arise as a consequence of firing rate modulation of putative excitatory or inhibitory neurons. Instead, infraslow SWA was associated with alterations in neuronal synchrony surrounding “On”/“Off” periods and changes in the number and duration of “Off” periods. Collectively, these data provide a mechanism by which SWA can be coordinated across disparate cortical locations and thereby connect local and global expression of this patterned neuronal activity. In doing so, infraslow SWA may contribute to the regulation of cortical circuits during sleep and thereby play a critical role in sleep function. slow wave activity, infraslow, sleep regulation, neuronal synchrony, rat Statement of Significance Patterns of neuronal communication during sleep differ dramatically from the waking brain and appear to be causally related to sleep function. For example, cortical slow wave activity (SWA) appears critical for sleep-associated memory improvements. We provide a novel characterization of how infraslow fluctuations (~40- to 120-s cycles) coordinate the expression of SWA across disparate cortical locations. We further demonstrate that changes in SWA across infraslow cycles arise from changes in neuronal synchrony, rather than changes in firing rates. Understanding how neuronal activity during sleep is regulated provides a foundation for understanding both healthy sleep function and sleep pathology. Introduction Cortical slow wave activity (SWA; the EEG power between 0.5 and 4 Hz) is the best established electrophysiological marker of sleep need [1], exhibiting a canonical homeostatic decline across periods of prolonged sleep. In addition to its well-characterized homeostatic decline, however, SWA also exhibits extensive spatial and temporal variability. Given that the expression of SWA is important for sleep function [2–4], it is critical to understand the physiological mechanisms responsible for the coordination and regulation of SWA. Individual slow waves originate within local cortical circuits and typically propagate through synaptic connections to generate a global slow wave [5, 6]. Propagation failures can result in the absence of slow waves within a local cortical circuit despite the presence of a slow wave elsewhere [7, 8], while the convergence of independently generated local slow waves may give rise to multipeak waves [9, 10]. Individual slow wave amplitudes vary considerably in conjunction with changes in the degree of neuronal synchronization [11]. Moreover, there appear to be distinct types of slow waves; local slow waves that are largely dependent upon corticocortical synchronization and global slow waves that may arise from subcorticocortical synchronization [12]. Each slow wave, therefore, represents a unique spatiotemporal sequence of neuronal activation and silence that depends upon both local and global regulators. Like individual slow waves, SWA also exhibits widespread variability. SWA decreases along the anterior-posterior axis [13, 14] and changes locally as a consequence of previous waking activity [15–19]. Extensive variability in SWA across multiple timescales (e.g. years [20], hours [21], and minutes/seconds [22, 23]) further characterizes this dynamic process. The complex spatiotemporal expression of SWA can potentially lead to different functional outcomes; slow waves preferentially downscale weak synapses while preserving synaptic strength of neurons within active ensembles [24, 25]. Consequently, to fully characterize the functional consequences of sleep, it is critical to understand how the timing and expression of SWA is coordinated across cortical networks. Infraslow fluctuations (<0.1 Hz) have previously been shown to modulate the power of higher frequency cortical activity during both sleep and wake states [22, 23, 26]. Moreover, these very slow fluctuations afford a powerful mechanism by which neuronal activity can be synchronized across disparate cortical regions [27–29]. For the present report, we investigated how infraslow activity modulates SWA expression. We observe that infraslow fluctuations are a significant source of SWA variability and coordinate SWA expression across two heterotypic, contralateral cortical regions (the right motor cortex and left parietal cortex). By examining neuronal firing rates and patterns across infraslow SWA cycles, we further show that changes in neuronal synchrony, rather than changes in neuronal firing rate, are associated with altered SWA across infraslow timescales. Methods The present study relies upon two distinct data sets: (1) chronic recordings of the electroencephalogram (EEG) and local field potential (LFP) recordings across the sleep/wake cycle in freely behaving male Sprague-Dawley rats (N = 7) and (2) LFP and neuronal spiking data recorded from frontal cortices during periods of consolidated sleep (at least 20-min episodes) and wake in male Long-Evans rats (N = 10). The first data set was recorded at Middlebury College, while the second was obtained from CRCNS.org [30], an online repository of freely available data. CRCNS.org data used this study were originally collected to address a different series of research questions [31]. Chronic EEG/LFP recordings in Sprague-Dawley rats Stereotactic surgery. Three- to four-month-old, male Sprague-Dawley rats (n = 7, Charles River; Wilmington, MA) were housed individually under standard laboratory conditions (12-h light/dark cycle, access to food and water ad libitum). To prepare for stereotactic surgery, rats were given a subcutaneous preoperative analgesic (Meloxicam; 2 mg/kg; MWI, Boise, ID), an intramuscular preoperative antibiotic (Penicillin; 100,000 units/kg), and isoflurane anesthesia (3.5% induction, 2%–3% maintenance). An LFP (0.003″ bare diameter Teflon-coated stainless steel wire, A-M Systems, Sequim, WA) was implanted into the right motor cortex (anterior-posterior [AP]: +2.0 mm, medial-lateral [ML]: +3.0 mm, dorsal-ventral [DV]: −1.5 mm). A second wire was affixed to a screw above the left parietal cortex (approximate AP: −4.0 mm, DV: −4.3 mm) and served as an electroencephalogram (EEG). Two additional stainless steel wires were affixed to screws above the cerebellum and served as a reference for the LFP/EEG leads and as a ground. To record the electromyogram (EMG), two braided stainless steel wires were inserted under the nuchal muscles. Two additional screws were attached to the skull to serve as anchors. All wires were connected to a headmount (8239 2EEG/1EMG Rat Headmount, Pinnacle Technologies, Lawrence, KS), and affixed in place with dental acrylic (Lang Dental, Wheeling, IL). One day post-surgery, each rat received a postoperative analgesic (Meloxicam; 2 mg/kg). These methods were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by Middlebury College’s Institutional Animal Care and Use Committee. Data collection, processing, and analyses. Rats were allowed a minimum of seven complete days to recover following surgery before continuous EEG/LFP/EMG recordings began. A flexible preamplifier (100× amplification, EEG/LFP high pass filter: 0.5 Hz, EMG high pass filter: 10 Hz; Pinnacle Technologies) was attached to both the rat’s headmount and a commutator (SL6C, Plastics One, Roanoke, VA) to pass electrical signals into the data acquisition system (8401 DACS, Pinnacle Technologies) and enable unobstructed movement throughout the home cage. For each rat, data were continuously recorded (250 Hz; Sirenia Acquisition, Pinnacle Technologies) across a complete 24-h light/dark cycle. Behavioral state was determined offline through manual classification of LFP/EEG/EMG signals. Waking (high-frequency, low-voltage LFP/EEG activity coupled with EMG activity), non-rapid eye movement (NREM) sleep (low-frequency, high-voltage LFP/EEG activity absent EMG activity), and REM sleep (high-frequency, low-voltage LFP/EEG activity absent EMG activity) epochs were scored in 4-s epochs and vigilance state could be resolved for all epochs. To process electrophysiological data, all EEG/LFP/EMG signals were first imported into Mathworks MATLAB (Natick, MA) and subsequently analyzed using custom scripts. As a measure of sleep need [1], we calculated SWA from the power spectra (Welch’s method, hamming window) by summing EEG or LFP power between 0.5 and 4 Hz. Average SWA was calculated in 30-min bins across the light period to examine the canonical homeostatic decline in SWA. Similar to previous reports [10, 32], to examine the underlying characteristics of individual slow waves that can contribute to SWA, EEG/LFP signals were band-pass filtered in the delta frequency range (0.5–4.0 Hz) with a zero-phase Chebyshev Type II filter. Local maxima during NREM sleep in these filtered tracings were used to identify individual slow waves. Maxima within 200 ms were considered part of the same individual slow wave (i.e. multipeak waves) while maxima separated by greater than 200 ms were considered unique individual slow waves. The amplitude, slope, and the number of peaks were quantified for each individual slow wave. Much of the analyses of the present work focused on infraslow fluctuations in SWA. We first identified all consolidated NREM episodes (min duration 168 s, absent any visually scored brief arousals or REM sleep attempts). SWA across each consolidated NREM episode was then filtered in the infraslow range (zero-phase Chebyshev Type II filter, band pass: 0.004–0.025 Hz). From these filtered signals, instantaneous phase was calculated from the angle of the complex vector obtained from the Hilbert transform [23, 33]. Instantaneous phase data were primarily used for two sets of analyses: (1) examining the coordination of SWA across motor and parietal cortices and (2) examining the relationship between characteristics of individual slow waves and infraslow SWA phase. To compare infraslow SWA phase across the motor and parietal cortices, instantaneous phase differences (IPD) were calculated (IPD = e^(i * (P1 − P2))), in which P1 and P2 are the instantaneous phase of the motor and parietal cortices, respectively. The average phase difference for that NREM episode was obtained from the imaginary component of the resultant complex number. To explore whether these infraslow phase differences are consistent across NREM episodes and rats, the infraslow cycle was first split into eight nonoverlapping phase bins. For each NREM episode, the percent of time that the motor cortex and parietal cortex exhibited an instantaneous phase difference within each of these bins was calculated. The percentages were then averaged across all NREM episodes and across all rats. To further explore the relationships between infraslow SWA in the motor and parietal cortex, another set of analyses were conducted in the time domain. The cross-correlation (max lag: 150 s) between infraslow SWA in motor and parietal cortices was calculated for each NREM episode. Resultant cross-correlograms were averaged across NREM episodes and rats. To determine whether the average cross-correlogram was likely to have arisen by chance alone, a similar approach was taken except motor and parietal cortex infraslow episodes were randomly shuffled (1,000 times for each rat) prior to calculating the cross-correlation. From this bootstrap approach, a confidence interval depicting a cross-correlogram expected by chance alone was generated and served as the comparison distribution for the observed unshuffled cross-correlogram. To investigate the relationships between infraslow SWA and infraslow fluctuations in different frequency bands, we first calculated band-limited power (BLP) for each 4-s epoch across a range of frequencies (SWA/delta: 0.5–4 Hz; theta: 6–9 Hz; alpha: 10–14 Hz; low_beta: 16–20 Hz; high_beta: 21–30 Hz; low_gamma: 40–59 Hz; and high_gamma: 61–80 Hz). Within each rat, the instantaneous phase of infraslow SWA (see above) was used to generate eight nonoverlapping phase bins that encompass the entire infraslow cycle. For each consolidated NREM episode, BLP values were normalized to the mean BLP of that NREM episode and averaged within each phase bin. Average infraslow BLP values were then correlated with average infraslow SWA. Coupled LFP/neuronal spiking recordings in Long-Evans rats To investigate potential physiological mechanisms that underlie infraslow SWA, a freely available data set was used (for full methodological details, see refs. [30, 31]). Briefly, this data set consists of continuous LFP and neuronal spiking activity recorded from deep cortical layers in frontal regions (Anterior Cingulate, Premotor, Medial Prefrontal, or Orbitofrontal Cortex) across consolidated episodes of sleep and wake in Long-Evans rats (N = 11). In the current report, a single animal’s data from the original data set could not be loaded and therefore results are presented with an N = 10. In addition to LFPs and spiking data from putative excitatory and inhibitory neurons, this data set includes scored behavioral state and time stamps that identify the presence of “On” and “Off” firing periods during NREM sleep. Firing rates and infraslow SWA. A single, cortical LFP channel (identified in the original work [31] as a clean, artifact free signal) was used for each rat for power spectral analyses. Herein, SWA, infraslow SWA, and infraslow SWA phase within each NREM episode were calculated as described above. For each putative excitatory or inhibitory neuron, firing rate data were transformed into spike counts during nonoverlapping 4-s bins. Within each NREM episode, spike counts were normalized for each neuron as a percent of that neuron’s mean firing rate across that NREM episode. Normalized firing rates were filtered in the infraslow range and instantaneous phase calculated for each neuron’s firing rate across each NREM episode. Instantaneous phase differences between these filtered, normalized firing rates and infraslow SWA were calculated as described above. To test the stability of these phase differences across time, a Rayleigh test for uniformity was performed using each neuron’s calculated phase differences across the entirety of the recording period. Deviations from uniformity (p < 0.05) reflect stability of observed phase differences. Across the entire population of neurons, an Omnibus test for circular uniformity was applied to average phase differences as these data appeared bimodal and therefore do not meet the assumptions of the Rayleigh test. To assess the relationships between neuronal firing rates and infraslow SWA across all episodes and all rats, infraslow SWA was again divided into eight nonoverlapping phase bins. Average SWA and firing rates were calculated within each bin, averaged across all NREM episodes, and then averaged across the 10 rats. “On” and “Off” periods and transition synchrony. During NREM sleep, cortical neurons exhibit a characteristic firing pattern consisting of alternations between “On” periods (brief periods of widespread neuronal activity) and “Off” periods (brief periods of neuronal silence). “On” and “Off” periods in this data set were previously detected [31] by largely following an earlier methodology [11]. Briefly, “Off” periods were characterized as short (75–1,250 ms) periods of population-wide silence. “On” periods were defined as population firing (at least 10 total spikes, 200–4,000 ms duration) occurring between “Off” periods. Transition synchrony was defined in a manner similar to that previously described [11]. Briefly, the latency from the last action potential fired for each neuron prior to an “Off” period (“On”–“Off” synchrony) or immediately following the cessation of an “Off” period (“Off”–“On” synchrony) was calculated. As these analyses were focused on neuronal activity immediately surrounding these transitions, spike latencies exceeded 50 ms were excluded. Transition synchrony was then defined as 1/(latency standard deviation). For comparisons with infraslow SWA, average “On” and “Off” period durations and transition synchrony values were calculated in nonoverlapping 4-s bins across each NREM episode as described above. To assess whether transition synchrony was associated with infraslow SWA, two complementary approaches were then undertaken: (1) average transition synchrony was calculated within eight phase bins across the infraslow cycle and a repeated measures ANOVA was used to determine whether phase significantly affected synchrony and (2) the average transition synchrony values were correlated with SWA. A similar methodological approach was taken to assess whether the number of “On” and “Off” periods varied in association with infraslow SWA. Statistical approaches. Similar statistical approaches were taken for both data sets. Repeated measures ANOVAs, correlations, and linear regressions were performed with SPSS (IBM, Armonk, NY). Circular statistics (i.e. circular mean tests, Watson-Williams tests, and circular-linear correlations) were performed in Matlab with the CircStat toolbox [34]. All linear data are presented as mean ± the standard error of the mean. All circular data are presented as the circular mean ± the standard error of the mean. Statistical significance was assessed at p < 0.05. Results Cortical SWA is homeostatically regulated across the light period SWA (the EEG/LFP power between 0.5 and 4 Hz during NREM sleep) is the best electrophysiological marker of homeostatic sleep need [1]. In Sprague-Dawley rats (N = 7), we observed a canonical decline in SWA across the light period in both the right motor cortex LFP and the contralateral parietal EEG (main effect of time, F(11,66) = 6.71, p < 0.001; no significant effects of recording site, F(1,6) = 1.48, p = 0.27, or interaction, F(11,66) = 0.45, p = 0.93; Figure 1A). Consistent with previous reports [9, 10, 12], this homeostatic decline in SWA was associated with a significant reduction in individual slow wave amplitude (F(2,12) = 63.01, p < 0.001), ascending slope (F(2,12) = 62.18, p < 0.001), and a significant increase in the proportion of multipeak slow waves (F(2,12) = 9.69, p < 0.01; see Figure 1B). Thus, as sleep need dissipates the aggregate characteristics of individual slow waves and SWA are predictably modulated. Figure 1. Open in new tabDownload slide Homeostatic sleep need reliably regulates SWA and individual slow wave characteristics. (A) Average SWA across the light period in the motor cortex (gray) and parietal cortex (black). (B) Changes in motor cortex (gray) and parietal cortex (black) individual slow wave amplitude, ascending slope, and proportion of multipeak waves as a function of sleep pressure (High: first 3 h of light period; Medium: hours 4.5–7.5, and Low: last 3 h of light period). Brackets depict significant differences (p < 0.05) from post hoc analyses of the significant main effect of sleep pressure. Inset depicts the average slow wave observed under high (gray) and low (black) sleep pressure. Figure 1. Open in new tabDownload slide Homeostatic sleep need reliably regulates SWA and individual slow wave characteristics. (A) Average SWA across the light period in the motor cortex (gray) and parietal cortex (black). (B) Changes in motor cortex (gray) and parietal cortex (black) individual slow wave amplitude, ascending slope, and proportion of multipeak waves as a function of sleep pressure (High: first 3 h of light period; Medium: hours 4.5–7.5, and Low: last 3 h of light period). Brackets depict significant differences (p < 0.05) from post hoc analyses of the significant main effect of sleep pressure. Inset depicts the average slow wave observed under high (gray) and low (black) sleep pressure. Cortical SWA is additionally regulated across infraslow timescales Although average hourly values of SWA reliably reflect sleep need, the generation of SWA is a dynamic process that exhibits extensive variability across shorter timescales (i.e. seconds to minutes). For example, the average standard deviation of SWA during each consolidated NREM episode was 83.15 ± 1.24% and 83.73 ± 1.25% of the total standard deviation observed across the entire light period for the RFLFP and LPEEG, respectively. To identify and characterize potential sources of this variability, we examined SWA within each consolidated NREM episode (see Figure 2A for an individual example). Fluctuations in SWA within NREM episodes do not appear to be fully stochastic and instead appear to exhibit spontaneous fluctuations across infraslow timescales (i.e. ~40- to 120-s periods). Consequently, SWA was filtered in the infraslow range and the instantaneous phase of this infraslow activity was calculated (Figure 2A). On average, infraslow phase alone explained 15.08 ± 1.67% of SWA variability within NREM episodes. To further identify the extent to which infraslow activity modulates SWA, for each rat recorded, linear regressions with infraslow phase and light-period time as predictors of SWA were performed (see Figure 2B for an individual example and Figure 2C for regression outputs). Including infraslow phase in addition to time of day significantly improved the regression model fit for every rat (F-change statistic; all p’s < 0.01), and in every regression the coefficients for both predictors (i.e. infraslow phase and time of day) were statistically significant (t-scores; all p’s < 0.01). On average, including these two predictors explained 26.0 ± 2.0% of total SWA variance. Thus, both time of day and infraslow phase appear to be significant sources of SWA variability. Figure 2. Open in new tabDownload slide Infraslow oscillations coordinate and modulate SWA. (A) Top panel: SWA within the right motor cortex (gray) and left parietal cortex (black) is depicted for each 4-s epoch of a consolidated NREM episode; overlay depicts filtered infraslow SWA traces. Bottom panel: instantaneous phase of infraslow SWA recorded from each cortical location. (B) SWA from all 4-s NREM epochs during the light period in a single rat as a function of the absolute value of infraslow phase and light-period time (from light onset, t = 0 to light offset, t = 12). Black lines depict the best-fit line for each predictor. Model results from this regression (Rat # 1) and those from all other rats are presented in C. (D) Average phase difference (black vector) and instantaneous phase differences (gray vectors, each 4-s epoch) between infraslow SWA in the motor and parietal cortex during the NREM episode depicted in A. (E) Average phase difference observed during all NREM episodes from all rats. Infraslow SWA is typically in-phase between motor cortex and parietal electrodes. Dashed line depicts expected phase differences by chance alone. (F) Cross-correlation between motor and parietal cortex infraslow SWA during the episode depicted in A. (G) Average cross-correlation between motor and parietal cortex infraslow across all rats and all NREM episodes (black). Gray tracing depicts a 95% confidence interval derived from shuffled data (see Methods section). Figure 2. Open in new tabDownload slide Infraslow oscillations coordinate and modulate SWA. (A) Top panel: SWA within the right motor cortex (gray) and left parietal cortex (black) is depicted for each 4-s epoch of a consolidated NREM episode; overlay depicts filtered infraslow SWA traces. Bottom panel: instantaneous phase of infraslow SWA recorded from each cortical location. (B) SWA from all 4-s NREM epochs during the light period in a single rat as a function of the absolute value of infraslow phase and light-period time (from light onset, t = 0 to light offset, t = 12). Black lines depict the best-fit line for each predictor. Model results from this regression (Rat # 1) and those from all other rats are presented in C. (D) Average phase difference (black vector) and instantaneous phase differences (gray vectors, each 4-s epoch) between infraslow SWA in the motor and parietal cortex during the NREM episode depicted in A. (E) Average phase difference observed during all NREM episodes from all rats. Infraslow SWA is typically in-phase between motor cortex and parietal electrodes. Dashed line depicts expected phase differences by chance alone. (F) Cross-correlation between motor and parietal cortex infraslow SWA during the episode depicted in A. (G) Average cross-correlation between motor and parietal cortex infraslow across all rats and all NREM episodes (black). Gray tracing depicts a 95% confidence interval derived from shuffled data (see Methods section). Infraslow activity coordinates SWA between cortical regions Infraslow fluctuations can coordinate neuronal activity between spatially segregated brain regions [27–29] and therefore may coordinate the expression of SWA throughout the cortex. To test this possibility, we used two complementary approaches to examine whether infraslow activity is coordinated between the right motor cortex LFP and contralateral parietal EEG. In the first approach, instantaneous infraslow phase was calculated within each brain region and instantaneous phase differences were derived by comparing those signals. Figures 2A and 2D depict the results of these analyses for a single NREM episode; during this episode, infraslow SWA was consistently in-phase between these cortical regions. Across all rats and NREM episodes, significant phase locking was observed between the right motor and left parietal cortices (F(7,42) = 95.95, p < 0.01; Figure 2E) with an average instantaneous phase difference that was not statistically significant from zero (circular mean ± circular standard deviation: 0.08 ± 0.10 radians; circular mean test: p > 0.05). For our second approach, we calculated the cross-correlation of infraslow SWA from each cortical region for each consolidated NREM episode. Cross-correlations from individual NREM episodes consistently exhibited high positive correlations with zero lag as well as a clear rhythmicity within infraslow timescales (see Figure 2F for an individual example). Across all animals and NREM episodes, the cross-correlation revealed high levels of synchrony around zero lag (average Pearson’s r: 0.57 ± 0.04; Figure 2G). Notably, unlike in examples from individual NREM episodes, clear infraslow rhythmicity was not observed across the total population, likely because of the nonharmonic characteristic of infraslow fluctuations [28]. Collectively, however, the significant phase locking and cross-correlations described above indicate that infraslow fluctuations can serve to coordinate SWA across disparate cortical locations. Given the clear coordination of infraslow SWA between cortical regions described above, we further explored how infraslow activity affects SWA by examining the characteristics of individual slow waves across the infraslow cycle. We first subdivided infraslow SWA recorded within the right motor cortex into eight phase bins (similar results were observed if left parietal infraslow activity was used instead, data not shown). Individual slow waves that occurred within each phase bin were then averaged to identify how infraslow phase affects slow wave characteristics (Figure 3A). Infraslow SWA phase significantly affected individual slow wave amplitude (F(7,42) = 87.24, p < 0.001), ascending slow wave slope (F(7,42) = 87.13, p < 0.001), and wave duration (F(7,42) = 5.30, p < 0.001), with peaks in infraslow SWA associated with increases in each of these characteristics. Infraslow SWA phase also significantly affected the total number of slow waves observed (F(7,42) = 5.61, p < 0.001) and the proportion of multipeak waves (F(7,42) = 15.03, p < 0.001), with peaks in infraslow SWA associated with fewer total waves and smaller proportion of multipeak waves observed. Thus, peaks in SWA across infraslow timescales appear to be arise from large, long-duration, single-peak slow waves as opposed to a general increase in the number of waves generated. Figure 3. Open in new tabDownload slide Infraslow SWA oscillations modulate individual slow wave characteristics. (A) Individual slow wave data from the right motor cortex of seven Sprague-Dawley rats. (B) Individual slow wave data from frontal cortical regions of 10 Long-Evans rats. Upper left panels in A/B depict average slow wave waveforms during infraslow SWA peaks (infraslow phase range: −Pi/4 to Pi/4; black lines) and infraslow SWA nadirs (infraslow phase ranges: −Pi to −3Pi/4 and 3Pi/4 to Pi; gray lines). The remaining panels in A/B depict how infraslow SWA phase affects the average characteristics of individual slow waves including wave amplitude, ascending slope, wave duration, total number of waves observed, and the proportion of multipeak waves observed. Figure 3. Open in new tabDownload slide Infraslow SWA oscillations modulate individual slow wave characteristics. (A) Individual slow wave data from the right motor cortex of seven Sprague-Dawley rats. (B) Individual slow wave data from frontal cortical regions of 10 Long-Evans rats. Upper left panels in A/B depict average slow wave waveforms during infraslow SWA peaks (infraslow phase range: −Pi/4 to Pi/4; black lines) and infraslow SWA nadirs (infraslow phase ranges: −Pi to −3Pi/4 and 3Pi/4 to Pi; gray lines). The remaining panels in A/B depict how infraslow SWA phase affects the average characteristics of individual slow waves including wave amplitude, ascending slope, wave duration, total number of waves observed, and the proportion of multipeak waves observed. Infraslow variation in SWA is associated with altered neuronal synchrony To identify physiological mechanisms that contribute to the generation of infraslow SWA, we turned to a freely available online data set [30] which contains coupled recordings of neuronal firing and LFP activity across periods of consolidated sleep in male Long-Evans rats. Similar to the data described above, the phase of infraslow SWA from this second data set strongly modulated characteristics of individual slow waves (Figure 3B). Peaks in infraslow SWA were associated with significant increases in the amplitude (F(7,63) = 23.59, p < 0.001), ascending slope (F(7,63) = 23.40, p < 0.001), and duration (F(7,63) = 5.55, p < 0.001) of slow waves. Meanwhile, significant decreases in the total number of waves (F(7,63) = 2.75, p < 0.05) and the proportion of multipeak waves (F(7,63) = 2.67, p < 0.05) were observed in association with peak infraslow SWA. Thus, infraslow SWA is present within both Sprague-Dawley and Long-Evans rats and appears to arise from similar alterations to the characteristics of individual slow waves within each rat strain. As SWA and individual slow waves arise from underlying patterns and levels of neuronal firing, we explored whether the firing activity of putative excitatory and inhibitory neurons was associated with infraslow SWA. Within each consolidated NREM sleep episode, spike counts for each neuron were calculated across 4-s epochs. Spike counts for each neuron were then normalized to the mean firing rate of that neuron across the NREM episode. As evident in Figure 4A, the firing rates of both excitatory and inhibitory neurons typically change in association with infraslow SWA phase; individual neurons, however, exhibit widespread heterogeneity in regards to how their firing rates changes relative to the infraslow SWA cycle. As evident from the instantaneous phase differences (PD) between infraslow SWA and the firing rates of each individual neuron (Figure 4B), some neurons exhibit peak firing in association with peak infraslow SWA (mean PD = 0), others exhibit peak firing during infraslow SWA nadirs (mean PD = −Pi/Pi), while others exhibit peak firing between these extremes. Despite this heterogeneity, phase differences between individual neurons’ firing rates and infraslow SWA across all rats were not uniform for excitatory or inhibitory neurons (Omnibus test for circular uniformity [34], both p’s < 0.001). Instead, bimodal phase differences were observed for both excitatory and inhibitory neurons with more neurons exhibiting peak firing rates around either infraslow SWA peaks or nadirs (Figure 4C/D). Phase differences for the majority of individual neurons appear to be stable across the recording session; 85.31% of neurons exhibited instantaneous phase distributions derived across all consolidated NREM episodes that significantly deviate from uniform as assessed by the Rayleigh test for nonuniformity. Thus, while the firing rates of individual neurons exhibit heterogeneous phase differences with respect to ongoing infraslow SWA, these phase differences appear largely stable across time. The underlying mechanism for this observed heterogeneity remains unclear as neither cortical region (Watson-Williams test [34]: F(3,828) = 1.73, p = 0.16), average firing rate (circular-linear correlation [34]: c = 0.02, p = 0.82), nor putative cell type (i.e. excitatory/inhibitory; Watson-Williams test: F(1,830) = 0.76, p = 0.38) significantly accounted for observed phase differences. Figure 4. Open in new tabDownload slide Infraslow SWA is not correlated with excitatory or inhibitory firing rates. (A) Top trace: infraslow SWA across a consolidated NREM episode. Heatmap: firing rates (normalized to the mean firing rate of this NREM episode for each neuron) for each putative excitatory and inhibitory neuron are depicted as separate rows in 4-s bins across the NREM episode. Composite excitatory and inhibitory firing rates are also depicted. Excitatory and inhibitory neurons have been sorted by their average phase difference with the infraslow SWA to better visualize the observed heterogeneity of phase differences. To limit the effect of outliers on the heatmap visualization, all firing rats >300% of the episode mean are plotted as equal to 300%. (B) Average phase differences between infraslow SWA and each individual neuron that is depicted in A. (C/D) Histogram of observed phase differences between infraslow SWA and excitatory (C) or inhibitory neurons (D). Data are double-plotted to better visualize the bimodal distribution of phase differences. (E/F) Across all NREM episodes and all rats, neither the mean firing rates of excitatory (E) nor inhibitory neurons (F) were significantly correlated with SWA across the infraslow cycle. Figure 4. Open in new tabDownload slide Infraslow SWA is not correlated with excitatory or inhibitory firing rates. (A) Top trace: infraslow SWA across a consolidated NREM episode. Heatmap: firing rates (normalized to the mean firing rate of this NREM episode for each neuron) for each putative excitatory and inhibitory neuron are depicted as separate rows in 4-s bins across the NREM episode. Composite excitatory and inhibitory firing rates are also depicted. Excitatory and inhibitory neurons have been sorted by their average phase difference with the infraslow SWA to better visualize the observed heterogeneity of phase differences. To limit the effect of outliers on the heatmap visualization, all firing rats >300% of the episode mean are plotted as equal to 300%. (B) Average phase differences between infraslow SWA and each individual neuron that is depicted in A. (C/D) Histogram of observed phase differences between infraslow SWA and excitatory (C) or inhibitory neurons (D). Data are double-plotted to better visualize the bimodal distribution of phase differences. (E/F) Across all NREM episodes and all rats, neither the mean firing rates of excitatory (E) nor inhibitory neurons (F) were significantly correlated with SWA across the infraslow cycle. To further explore the relationships between firing rate and infraslow SWA at a population level, we first subdivided the infraslow SWA cycle into eight phase bins. For each rat, normalized, mean firing rates for putative excitatory or inhibitory neurons across all NREM episodes were averaged within each phase bin. Average SWA was likewise calculated within each phase bin. Neither excitatory (r = −0.01, p = 0.95; Figure 4E) nor inhibitory (r = 0.09, p = 0.47; Figure 4F) firing rates were correlated with SWA across infraslow cycles. Thus, although the firing rate of many neurons appears to be modulated across infraslow timescales, the relationship of these firing rates to infraslow SWA is highly variable. Infraslow SWA does not appear, therefore, to arise from consistent alterations in population firing rates. Alternating periods of neuronal silence (“Off” periods) and neuronal firing (“On” periods) underlie sleep, slow waves (Figure 5A). Previous reports have indicated that homeostatic alterations in SWA are associated with changes in neuronal synchrony during transitions between “On” and “Off” periods [11]. Similar changes in synchrony could contribute to the observed infraslow SWA variability observed in the present study. Consequently, we calculated indices of neuronal synchrony surrounding transitions from “On” to “Off” states, and vice versa. Unlike the firing rate modulation described above, transition synchrony was reliably modulated across the infraslow cycle (“On”–“Off” transitions: F(7,63) = 3.89, p < 0.01; “Off”–“On” transitions: F(7,63) = 4.30, p < 0.01; Figure 5B). Across infraslow cycles, SWA was significantly correlated with transition synchrony (“On”–“Off” synchrony: r = 0.47, p < 0.001, “Off”–“On” synchrony: r = 0.39, p < 0.001; Figure 5C). Together, these results appear to indicate that enhanced neuronal recruitment (i.e. from “Off” to “On” periods) and decruitment (i.e. from “On” to “Off” periods) contribute to infraslow variations in SWA. Figure 5. Open in new tabDownload slide Infraslow SWA is associated with changes in neuronal synchrony and the number and duration of off periods. (A) A raster plot of putative excitatory and inhibitory firing across “On” and “Off” periods (white and gray, respectively). Each row represents an individual neuron and each vertical tick an action potential. Composite excitatory and inhibitory firing are also depicted. (B) Average SWA (light gray), neuronal synchrony during transitions to “Off” periods from “On” periods (gray), and neuronal synchrony during transitions to “On” periods from “Off” periods (black) are depicted as a function of the infraslow SWA cycle. (C) Neuronal synchrony is significantly correlated with SWA across the infraslow cycle. (D) Average SWA (light gray), “Off” period duration (gray), and number of “Off” periods (black) are depicted as a function of the infraslow SWA cycle. (E) The number and duration of “Off” periods are significantly correlated with SWA across the infraslow cycle. B/D are each double-plotted to better visualize infraslow rhythmicity. Figure 5. Open in new tabDownload slide Infraslow SWA is associated with changes in neuronal synchrony and the number and duration of off periods. (A) A raster plot of putative excitatory and inhibitory firing across “On” and “Off” periods (white and gray, respectively). Each row represents an individual neuron and each vertical tick an action potential. Composite excitatory and inhibitory firing are also depicted. (B) Average SWA (light gray), neuronal synchrony during transitions to “Off” periods from “On” periods (gray), and neuronal synchrony during transitions to “On” periods from “Off” periods (black) are depicted as a function of the infraslow SWA cycle. (C) Neuronal synchrony is significantly correlated with SWA across the infraslow cycle. (D) Average SWA (light gray), “Off” period duration (gray), and number of “Off” periods (black) are depicted as a function of the infraslow SWA cycle. (E) The number and duration of “Off” periods are significantly correlated with SWA across the infraslow cycle. B/D are each double-plotted to better visualize infraslow rhythmicity. Changes in the efficacy of recruitment/decruitment could further influence SWA by altering the duration and/or number of “On” and/or “Off” periods. Similar to the changes in neuronal synchrony described above, the number F(7,63) = 13.21, p < 0.001) and duration (F(7,63) = 5.34, p < 0.001) of “Off” periods were significantly modulated across the infraslow SWA cycle; increases in SWA across the infraslow cycle were associated with more frequent and longer “Off” periods (Figure 5D). Consistent with this observation, SWA across the infraslow cycle was significantly correlated with both the number (r = 0.79, p < 0.001) and duration (r = 0.66, p < 0.001) of “Off” periods (Figure 5E). Although the number of “On” periods was similarly modulated across the infraslow SWA cycle (F(7,63) = 9.24, p < 0.001), we did not observe a significant effect of infraslow SWA cycle on “On” period duration (F(7,63) = 1.44, p = 0.21). Thus, changes in the number of “On”/“Off” periods, “Off” period duration, and transition synchrony are associated with infraslow SWA. Infraslow modulation extends beyond sleep SWA Previous reports indicate that infraslow activity can modulate EEG power across a wide range of higher frequencies during both sleep and wake [23, 26, 35]. We therefore examined whether the infraslow alterations in SWA documented above were additionally associated with changes in EEG power within other frequency bands (see Methods section for detailed explanation). Across infraslow cycles, SWA was significantly correlated with both theta and alpha power (r(110) = 0.90, p < 0.01 and r(110) = 0.32, p < 0.05, respectively) but not with power in other frequency bands (low_beta: p = 0.80, high_beta: p = 0.74, low_gamma: p = 0.15, high_gamma: p = 0.61). Using a similar approach, we further investigated whether higher frequency power was also modulated by infraslow activity during consolidated waking bouts. Strikingly, waking infraslow activity in the delta range (0.5–4 Hz, the same frequency range used to calculate SWA during NREM sleep) was significantly correlated with infraslow activity across all of the higher frequency bands investigated (theta: r = 0.89, alpha: r = 0.60, low_beta: r = 0.80, high_beta: r = 0.69, low_gamma: r = 0.88, and high_gamma: r = 0.78; all df = 110, all p < 0.01). Thus, similar to previous reports [23, 26, 35], infraslow fluctuations observed in the present study contribute to significant nesting of higher frequency power across multiple frequency bands. While significant nesting persists during NREM sleep, it appears to only do so within relatively lower frequencies (delta, alpha, and theta). Discussion The expression of cortical SWA is homeostatically regulated with a well-characterized decline across prolonged periods of sleep. We observe that SWA is additionally regulated across shorter timescales; extensive infraslow SWA fluctuations (<0.1 Hz) are present within each NREM episode and contribute to intra-episode SWA variability that is ~80%–85% of that observed across the typical homeostatic decline. Infraslow SWA fluctuations were typically in-phase between the right motor and left parietal cortices and thereby appear to coordinate SWA across disparate cortical locations. Peak SWA across the infraslow cycle was associated with increased amplitude, slope, and duration of individual slow waves with a concurrent decrease in the number of slow waves and proportion of multipeak waves. SWA peaks were additionally associated with increased neuronal synchrony surrounding transitions between “On” and “Off” periods and more frequent and longer “Off” period durations. By contrast, firing rates of putative excitatory or inhibitory neurons were not significantly associated with infraslow SWA variability. Through characterizing the coordination and control of SWA expression, we can better understand how this neuronal activity directly contributes to sleep function. The function of sleep, and by association the physiological role of SWA, remains unclear. A growing consensus [36–41], however, suggests that SWA is important for the regulation of memory and corresponding synaptic plasticity. Indeed, enhancing SWA through electrical [3] or auditory [4, 42] stimulation produces marked improvements in memory while selective SWA deprivation impairs memory [2, 43]. Two primary mechanisms have been proposed to account for SWA-dependent memory improvement: (1) the selective strengthening and consolidation of specific memory traces [44–47] and (2) nonselective reductions in global synaptic strength that could improve signal-to-noise ratio within cortical networks [48–50]. Recent evidence may provide a basis to integrate these otherwise contrasting roles; SWA may broadly downscale synaptic strength while maintaining and/or consolidating synaptic strength of specific memory traces [24, 25, 51–53]. Herein, the depolarized “Up” state of the slow wave promotes synaptic weakening unless concurrent suprathreshold postsynaptic spiking activity is present [24, 25]. Memory replay during SWA, therefore, may preserve synaptic weight within active neuronal circuits while enabling widespread reductions in synaptic strength elsewhere [51–53]. Such a mechanism could underlie selective memory facilitation observed when memory-specific odor [54] or auditory cues [55] are delivered during slow wave sleep. Understanding the spatiotemporal expression of SWA, therefore, appears critical for understanding its role in shaping cortical plasticity. We observe that SWA exhibits extensive intra-episode variability that is coordinated between the right motor and left parietal cortices across infraslow timescales. During waking, correlated infraslow fluctuations in resting state fMRI activity form the basis for a diverse array of functional networks [56–58]. Although similar network connectivity across infraslow timescales appears to be maintained during light sleep [59], sleep can induce robust alterations to this network communication [60, 61]. Consistent, in-phase infraslow SWA between heterotypic, contralateral recording sites in the present study further demonstrates the ability of these fluctuations to coordinate activity across the cortex. This coordination could facilitate the highly correlated expression of slow waves across cortical hemispheres [29]. Although the underlying mechanisms responsible for the generation of infraslow activity are still being characterized, these slow fluctuations have been associated with changes in brain metabolism [62] and widespread alterations in neuronal excitability [27, 63]. Astrocytic adenosine has been proposed to contribute to infraslow activity as (1) infraslow fluctuations appear dependent upon the activation of A1 receptors by adenosine [62], (2) astrocytic syncytia exhibit coordinated changes in intracellular calcium signaling across large spatial scales that fluctuate across infraslow timescales [64, 65], and (3) calcium-dependent gliotransmission can modulate neuronal activity through A1 receptors [66–68]. Infraslow alterations in adenosine could contribute to fluctuations in infraslow SWA and neuronal synchrony observed in the present study. Astrocytic calcium transients and adenosine play causal roles in the generation of the cortical slow oscillation underlying NREM SWA [67, 69]. Indeed, electrically stimulating astrocytes increases the frequency of neuronal Up states while blocking astrocytic calcium or A1 receptors decreases Up state frequency [70]. Thus, in addition to playing a causal role in the homeostatic regulation of cortical SWA [71, 72], astrocytic adenosine may additionally regulate SWA across infraslow cycles. Infraslow activity has also been shown to modulate neuronal excitability including alterations in higher frequency EEG/LFP power [22, 23, 26, 35], cortical interictal events in epileptic patients [23], and spontaneous hippocampal afterdischarges in Wistar rats [73]. In the present study, we replicate previous findings [23, 26, 35] that demonstrate significant nesting of higher frequency BLP (0.5–80 Hz) across infraslow cycles during waking. Moreover, we show that significant reorganization of these associations occurs during NREM sleep with infraslow activity at higher frequencies (>16 Hz) no longer correlating with lower frequency activity (0.5–14 Hz). Although the functional significance of this reorganization remains to be elucidated, sleep-associated reorganization of infraslow propagation has previously been proposed to contribute to sleep function, thalamic gating, and/or sleep-associated alterations in consciousness [74]. We further observe that infraslow fluctuations alter the characteristics of spontaneous slow waves, with infraslow SWA nadirs associated with decreased slope, amplitude, and duration of individual slow waves and an increased proportion of multipeak waves. Strikingly, these changes parallel homeostatic alterations in slow wave characteristics [10] (and see Figure 1B) thought to arise as a consequence of persistent reductions in synaptic strength [9, 10, 49]. While similar changes in structural connectivity are unlikely to account for changes in slow waves across infraslow cycles, infraslow activity is dependent upon synaptic transmission (e.g. attenuated by antagonists of voltage-gated sodium channels or glutamate receptors [27]). Consequently, changes in functional connectivity that alter synaptic efficacy across infraslow timescales could account for the similarity with homeostatic slow wave modulation. Consistent with this idea, coupling between electrically evoked postsynaptic excitatory potentials and action potential generation has been shown to vary across infraslow cycles [33]. To further understand the physiological underpinnings of infraslow SWA, we examined how neuronal firing rates and patterns vary across the infraslow cycle. During spontaneous sleep across the major sleep phase, firing rates during “On” periods are significantly correlated with SWA [11], with average population firing rates declining over periods of sleep [11, 31]. However, elevated firing per se, does not appear to cause increased SWA; optogenetic stimulation to increase firing rates during NREM sleep was associated with reductions in SWA [75]. Moreover, NREM sleep bidirectionally regulates firing rates of individual neurons by increasing/decreasing firing rates in neurons with low/high basal firing levels, respectively [31]. Consistent with these results, we observed that neither the average firing rates of putative excitatory nor inhibitory neurons were correlated with SWA across infraslow cycles. This absence of significant correlation may arise from extensive heterogeneity in the phase relationship between an individual neuron’s firing rate and infraslow SWA (Figure 4A/B). Across our population of neurons (Figure 4C/D), firing rates exhibited a bimodal distribution with elevated firing rates predominantly occurring around the peaks and nadirs of infraslow SWA. A strikingly similar relationship between individual neuronal firing rates and population activity has recently been described in the medial prefrontal cortex of freely behaving mice across slow timescales (<0.1 Hz) that is absent across fast timescales [76]. Firing rate phase heterogeneity, with respect to infraslow SWA, may therefore reflect an emergent organizational property of neuronal activity across slow timescales. Indeed, neither cortical location, average firing rate, nor putative cell type significantly accounted for this phase heterogeneity. Alternatively, it is intriguing to hypothesize whether this phase heterogeneity reflects the existence of multiple, distinct infraslow ensembles (e.g. [27]), though the present data set does not facilitate a direct test of this hypothesis. Unlike firing rates, measures of neuronal synchrony and the number and duration of “Off” periods were significantly correlated with infraslow SWA. Cortical slow waves arise from alternating patterns of widespread neuronal activity and neuronal silence (“On” and “Off” periods, respectively [11, 31]). In association with the homeostatic regulation of SWA, the number and duration of “Off” periods [11, 75] decrease across prolonged periods of sleep; a similar pattern is observed for LFP “Down” states that are associated with “Off” periods [31]. Likewise, neuronal synchrony surrounding “On”/“Off” and “Off”/“On” transitions decreases across sleep [11]. These alterations appear, in part, to be driven by widespread reductions in synaptic strength [10, 50, 77]. It is unlikely that persistent changes in synaptic strength are responsible for similar changes in neuronal synchrony associated with infraslow SWA observed in the current study. Rather, as discussed above, transient changes in neuronal excitability and/or synaptic efficacy across the infraslow cycle could account for these observations. To wit, computer simulations indicate that the synchronous termination of active firing across slow waves can be mediated by the strength of synaptic inhibition and/or the excitability of inhibitory neurons [78]. Therefore, changes in the excitability of inhibitory neurons across the infraslow cycle may provide a functional mechanism to alter synchrony surrounding “On”/“Off” transitions and thereby produce the observed infraslow fluctuations in SWA. From “Off” period number and duration, to neuronal transition synchrony, to characteristics of individual slow waves (i.e. slope, amplitude, and proportion of multipeak waves), we observe changes across infraslow cycles that contribute to significant SWA variability. Infraslow SWA was consistently in-phase across disparate cortical regions. Consequently, infraslow fluctuations provide an important mechanism to regulate the spatiotemporal expression of SWA. Infraslow and homeostatic regulators of SWA appear to alter slow wave generation in a strikingly similar manner, albeit across vastly different timescales. Sleep spindles also appear to be regulated across both the sleep cycle [7, 79] and infraslow timescales [80] in order to balance sensory reactivity with off-line memory consolidation [80]. The precise functional significance of regulating SWA across both infraslow and homeostatic timescales, however, remains to be elucidated. Acknowledgments We thank the Collaborative Research in Computational Neuroscience program (CRCNS; crcns.org) for hosting the freely available data set used within this report and the authors [31] who collected the data originally. Funding Research reported in this publication was supported by an Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health under grant number P20GM103449. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NIGMS or NIH. Conflict of interest statement. None declared. References 1. Achermann P , et al. Mathematical models of sleep regulation . Front Biosci . 2003 ; 8 : s683 – s693 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Landsness EC , et al. Sleep-dependent improvement in visuomotor learning: a causal role for slow waves . Sleep . 2009 ; 32 ( 10 ): 1273 – 1284 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Marshall L , et al. Boosting slow oscillations during sleep potentiates memory . Nature . 2006 ; 444 ( 7119 ): 610 – 613 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Ngo HV , et al. Auditory closed-loop stimulation of the sleep slow oscillation enhances memory . Neuron . 2013 ; 78 ( 3 ): 545 – 553 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Murphy M , et al. Source modeling sleep slow waves . Proc Natl Acad Sci USA . 2009 ; 106 ( 5 ): 1608 – 1613 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Massimini M , et al. The sleep slow oscillation as a traveling wave . J Neurosci . 2004 ; 24 ( 31 ): 6862 – 6870 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Nir Y , et al. Regional slow waves and spindles in human sleep . Neuron . 2011 ; 70 ( 1 ): 153 – 169 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Vyazovskiy VV , et al. Local sleep in awake rats . Nature . 2011 ; 472 ( 7344 ): 443 – 447 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Riedner BA , et al. Sleep homeostasis and cortical synchronization: III. A high-density EEG study of sleep slow waves in humans . Sleep . 2007 ; 30 ( 12 ): 1643 – 1657 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Vyazovskiy VV , et al. Sleep homeostasis and cortical synchronization: II. A local field potential study of sleep slow waves in the rat . Sleep . 2007 ; 30 ( 12 ): 1631 – 1642 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Vyazovskiy VV , et al. Cortical firing and sleep homeostasis . Neuron . 2009 ; 63 ( 6 ): 865 – 878 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Bernardi G , et al. Local and widespread slow waves in stable NREM sleep: evidence for distinct regulation mechanisms . Front Hum Neurosci . 2018 ; 12 : 248 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Werth E , et al. Brain topography of the human sleep EEG: antero-posterior shifts of spectral power . Neuroreport . 1996 ; 8 ( 1 ): 123 – 127 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Cajochen C , et al. Frontal predominance of a relative increase in sleep delta and theta EEG activity after sleep loss in humans . Sleep Res Online . 1999 ; 2 ( 3 ): 65 – 69 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 15. Huber R , et al. Local sleep and learning . Nature . 2004 ; 430 ( 6995 ): 78 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Hanlon EC , et al. Effects of skilled training on sleep slow wave activity and cortical gene expression in the rat . Sleep . 2009 ; 32 ( 6 ): 719 – 729 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Kattler H , et al. Effect of unilateral somatosensory stimulation prior to sleep on the sleep EEG in humans . J Sleep Res . 1994 ; 3 ( 3 ): 159 – 164 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Korf EM , et al. Blindfolding during wakefulness causes decrease in sleep slow wave activity . Physiol Rep . 2017 ; 5 ( 7 ): e13239 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Huber R , et al. Arm immobilization causes cortical plastic changes and locally decreases sleep slow wave activity . Nat Neurosci . 2006 ; 9 ( 9 ): 1169 – 1176 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Varga AW , et al. Effects of aging on slow-wave sleep dynamics and human spatial navigational memory consolidation . Neurobiol Aging . 2016 ; 42 : 142 – 149 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Borbély AA , et al. The two-process model of sleep regulation: a reappraisal . J Sleep Res . 2016 ; 25 ( 2 ): 131 – 143 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Liu Z , et al. Large-scale spontaneous fluctuations and correlations in brain electrical activity observed with magnetoencephalography . Neuroimage . 2010 ; 51 ( 1 ): 102 – 111 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Vanhatalo S , et al. Infraslow oscillations modulate excitability and interictal epileptic activity in the human cortex during sleep . Proc Natl Acad Sci USA . 2004 ; 101 ( 14 ): 5053 – 5057 . Google Scholar Crossref Search ADS PubMed WorldCat 24. González-Rueda A , et al. Activity-dependent downscaling of subthreshold synaptic inputs during slow-wave-sleep-like activity in vivo . Neuron . 2018 ; 97 ( 6 ): 1244 – 1252.e5 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Bartram J , et al. Cortical up states induce the selective weakening of subthreshold synaptic inputs . Nat Commun . 2017 ; 8 ( 1 ): 665 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Monto S , et al. Very slow EEG fluctuations predict the dynamics of stimulus detection and oscillation amplitudes in humans . J Neurosci . 2008 ; 28 ( 33 ): 8268 – 8272 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Chan AW , et al. Mesoscale infraslow spontaneous membrane potential fluctuations recapitulate high-frequency activity cortical motifs . Nat Commun . 2015 ; 6 : 7738 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Palva JM , et al. Infra-slow fluctuations in electrophysiological recordings, blood-oxygenation-level-dependent signals, and psychophysical time series . Neuroimage . 2012 ; 62 ( 4 ): 2201 – 2211 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Mohajerani MH , et al. Mirrored bilateral slow-wave cortical activity within local circuits revealed by fast bihemispheric voltage-sensitive dye imaging in anesthetized and awake mice . J Neurosci . 2010 ; 30 ( 10 ): 3745 – 3751 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Watson B , et al. Multi-unit spiking activity recorded from rat frontal cortex (brain regions MPFC, OFC, ACC, and M2) during wake-sleep episode wherein at least 7 minutes of wake are followed by 20 minutes of sleep . CRCNS.org . doi: 10.6080/K02N506Q OpenURL Placeholder Text WorldCat Crossref 31. Watson BO , et al. Network homeostasis and state dynamics of neocortical sleep . Neuron . 2016 ; 90 ( 4 ): 839 – 852 . Google Scholar Crossref Search ADS PubMed WorldCat 32. Panagiotou M , et al. Differences in electroencephalographic non-rapid-eye movement sleep slow-wave characteristics between young and old mice . Sci Rep . 2017 ; 7 : 43656 . Google Scholar Crossref Search ADS PubMed WorldCat 33. Dash MB , et al. Spontaneous infraslow fluctuations modulate hippocampal EPSP-PS coupling . eNeuro . 2018 ; 5 ( 1 ): 403 – 417 . Google Scholar Crossref Search ADS WorldCat 34. Berens P . CircStat: a MATLAB toolbox for circular statistics . J Stat Softw . 2009 ; 31 ( 10 ): 1 – 21 . Google Scholar Crossref Search ADS WorldCat 35. Mantini D , et al. Electrophysiological signatures of resting state networks in the human brain . Proc Natl Acad Sci USA . 2007 ; 104 ( 32 ): 13170 – 13175 . Google Scholar Crossref Search ADS PubMed WorldCat 36. Puentes-Mestril C , et al. Linking network activity to synaptic plasticity during sleep: hypotheses and recent data . Front Neural Circuits . 2017 ; 11 : 61 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Niethard N , et al. Plasticity during sleep is linked to specific regulation of cortical circuit activity . Front Neural Circuits . 2017 ; 11 : 65 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Levenstein D , et al. Sleep regulation of the distribution of cortical firing rates . Curr Opin Neurobiol . 2017 ; 44 : 34 – 42 . Google Scholar Crossref Search ADS PubMed WorldCat 39. Timofeev I , et al. Sleep slow oscillation and plasticity . Curr Opin Neurobiol . 2017 ; 44 : 116 – 126 . Google Scholar Crossref Search ADS PubMed WorldCat 40. Abel T , et al. Sleep, plasticity and memory from molecules to whole-brain networks . Curr Biol . 2013 ; 23 ( 17 ): R774 – R788 . Google Scholar Crossref Search ADS PubMed WorldCat 41. Tononi G , et al. Sleep and synaptic down-selection . Eur J Neurosci . 2019 ; 1 – 9 . doi: 10.1111/ejn.14335 OpenURL Placeholder Text WorldCat Crossref 42. Papalambros NA , et al. Acoustic enhancement of sleep slow oscillations and concomitant memory improvement in older adults . Front Hum Neurosci . 2017 ; 11 : 109 . Google Scholar Crossref Search ADS PubMed WorldCat 43. Aeschbach D , et al. A role for non-rapid-eye-movement sleep homeostasis in perceptual learning . J Neurosci . 2008 ; 28 ( 11 ): 2766 – 2772 . Google Scholar Crossref Search ADS PubMed WorldCat 44. Chauvette S , et al. Sleep oscillations in the thalamocortical system induce long-term neuronal plasticity . Neuron . 2012 ; 75 ( 6 ): 1105 – 1113 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Aton SJ , et al. Mechanisms of sleep-dependent consolidation of cortical plasticity . Neuron . 2009 ; 61 ( 3 ): 454 – 466 . Google Scholar Crossref Search ADS PubMed WorldCat 46. Walker MP . The role of slow wave sleep in memory processing . J Clin Sleep Med . 2009 ; 5 ( 2 Suppl. ): S20 – S26 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 47. Frank MG , et al. Sleep enhances plasticity in the developing visual cortex . Neuron . 2001 ; 30 ( 1 ): 275 – 287 . Google Scholar Crossref Search ADS PubMed WorldCat 48. Tononi G , et al. Sleep function and synaptic homeostasis . Sleep Med Rev . 2006 ; 10 ( 1 ): 49 – 62 . Google Scholar Crossref Search ADS PubMed WorldCat 49. Vyazovskiy VV , et al. Molecular and electrophysiological evidence for net synaptic potentiation in wake and depression in sleep . Nat Neurosci . 2008 ; 11 ( 2 ): 200 – 208 . Google Scholar Crossref Search ADS PubMed WorldCat 50. Olcese U , et al. Sleep and synaptic renormalization: a computational study . J Neurophysiol . 2010 ; 104 ( 6 ): 3476 – 3493 . Google Scholar Crossref Search ADS PubMed WorldCat 51. Gulati T , et al. Neural reactivations during sleep determine network credit assignment . Nat Neurosci . 2017 ; 20 ( 9 ): 1277 – 1284 . Google Scholar Crossref Search ADS PubMed WorldCat 52. Wei Y , et al. Differential roles of sleep spindles and sleep slow oscillations in memory consolidation . PLoS Comput Biol . 2018 ; 14 ( 7 ): e1006322 . Google Scholar Crossref Search ADS PubMed WorldCat 53. Hashmi A , et al. Sleep-dependent synaptic down-selection (II): single-neuron level benefits for matching, selectivity, and specificity . Front Neurol . 2013 ; 4 : 148 . Google Scholar Crossref Search ADS PubMed WorldCat 54. Rasch B , et al. Odor cues during slow-wave sleep prompt declarative memory consolidation . Science . 2007 ; 315 ( 5817 ): 1426 – 1429 . Google Scholar Crossref Search ADS PubMed WorldCat 55. Antony JW , et al. Cued memory reactivation during sleep influences skill learning . Nat Neurosci . 2012 ; 15 ( 8 ): 1114 – 1116 . Google Scholar Crossref Search ADS PubMed WorldCat 56. Smith SM , et al. Correspondence of the brain’s functional architecture during activation and rest . Proc Natl Acad Sci USA . 2009 ; 106 ( 31 ): 13040 – 13045 . Google Scholar Crossref Search ADS PubMed WorldCat 57. Karahanoğlu FI , et al. Transient brain activity disentangles fMRI resting-state dynamics in terms of spatially and temporally overlapping networks . Nat Commun . 2015 ; 6 : 7751 . Google Scholar Crossref Search ADS PubMed WorldCat 58. Biswal B , et al. Functional connectivity in the motor cortex of resting human brain using echo-planar MRI . Magn Reson Med . 1995 ; 34 ( 4 ): 537 – 541 . Google Scholar Crossref Search ADS PubMed WorldCat 59. Larson-Prior LJ , et al. Cortical network functional connectivity in the descent to sleep . Proc Natl Acad Sci USA . 2009 ; 106 ( 11 ): 4489 – 4494 . Google Scholar Crossref Search ADS PubMed WorldCat 60. Tagliazucchi E , et al. Decoding wakefulness levels from typical fMRI resting-state data reveals reliable drifts between wakefulness and sleep . Neuron . 2014 ; 82 ( 3 ): 695 – 708 . Google Scholar Crossref Search ADS PubMed WorldCat 61. Tagliazucchi E , et al. Large-scale brain functional modularity is reflected in slow electroencephalographic rhythms across the human non-rapid eye movement sleep cycle . Neuroimage . 2013 ; 70 : 327 – 339 . Google Scholar Crossref Search ADS PubMed WorldCat 62. Lörincz ML , et al. ATP-dependent infra-slow (<0.1 Hz) oscillations in thalamic networks . PLoS One . 2009 ; 4 ( 2 ): e4447 . Google Scholar Crossref Search ADS PubMed WorldCat 63. Watson BO . Cognitive and physiologic impacts of the infraslow oscillation . Front Syst Neurosci . 2018 ; 12 : 44 . Google Scholar Crossref Search ADS PubMed WorldCat 64. Wang M , et al. Brain-state dependent astrocytic Ca2+ signals are coupled to both positive and negative BOLD-fMRI signals . Proc Natl Acad Sci USA . 2018 ; 115 ( 7 ): E1647 – E1656 . Google Scholar Crossref Search ADS PubMed WorldCat 65. Kuga N , et al. Large-scale calcium waves traveling through astrocytic networks in vivo . J Neurosci . 2011 ; 31 ( 7 ): 2607 – 2614 . Google Scholar Crossref Search ADS PubMed WorldCat 66. Parpura V , et al. Gliotransmission: exocytotic release from astrocytes . Brain Res Rev . 2010 ; 63 ( 1–2 ): 83 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat 67. Fellin T , et al. Endogenous nonneuronal modulators of synaptic transmission control cortical slow oscillations in vivo . Proc Natl Acad Sci USA . 2009 ; 106 ( 35 ): 15037 – 15042 . Google Scholar Crossref Search ADS PubMed WorldCat 68. Pascual O , et al. Astrocytic purinergic signaling coordinates synaptic networks . Science . 2005 ; 310 ( 5745 ): 113 – 116 . Google Scholar Crossref Search ADS PubMed WorldCat 69. Szabó Z , et al. Extensive astrocyte synchronization advances neuronal coupling in slow wave activity in vivo . Sci Rep . 2017 ; 7 ( 1 ): 6018 . Google Scholar Crossref Search ADS PubMed WorldCat 70. Poskanzer KE , et al. Astrocytic regulation of cortical UP states . Proc Natl Acad Sci USA . 2011 ; 108 ( 45 ): 18453 – 18458 . Google Scholar Crossref Search ADS PubMed WorldCat 71. Nadjar A , et al. Astrocyte-derived adenosine modulates increased sleep pressure during inflammatory response . Glia . 2013 ; 61 ( 5 ): 724 – 731 . Google Scholar Crossref Search ADS PubMed WorldCat 72. Halassa MM , et al. Astrocytic modulation of sleep homeostasis and cognitive consequences of sleep loss . Neuron . 2009 ; 61 ( 2 ): 213 – 219 . Google Scholar Crossref Search ADS PubMed WorldCat 73. Penttonen M , et al. Ultra-slow oscillation (0.025 Hz) triggers hippocampal afterdischarges in Wistar rats . Neuroscience . 1999 ; 94 ( 3 ): 735 – 743 . Google Scholar Crossref Search ADS PubMed WorldCat 74. Mitra A , et al. Propagated infra-slow intrinsic brain activity reorganizes across wake and slow wave sleep . eLife . 2015 ; 4 : e10781 . doi: 10.7554/eLife.10781 Google Scholar Crossref Search ADS PubMed WorldCat Crossref 75. Rodriguez AV , et al. Why does sleep slow-wave activity increase after extended wake? Assessing the effects of increased cortical firing during wake and sleep . J Neurosci . 2016 ; 36 ( 49 ): 12436 – 12447 . Google Scholar Crossref Search ADS PubMed WorldCat 76. Okun M , et al. Distinct structure of cortical population activity on fast and infraslow timescales . Cereb Cortex . 2019 ; 29 ( 5 ): 2196 – 2210 . Google Scholar Crossref Search ADS PubMed WorldCat 77. Esser SK , et al. Sleep homeostasis and cortical synchronization: I. Modeling the effects of synaptic strength on sleep slow waves . Sleep . 2007 ; 30 ( 12 ): 1617 – 1630 . Google Scholar Crossref Search ADS PubMed WorldCat 78. Chen JY , et al. Interneuron-mediated inhibition synchronizes neuronal activity during slow oscillation . J Physiol . 2012 ; 590 ( 16 ): 3987 – 4010 . Google Scholar Crossref Search ADS PubMed WorldCat 79. Andrillon T , et al. Sleep spindles in humans: insights from intracranial EEG and unit recordings . J Neurosci . 2011 ; 31 ( 49 ): 17821 – 17834 . Google Scholar Crossref Search ADS PubMed WorldCat 80. Lecci S , et al. Coordinated infraslow neural and cardiac oscillations mark fragility and offline periods in mammalian sleep . Sci Adv . 2017 ; 3 ( 2 ): e1602026 . Google Scholar Crossref Search ADS PubMed WorldCat © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail [email protected]. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)