SLEEP

Gaston, S A; Jackson, W; Jackson, C L

doi: 10.1093/sleep/zsy061.740pmid: N/A

Abstract Introduction Many racial/ethnic minorities disproportionately live in rental properties and environments that may contribute to suboptimal sleep and cardiometabolic health, but few studies have investigated racial disparities in sleep and cardiometabolic health by housing status. Methods To study suboptimal sleep and cardiometabolic health disparities between US-born Hispanics/Latinos and non-Hispanic (NH) Whites by rental/homeownership status, we used pooled cross-sectional National Health Interview Survey data (2004–2016). Hispanic/Latino subgroups included: Mexicans, Puerto Ricans, Cubans/Dominicans, Central/South Americans, and other Hispanic/Spanish/multiple-ethnicities. Categories of self-reported habitual sleep duration were short (6-<7 hours), recommended (7-<9 hours), and long (≥9 hours). Using Poisson regression with robust variance, stratified by housing status (i.e., government-assisted and unassisted rental; homeownership), we estimated prevalence ratios (PRs) for Hispanics/Latinos compared to NH-Whites for sleep duration and cardiometabolic health outcomes within sleep duration categories. All models were adjusted for sociodemographics and overall health. Results Among 236,040 adults (51% women), most were NH-White (90%), Mexican (7%), or Puerto Rican (1%). Government-assisted and unassisted renters were younger than homeowners, and most government-assisted renters were female (57–71%, depending on racial/ethnic group). Among government-assisted renters, short sleep was less prevalent among Central/South Americans than NH-Whites. We observed no Central/South American-White cardiometabolic health disparities among government-assisted renters with short sleep. Among unassisted renters, Central/South Americans had a higher prevalence of short sleep than NH-Whites (PR=1.22 [95% CI:1.03–1.46]). Among short-sleepers, overweight (PR=1.16 [95% CI:1.04–1.29]) and obesity (PR=1.29 [95% CI:1.06–1.56]) were more prevalent among Central/South Americans compared to NH-White unassisted renters, although there were no racial/ethnic differences among government-assisted renters and homeowners. Mexican and Puerto Rican homeowners had higher prevalence of short sleep (PR=1.06 [95% CI:1.00–1.13] and PR=1.40 [95% CI:1.25–1.57]) compared to NH-Whites. Mexican and Puerto Rican homeowners had higher prevalence of diabetes (except Puerto Rican long-sleepers), overweight, and obesity than NH-White counterparts, regardless of sleep duration. Conclusion Sleep duration and cardiometabolic health disparities varied by rental/homeownership status and between Hispanic/Latino subgroups compared to NH-Whites. Future research should investigate differential housing environments and sleep disparities. Support (If Any) NIEHS, Z1AES103325-01. This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Grandner, M; Perlis, M; Parthasarathy, S; Pack, A

doi: 10.1093/sleep/zsy061.876pmid: N/A

Abstract Introduction Few real-world strategies exist for increasing sleep duration. As with other health behaviors, short sleep may be relatively intractable, as people give up sleep for increased work time, social activities, etc. Accordingly, a protocol was developed to evaluate whether an individualized, systematic upward titration of sleep opportunity would result in increased sleep duration. Methods The sleep extension intervention was conducted over 8 weeks and entailed (1) standardizing the sleep-wake schedule by assigning bed/wake times based on mean sleep ability determined by 2 weeks of sleep diary average total sleep time, and (2) increasing sleep opportunity by 15 minutes per week provided sleep efficiency remained at the standard CBTI cutoff of >90%. If sleep efficiency was moderate, no additional time was added for that week. If sleep efficiency was lower than 85%, sleep opportunity was reduced by 15 minutes for the week. Five men aged 26–51 (M=33.4) who self-reported ≤6 hours of habitual sleep and did not have other medical, psychiatric, or sleep disorders were recruited. Subjects were not recruited on the basis of motivation to increase sleep duration. Results Based on electronic daily sleep diaries, mean sleep duration at Week 1 was 5.8 hours (SD=1.2 hours) with a corresponding mean sleep efficiency of 86.8% (SD=7.8%). By the end of 8 weeks, these were increased to a mean 7.4 hours (SD=1.4 hours) with a sleep efficiency of 93.6 (SD=3.5%). Mean increase in sleep duration was 97.25 minutes per person (SD=23.9, range=57.7–118.0, effect size d=1.23). A paired t-test was significant (t(4)=-9.10,p=0.0008). Conclusion In summary, a “real-world,” data-driven protocol that slowly increases sleep opportunity can be effective in extending sleep duration. The cardio-metabolic and neurocognitive health benefits of such an effective intervention need to be explored. Support (If Any): This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Artinian, Hovig; Gandotra, Kamal; Strohl, Kingman

doi: 10.1093/sleep/zsy063.1119pmid: N/A

Abstract Introduction Case of mirtazapine treatment resulting in REM-behavior disorder, nightmares and somnambulism in a patient without Parkinsonism. Report of Case 67-year-old male Vietnam Veteran on treatment for major depressive disorder (MDD) and post-traumatic stress disorder (PTSD) was found to have mild OSA (AHI 13.9) with poor compliance on autoPAP therapy due to claustrophobia with the mask and frequent episodes of dream enactment. The patient reported a temporal relationship with his nocturnal disruptive behaviors and commencement of mirtazapine. The mirtazapine dose had recently been increased to 45mg which the patient associated with worsening somnambulism, RBD and nightmares. Additionally, worsening insomnia was associated with twice daily bupropion SR use. In consultation with the psychiatrist, we recommended bupropion XL once daily in the morning, which resulted in significant and swift improvement in insomnia, and initiated a mirtazapine taper, with goal to discontinue the medication. As the mirtazapine dose was reduced, the patient reported significant reduction in nocturnal disruptive behaviors, including dramatic improvement of RBD and significant reduction in nightmares and somnambulism. Additionally, low-dose clonazepam was initiated transiently to prevent withdrawal symptoms such as rebound anxiety and irritability from the alpha-2 antagonist taper, to decrease arousals and to improve CPAP compliance. Conclusion Medications used to treat mood disorders can result in significant sleep comorbidities. Identifying appropriate medications that can treat the mood and administering them at appropriate times can result in maximal benefit with minimal side effects to the patient. This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Low, D; Wu, M; Spira, A

doi: 10.1093/sleep/zsy061.1019pmid: N/A

Abstract Introduction Excessive and insufficient sleep have been associated with cognitive dysfunction. The sleep duration recommended for adults aged 65+ by the National Sleep Foundation is 7–8 hours, with 5, 6, or 9 hours considered potentially appropriate. We investigated the association between sleep duration and cognitive performance in adults aged 65+. Methods Participants were 901 adults aged 65+ from the National Health and Nutrition Examination Survey 2011–12 dataset, who had data on self-reported sleep duration, covariates, and tests of cognitive function: Consortium to Establish a Registry for Alzheimer’s Disease Word Learning (CERAD-WL immediate and delayed), Animal Fluency Test (AFT), and Digital Symbol Substitution Test (DSST). Sleep duration (hours) was categorized as: ≤4, 5–6, 7–8 (reference), 9, and ≥10. We excluded outliers of 2 or ≥12 hours. Regression analyses were performed with sample weights applied to obtain nationally representative results. Results The mean±SD age of the participants was 72.7 ± 0.2; 44.3% were men; 80.8% were non-Hispanic White; 57.1% received some college education; 3.3% slept for ≤4 hours, 24.2% for 5–6 hours, 61.7% for 7–8 hours, 7.9% for 9 hours, and 3.0% for ≥10 hours. In unadjusted analyses, compared to participants with 7–9 hours of sleep, those with ≥10 had poorer immediate CERAD-WL[β(95%CI) -3.57(-6.04,-1.10)] and DSST[β(95%CI) -11.11(-18.78,-3.44)]; those with ≤4 or ≥10 hours had poorer AFT[β(95%CI) -2.44(-4.16,-0.73)] and [β(95%CI) -4.23(-5.36,-3.10)], respectively; those with 5–6 hours had better delayed CERAD-WL[β(95%CI) 0.63(0.16,1.11)]; those with 9 or ≥10 hours had poorer delayed CERAD-WL[β(95%CI) -0.91(-1.71,-0.10)] and [β(95%CI) -1.44(-2.76,-0.11)], respectively. After adjusting for age, race/ethnicity, gender, and education, compared to participants with 7–9 hours of sleep, those with ≥10 hours had poorer immediate CERAD-WL[β(95%CI) -2.35(-4.47,-0.23)] and AFT[β(95%CI) -2.31(-3.71,-0.91)] performance; those with 5–6 hours had better delayed CERAD[β(95%CI) 0.61(0.20,1.02)]. There was no significant association with DSST. Conclusion In adults aged 65+, prolonged sleep duration is associated with poorer immediate verbal learning and semantic fluency performance. Shorter sleep duration is associated with better delayed verbal learning. Further prospective studies with objective sleep measures in representative cohorts are needed. Support (If Any) NIA R01AG050507, RF1AG050745, R01AG049872, U01AG052445, R01AG054771. NHANES 2011–12 https://wwwn.cdc.gov/nchs/nhanes/Search/DataPage.aspx?Component=Questionnaire&CycleBeginYear=2011. This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Wakschal, E; Boland, E; Gehrman, P; Haynes, P

doi: 10.1093/sleep/zsy061.934pmid: N/A

Abstract Introduction PTSD and major depressive disorder (MDD) are highly comorbid. It is well established that nightmares are a key feature of PTSD, though the role that depression plays in the expression of this symptom is less clear. The present study attempted to elucidate the relationship between depression, PTSD, nightmare frequency, and nightmare distress. Methods Data were obtained from 63 male veterans who met criteria for PTSD and MDD. Nightmare distress and frequency were measured through daily sleep diaries, and psychiatric symptoms were assessed using the Hamilton Rating Scale for Depression (HRSD) and Clinician Administered PTSD Scale (CAPS). Results When sleep items on the HRSD and CAPS were omitted, nightmare frequency was significantly associated with CAPS scores (p=.02) but not HRSD scores (p=.28), and nightmare distress was associated with HRSD scores at trend level (p=.07) but not significantly related to CAPS scores (p=.61). When sleep items were included in the HRSD and CAPS, the only change was that nightmare distress was significantly associated with HRSD scores (p= .03). Conclusion These findings suggest that increased PTSD severity contributes to greater number of nightmares, but not the level of distress associated with them. The fact that nightmare distress was associated with depression suggests that depression and insomnia (as assessed via the HRDS sleep items) may be implicated in the more debilitating aspects of nightmares in PTSD patients. Further research is necessary to determine the precise nature of this relationship. Support (If Any) Funding for this research was supported by Department of Defense #W81XWH-08-2-0121 (PI: Haynes) and American Sleep Medicine Foundation #37-CA-06 (PI: Haynes). Analysis and presentation of this research supported by: Department of Veterans Affairs IK2-CX001501 (PI: Boland). PDF This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

McCarty, T; Bourne, B; Wong, M; Travis, J K; Merritt, M P; Brown, M F; Krishna, J

doi: 10.1093/sleep/zsy061.756pmid: N/A

Abstract Introduction Standard management of sleep-related breathing disorders (SRBDs) in children requiring PAP includes the completion of a diagnostic polysomnography followed by full-night PAP titration (FNPT). However, the commonly used 2-step process in the pediatric context is costly, resource intensive, and may delay treatment initiation due to long wait times. While split-night PAP titration (SNPT) is frequently sufficient for treatment initiation in adult SRBD, there is currently limited data on the utility of SNPT in pediatric SRBD. This study examines the feasibility of SNPT in a tertiary-care pediatric hospital-based sleep laboratory and presents preliminary data from this project which ultimately seeks to identify patient characteristics associated with successful SNPTs in pediatric populations. Methods Demographic and pertinent sleep study data were gathered via retrospective sequential chart review for all completed FNPT and SNPT studies conducted between November of 2015 and November of 2017 as part of routine care. In-lab parent and patient education and mask fitting was performed prior to each titration. Using AASM guidelines, titration was classified “optimal”, “good”, or “adequate” by a board certified sleep physician, and deemed unsuccessful if classified “unacceptable”. A one-tailed t-test and two-tailed Fisher’s Exact test were used to assess group differences. Results Data from 166 PAP titration studies, including 132 FNPTs (55.8% male) and 34 SNPTs (70.6% male) were evaluated. Patient ages ranged from 1 to 21 years for FNPTs (mean 12.6 years) and 2 months to 20 years for SNPTs (mean 13.3 years). The average study time on pressures was not statistically significantly different (t(37) = 9.31, p = 1.54), with 472.6 minutes for FNPTs and 311.0 minutes for SNPTs. In the current sample, 124 (93.9%) of FNPTs and 29 (85.3%) of SNPTs were successful, revealing a non-significant relationship between type of PAP titration study and study outcome (p = 0.14). Conclusion We conclude SNPT is feasible and may be used in a subset of pediatric patients. Further examination of our data will be conducted to determine patient characteristics predictive of PAP study outcome. Support (If Any) None. This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Stewart, N; Walters, R; Mokhlesi, B; Arora, V

doi: 10.1093/sleep/zsy061.242pmid: N/A

Abstract Introduction Chronic obstructive pulmonary disease (COPD) is the 3rd leading cause of morbidity and mortality in the US, and results in significant cost to the health care system. Obstructive Sleep Apnea (OSA) is a common sleep disorder in 25% of US adults. “Overlap Syndrome” is COPD and OSA in the same patient. These patients have more frequent nocturnal oxygen desaturations, hypoxemia, and hypercapnea. Patients with overlap who do not wear continuous positive airway pressure (CPAP) have increased risk of death and rehospitalization. Many inpatients are at high risk for OSA. Patients at high risk of OSA have worse in-hospital sleep as measured by actigraphy. Poor sleep quality is independently associated with increased risk for acute COPD exacerbations. No studies characterize risk of undiagnosed sleep disorders among hospitalized patients with obstructive lung disease (COPD or asthma). Methods In this retrospective secondary analysis of prospective cohort study, medical inpatients 50 years of age were interviewed on admit, and their charts were reviewed to obtain demographic data and diagnosis. Initial sleep evaluation was based on survey questionnaires (Berlin, Pittsburg Sleep Quality Index, Epworth Sleepiness Scale, and Insomnia Severity Index) completed on day of admission, and examined by Cochran Mantel Haenszel (CMH) or Wilcoxon rank sum. Results A total of 585 patients (322 female and 415 African American) were enrolled. Roughly one-quarter (128, 22%) had a prior diagnosis of Obstructive Lung Disease (OLD). No association was found between OLD and risk of obstructive sleep apnea (RR 1.14 [95% CI 0.98, 1.33] p=0.109), risk of poor sleep quality (RR 1.06 [95% CI 0.90, 0.1.24] p=0.482), or risk of being excessively sleepy (RR 0.97 [95% CI 0.72, 1.32] p=0.867). A diagnosis of OLD carried a 88% higher odds (OR 1.88 [95% CI 1.43, 3.21] p=0.005) of having insomnia. Conclusion Among medical inpatients, having obstructive lung disease was independently associated with higher proportional odds of having insomnia, yet no association was found with risk of obstructive sleep apnea, risk of poor sleep quality, or risk of being excessively sleepy. Support (If Any) none This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Kumar, S; Hsieh, K; McGinty, D; Szymusiak, R

doi: 10.1093/sleep/zsy061.219pmid: N/A

Abstract Introduction CSR alters hypothalamic-pituitary-axis (HPA) axis function and can be viewed as a chronic mild stress. Homeostatic responses to sleep loss are present in rats during 1–2 days of CSR, but are attenuated or absent during subsequent days. We hypothesize that elevated corticotropin releasing factor (CRF) signaling in the brain contributes to the attenuation of sleep homeostasis during CSR. In this study we examined the effects of ICV infusion of CRF receptor-1 (CRF-R1) antagonist, in rats subjected to CSR. Methods CSR was achieved by a forced locomotion on a treadmill (speed=8cm/sec), with a 3/12 sec on/off sequence. This sequence was repeated for 3 hrs, followed by 1 hr sleep opportunity. This 3:1 hr schedule was applied continuously for five days. On CSR day 4, control rats (n=5) received ICV infusion of vehicle (1% DMSO in aCSF) while experimental rats received infusion of either 10 µg (n=7) or 1 µg (n=5) antalarmin (ANT) (0.2 µl/min over 3 hrs starting at ZT6). SWA in NREM sleep (% change from baseline) and time asleep during the 1 hr sleep opportunity at ZT9-10 were analyzed. Results Compared to baseline, EEG slow-wave activity (SWA) increased in all groups on CSR day 1 (Vehicle; 204 ± 27%, 1ug ANT; 209 ± 9% and 10ug ANT; 216 ± 11). SWA was reduced on CSR day 4 versus day 1 in Vehicle- (144 ± 10% vs 204 ± 27%) and 1 ug ANT- (153 ± 9% vs 209 ± 9% versus) treated rats. Infusion of 10ug ANT on day 4 partially restored the EEG SWA response (187 ± 20% versus 216 ± 11%). Total Sleep Time (TST) on CSR day 1 (in min) was Vehicle; 34.6 ± 3.1, 1ug ANT; 35.6 ± 3.7 and 10ug ANT; 36.7 ± 2.8. On CSR day 3 values were 27.2 ± 2.8, 27.7 ± 6.2, and 28.6 ± 2.7. Infusion of 10 ug ANT on CSR day 4 restored the TST response to sleep loss compared to Vehicle and 1 ug ANT infusion (36.9 ± 2.9 vs 28.7 ± 2.5 vs 27.9 ± 6.3). Conclusion Findings support the hypothesis that increased CRF signaling in the brain contributes to the suppression of sleep homeostasis during CSR. Support (If Any) VA Merit Award BX00155605 and BX003520. This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Javed, F; Armitstead, J; Benjafield, A; Blase, A; O’Connor, C; Aurora, R; Punjabi, N

doi: 10.1093/sleep/zsy061.467pmid: N/A

Abstract Introduction SDB is highly prevalent in heart failure (HF) patients. While LVEF is a well-known predictor of morbidity and mortality in HF, it remains to be determined whether characteristics of disordered breathing have any additional predictive value. Thus, the overall objective of this study was to determine whether parameters derived from nasal airflow and oximetry such as lung-to-peripheral circulation-time (LPCT) and time-to-peak flow (TTPF) were associated with death or cardiovascular-hospitalization independent of LVEF. Methods The CAT-HF study was a multi-center study (N=126) on the cardiovascular effects of adaptive-servo ventilation (ASV) in patients admitted for decompensated HF. The current analysis is based on the subsample of the randomized cohort with evidence of Cheyne-Stokes respiration (CSR). Patients were followed until death or cardiovascular-related hospitalization - the composite endpoint for the current analysis. Multivariate logistic regression analyses were used to examine whether the median values of LPCT and TTPF derived using an automated algorithm (ResCSRF) were associated with the composite endpoint after adjusting for LVEF. Results The study sample consisted of 78 patients with CSR (34 in the optimal medical treatment (OMT)-arm and 44 in the OMT-ASV arm). In total, 38 deaths or cardiovascular-hospitalizations were observed [OMT(18) and OMT-ASV(20); p=0.51]. Median LPCT and TTPF were highly correlated (r=0.83; 95% CI: 0.74–0.89) and each was associated with the composite endpoint independent of LVEF. However, multivariate models that included LVEF as a covariate showed that TTPF, but not LPCT, was independently associated with the composite endpoint. The multivariate-adjusted odds ratio for the composite outcome of death or cardiovascular-hospitalization for the four quartiles of TTPF were 1.00 (quartile 1: < 12.7 secs), 3.75 (95% CI: 0.86–16.36; quartile 2: 12.8–15.2 secs); 5.61 (95% CI: 1.01–31.18; quartile 3: 15.3–17.9 secs) and 14.19 (95% CI: 1.56–128.97; quartile 4: ≥18.0 secs). Conclusion In patients with decompensated HF and CSR, TTPF, but not LPCT, derived from a type 3 portable sleep test is predictive of death or cardiovascular-hospitalization independent of LVEF. Support (If Any) This study was supported by ResMed. This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]

Tuan, L; Lee, L

doi: 10.1093/sleep/zsy061.033pmid: N/A

Abstract Introduction Adolescent is a critical periods for neural circuit development. During this period, substantial neurological changes, including synaptic remodeling, are taking place in various brain regions. It is known that the synaptic organization is modulated by sleep and sleep loss has been shown to disrupt synaptic pruning particularly in young animals. The roles of microglia in synaptic pruning in the healthy developing brain have drawn much attention for their close interaction with synapses. They are also sensitive to the pro-inflammatory condition following sleep loss. We therefore hypothesized a role of microglia in sleep deprivation (SD)-induced synaptic changes in adolescent mice. Methods Male C57/BL6 mice of 5 weeks and 10 to 12 weeks old were used in this study to represent the periods of adolescence and adulthood, respectively. Subsequent to 72h of SD, which was conducted using modified multiple platform method, mice of SD and control groups were subjected to neurochemical and histological examinations. Results Compared with that in normal sleep controls, the density of dendritic spines in the granule cells of the dentate gyrus was increased after SD in adolescent but not in adult mice. The width of the spine heads was reduced in adolescent after SD but appeared normal in adult mice of SD group. Morphometric analyses revealed significant morphological changes in microglia after SD in adolescent but not in adult mice. Furthermore, there was a reduction in engulfment of postsynaptic density protein 95 within the microglia after SD in adolescents. Quantification of mRNA expression by real-time PCR showed CX3C chemokine receptor 1 and integrin alpha M, closely associated with the synaptic pruning via phagocytosis function in microglia, were also downregulated after SD in adolescent mice. Conclusion We provided evidence to support our hypothesis that microglia-mediated phagocytosis in synaptic pruning is impaired during SD in adolescent mice. Given synaptic downscaling during sleep is beneficial to the consequential learning and memory capacity, sufficient sleep for the adolescence is indispensable. Support (If Any) No This content is only available as a PDF. © Sleep Research Society 2018. 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/about_us/legal/notices) © Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: [email protected]