Abstract Study Objectives This study used ambulatory polysomnography (PSG) to investigate post-traumatic nightmares of post-traumatic stress disorder (PTSD). The key research question was whether post-traumatic nightmares occur in both rapid eye movement (REM) and non-REM sleep, and if so, whether nightmares in each sleep stage differed in content, phenomenology, and heart rate response. Underlying sleep disorders were investigated in an exploratory way. Methods Thirty-five treatment-seeking veterans, current serving military members, and emergency service personnel undertook full PSG using the Compumedics (Melbourne, Australia) SomtePSG V1 system, during an inpatient psychiatric admission. The PSG recording included an event button to be pressed when a nightmare occurred, allowing us to determine the stage of sleep, changes in heart rate, and associated sleep events. The content and phenomenological features of participants’ nightmares were recorded. Results Of the 35 participants, 29 reported a nightmare during their sleep study, but only 21 pressed the event button and could recall the content of one or more nightmare. This yielded sleep and nightmare data for 24 nightmares. Of the 24, 10 nightmares arose from REM sleep and 14 from non-REM (stages N1 and N2). Seven were accurate trauma replays and 17 were non-replay or a mixture of replay and non-replay. Most nightmares were associated with respiratory or leg movement events and increase in heart rate on awakening. Conclusions Post-traumatic nightmares of PTSD occur in both REM and non-REM sleep and are commonly associated with other sleep disturbances. These findings have important treatment implications. post-traumatic nightmares, PTSD, REM and non-REM sleep, obstructive sleep apnea, ambulatory polysomnography Statement of Significance Researchers have had limited success in investigating post-traumatic nightmares in sleep studies as they tend not to occur in the sleep laboratory. The use of ambulatory polysomnography in this study has led to two important findings. First, we found that post-traumatic nightmares, unlike normal dreams, occur in both rapid eye movement and non-rapid eye movement sleep, suggesting that post-traumatic nightmares might not be a single phenomenon but a range of different phenomena. Second, we found that post-traumatic nightmares frequently occur in the context of other sleep disturbances, with most nightmare awakenings associated with a respiratory event and/or leg movements. This highlights the need to diagnose and treat underlying sleep disorders as part of the approach to treatment of post-traumatic nightmares. INTRODUCTION Post-traumatic nightmares are a highly prevalent and distressing symptom of post-traumatic stress disorder (PTSD), which affect not only the quality of sleep but also daytime functioning. Nightmares and related sleep disturbances are considered if not the hallmark of PTSD,1 at least a core symptom2 or an underlying mechanism.3 Of particular importance, some researchers argue that disrupted sleep and trauma-related dreams even have a role in the development of PTSD4,5 and increase PSTD symptom severity, distress, and associated functional impairment.6 The nightmares of PTSD are classified as an intrusive symptom, arising from sensory memory of the traumatic event.7 Their occurrence during sleep, however, gives rise to the possibility that their nature is also influenced by sleep and normal dream–related factors.8 With respect to sleep factors, an important finding of previous research is that increased rates of sleep disordered breathing have been found to be associated with PTSD, with comorbidity between the two conditions reported to be between 50% and 90%.9 Krakow and colleagues10 propose a bi-directional relationship in which PTSD-related insomnia and nightmares lead to sleep fragmentation which increases upper airway collapsibility and sleep breathing events, which in turn increase sleep fragmentation, exacerbating insomnia, and nightmares. In addition, Mellman and colleagues11 report a range of parasomnia-like events in veterans with PTSD, most commonly nightmares (combat and noncombat), thrashing movements during sleep, and panic awakenings with no dream recall. Fifty percent of awakenings observed in the sleep laboratory were preceded by rapid eye movement (REM) sleep, 44% by stage 2 sleep, and 6% by stage 1 sleep. With respect to dream-related factors, differences in the form and content of dream mentation across stages of sleep12 may lead to the phenomenology of post-traumatic nightmares of PTSD varying according to the stage of sleep in which they occur. Previous research has outlined the potential differences between dreams and nightmares arising from REM versus non-REM (NREM) sleep.13 Post-traumatic nightmares arising from REM sleep, like normal dreams, are likely to have bizarre or distorted content and to be associated with a gradual increase in anxiety before awakening, with heart and respiration returning to normal on awakening. On the other hand, post-traumatic nightmares arising from NREM sleep are likely to be more realistic, related to real life experiences and associated with a sudden onset of intense anxiety with increased heart rate, respiration, and body movement. This is consistent with the way in which PTSD intrusions have been characterized as memory intrusions accompanied by the same psychological and physical reactions experienced at the time of trauma.8 In previous research into self-reported phenomenology of post-traumatic nightmares, Phelps and colleagues14,15 found that post-traumatic nightmares differ in the extent to which they involve actual memory intrusions or contain trauma-related but bizarre and imaginative content. The possibility that the post-traumatic nightmares of PTSD vary depending upon the stage of sleep in which they occur was first raised by van der Kolk and colleagues.16 Although nightmares are generally associated with REM sleep American Academy of Sleep Medicine,17 a small number of studies have recorded post-traumatic nightmares in both REM and non-REM sleep. In early research (prior to the diagnosis of PTSD), Fisher and colleagues18 reported that stage 4 nightmares (termed arousal reaction nightmares, including the nightmare of traumatic neurosis in which content is recalled) and REM nightmares were found to co-exist in some individuals following trauma. Schlossberg and Benjamin19 reported nightmares in stage 2 sleep in three veterans with combat fatigue. Kramer, Schoen, and Kinney20 investigated the dreams of eight veterans with PTSD in the laboratory setting and found that disturbing dreams with military references occurred on spontaneous awakenings from both REM and non-REM sleep. Similarly, van der Kolk et al.16 observed PTSD nightmares in both REM and stage 2 sleep in two participants. Unfortunately, any differences in dreams arising from REM and non-REM sleep were not reported in these studies. Unfortunately, advances in our knowledge of potential phenomenological differences in post-traumatic nightmares that occur in REM and non-REM sleep have been hindered by the fact that post-traumatic nightmares rarely occur in the sleep laboratory.21 We sought to overcome this difficulty in the current study by using ambulatory polysomnography, thus eliminating the impact of the potential protective sleep laboratory environment.22 Aim The primary aim of the current research was to investigate in what stage of sleep the nightmares of PTSD occur. The secondary aim was to investigate any differences in post-traumatic nightmares arising from REM or NREM sleep with respect to content, phenomenology, and their temporal relationship with increases in heart rate. The influence of other sleep disorders on dreams was investigated in an exploratory way. Hypotheses 1. Post-traumatic nightmares will occur across REM and NREM sleep. 2. Post-traumatic nightmares that arise from REM sleep will have non-replay content while those that arise from non-REM sleep will involve accurate replay of all or part of an actual traumatic event. 3. Post-traumatic nightmares that arise from REM sleep will be associated with a gradual increase in heart rate prior to waking while those arising from NREM will be associated with a sudden increase in heart rate on waking. METHODS Participants Veterans, current serving members of the Australian Defence Force (ADF) and emergency service personnel, were recruited from the inpatient unit of the Psychological Trauma and Recovery Service (PTRS) at Heidelberg Repatriation Hospital in Melbourne, Australia. Inclusion criteria included a diagnosis of PTSD, post-traumatic nightmares on a daily or almost daily basis, and being aged 18 years or older. Exclusion criteria included current psychosis, current suicidality, moderate-to-severe brain injury, current substance intoxication or withdrawal, and being unable or unwilling to abstain from alcohol or other nonprescribed drugs on the sleep study night(s). Measures The Clinician Administered PTSD Scale for DSM 5 CAPS-523 was used to confirm the diagnosis of PTSD. The CAPS-5 is a 30-item structured interview that measures DSM-5 PTSD symptoms in the past month. All participants underwent unattended full polysomnography (PSG) using the Compumedics (Melbourne, Australia) SomtePSG V1 system following the Australasian Sleep Association and Thoracic Society of Australia and New Zealand guidelines.24 The following parameters were recorded: electroencephalography (EEG) with electrode placement at F4/M1, C4/M1, and O2/M1; electro-oculography (EOG) at E1/M2 and E2/M2; submental electromyography (EMG); electrocardiography (ECG); nasal pressure; nasal airflow temperature with a thermocouple; thorax and abdominal movements with respiratory inductive plethysmography bands; oxygen saturation with a pulse oximeter on the finger; body position; snoring with a microphone on the upper chest; and leg movements with piezoelectric sensors. The sleep recording measured sleep stages, patterns of arousal and awakening, and the presence of any other sleep disorders such as sleep apnea. An event button, activated by the participant upon waking from a nightmare, was linked to the sleep record. Participants completed a sleep diary reporting the time they went to bed, time of lights out, how long it took to get to sleep, time of waking in the morning, and how long they slept for. A post-traumatic nightmare questionnaire used in previous research14,15 was administered to those who experienced a nightmare during their sleep study. The questionnaire involved rating the following features of the nightmare using a Likert scale of 0 = not at all; 1 = a little bit; 2 = moderately; 3 = quite a bit; 4 = extremely or exactly: (1) How similar is the content of the dream to waking memory of the traumatic event? (2) How disturbing is the dream? (3) How realistic is the dream (events could happen)? (4) How vivid are the images (sharp, clear, detailed)? (5) How vivid are other sensory details (e.g., sounds, smells, taste, and bodily sensations such as pain)? (6) Did you “act out” the dream in any way (e.g., kicking, punching, running, or screaming)? (7) Did you have strong physical sensations (e.g., heart rate, sweating, and trembling)? (8) Did you feel like you were reliving the incident, like it was happening again? They were also asked to rate to what extent the nightmare seemed like a thought, a feeling or a sensory experience on the same scale. Finally, participants were asked to nominate whether the nightmare was an accurate replay of part or all of a traumatic event, non-replay, or contained a mix of replay and non-replay elements. In addition, a range of self-report measures were used to assess the severity of PTSD and common comorbid mental health problems. The PTSD Checklist for DSM-5 (PCL-5)25 was used to assess the severity of PTSD symptoms. The PCL-5 lists the 20 symptoms of PTSD and the respondent is asked to rate how much they have been bothered by each symptom over the past month on a 5-point Likert scale from 1 to 5, yielding a total score between 20 and 100 with higher scores indicating more severe symptoms. The Alcohol Use Disorders Identification Test (AUDIT)26 is a brief screening tool that consists of 10 items measuring alcohol consumption, dependence symptoms, and the personal and social harm caused by drinking. Items are scored on a 5-point Likert scale of 0 to 4, yielding total scores of 0 to 40 with higher scores indicating more severe symptoms. A cutoff of 8 is generally accepted as identifying individuals with potentially problematic alcohol use. The Hospital Anxiety and Depression Scale (HADS)27 is divided into two subscales to assess both anxiety and depression. Each scale consists of seven items, scored on a 4-point Likert scale from 0 to 3 yielding total scores for each subscale from 0 to 21 with higher scores indicating more severe symptoms. Procedure The study was approved by the Austin Health Human Research Ethics Committee. Potential participants were referred by their treating psychiatrist or psychiatry registrar and screened against inclusion and exclusion criteria. Potential participants were then approached, the study was explained, and they gave written informed consent. Following this, participants were assessed using the CAPS-5 to confirm the diagnosis of PTSD and were then asked to complete the PCL-5 and AUDIT self-report measures. Referral was made to the Institute of Breathing and Sleep (IBAS) at the Austin Hospital for a sleep study to be undertaken within the following 2 to 3 days. On the night of their sleep study, participants attended the sleep laboratory at 4:30 pm to have the recording electrodes attached. They completed the HADS at this time. They then returned to the PTRS inpatient unit and undertook usual activities including sleeping. Although participants were in an inpatient unit, they were not sleeping under surveillance as they would be in the sleep laboratory. During the night, when participants woke from a nightmare, they were instructed to press the event button and to jot down the main theme or content of the nightmare as an aide memoire. The following morning, participants completed the sleep diary and the nightmare questionnaire. Data Coding and Analysis All studies were staged and scored using the 2016 American Academy of Sleep Medicine Manual for the Scoring of Sleep and Associated Events. When the event button was pressed to indicate that the participant has woken from a nightmare, the sleep recording was used to determine whether the preceding sleep was REM sleep, non-REM sleep, or wakefulness. The sleep recording was also used to determine the temporal relationship between changes in heart rate and dream awakening. Participants’ reports of nightmare phenomenology were categorized as replay versus non-replay or mixed. The relationships between stage of sleep (REM vs non-REM) and nightmare type (replay vs non-replay or mixed), and between sleep apnea and stage of sleep and nightmare type were tested using chi-square analysis. RESULTS Participants Thirty-five participants were recruited into the study, comprising 24 veterans, seven current serving ADF personnel, two current serving police officers, one former police officer, and one current serving firefighter. The average age of participants was 43.32 years (SD = 10.73; range = 31–72). Almost all were male, with a single female current serving ADF member. The majority were married (n = 15), a further eight were in a de facto relationship, eight were separated or divorced, and four were single. With respect to employment status, 11 were employed, 11 retired, seven were on a Department of Veterans’ Affairs pension, and six were unemployed. Almost all participants reported their ethnic background to be Caucasian, with a single participant identifying as Aboriginal or Torres Strait Islander. PTSD and Common Comorbid Mental Health Problems All participants met DSM-5 diagnostic criteria for PTSD according to the CAPS-5. Their mean CAPS-5 score was 46.06 (SD = 8.17, range = 33–72). Scores on self-report measures of PTSD and common comorbid mental health problems were as follows: the PCL-5 mean score was 53.45 (SD = 9.64, range 35–73), well above the suggested diagnostic cutoff 33; the AUDIT mean score was 17.06 (SD = 11.36, range 0–38), falling into the high risk or harmful category; the HADS anxiety mean score was 14.32 (SD = 2.86, range 10–21), indicating moderate anxiety; and the HADS depression mean score was 12.38 (SD 3.89, range 7–20), indicating moderate depression. Participants were taking a range of psychotropic medications, including some with known effects on sleep and nightmares, notably hypnotics (n = 15), benzodiazepines (n = 16), SSRI antidepressants (n = 12), SNRI antidepressants (n = 15), and prazosin (n = 13). Only three participants were taking no psychotropic medication, and the majority (n = 26) were taking more than one. Of particular note, all but six participants were taking antidepressant medication which have known REM-suppressing effects. Sleep There was a high degree of concordance between objective and subjective measures of sleep latency and total sleep duration. Subjective reports of time taken to get to sleep ranged from 5 to 150 minutes with an average estimate of 35.08 minutes (SD = 38.82). Objective reports ranged from 0 to 160.00 minutes with average sleep latency of 31.60 minutes (SD = 34.77). A Bland–Altmann plot was generated to investigate the level of agreement between the objective and subjective measures of sleep latency. Figure 1 is presented in Supplementary Material. A one-sample t test revealed no significant difference between the means of two measures, mean difference = −1.05, SD = 38.43, t = −0.205, p = .838, ns with 95% confidence limits of 74.28 (upper limit) and −76.39 (lower limit). Subjective reports ranged from 120.00 to 570.00 minutes with an average reported total sleep time of 330.00 minutes (SD = 105.60). Objective reports, based on the number of epochs staged as sleep multiplied by 30 seconds, ranged from 139.00 to 546.50 minutes with an average of 350.28 minutes (SD = 112.24). A Bland–Altmann plot was generated to investigate the level of agreement between the objective and subjective measures of total sleep time. Figure 2 is presented in Supplementary Material. A one-sample t test revealed no significant difference between the means of two measures, mean difference = 14.13, SD = 70.28, t = 1.31, p =.198, ns with 95% confidence limits of 152.89 (upper limit) and −124.63 (lower limit). With respect to other sleep parameters, the findings were as follows: total time in bed based on the times that the participant reported turning lights off and turning them on at the end of the study, ranged from 165.00 to 615.00 minutes, mean = 447.07 (SD = 106.43); wake after sleep onset, calculated as minutes of wakefulness stage from the first onset of sleep to the last epoch of sleep, ranged from 3 to 268.50 minutes, mean = 56.51 (SD = 53.21); early morning awakenings based on the number of times the patient woke between the first onset of sleep until the last epoch of sleep ranged from 1 to 75 times, mean = 29.43 (SD = 19.22); and sleep efficiency, the percentage of time asleep during the total time in bed ranged from 33.50% to 92.60%, mean = 76.76% (SD = 15.88). Sleep studies revealed greater than normal REM latency although again, there was considerable variability between participants. REM latency ranged from 57.50 to 485.50 minutes with an average of 213.40 minutes (SD = 91.38), considerably greater than normal REM latency of 70–90 minutes. The proportion of sleep time spent in REM was generally reduced with 23 (66%) of study participants recording less than normal 20% REM. Fifteen participants (43%) had 10% REM or less and four participants recorded between 0 and 2.5 minutes of REM. The potential impact of REM suppressing medications on these results needs to be acknowledged. The presence and severity of OSA was determined by the apnea–hypopnea index (AHI), the number of apnea and hypopnea events per hour of sleep, with AHI < 5 per hour = absent/minimal OSA, AHI ≥ 5 but <15 = mild OSA, AHI ≥ 15 but <30 = moderate, and AHI ≥ 30 = severe OSA. Twenty of the 35 participants had OSA. The severity was mild in 12 cases, moderate in four cases, and severe in four cases. Post-traumatic Nightmares Of the 35 participants, 29 reported a nightmare during their sleep study. Twenty-one reported a single nightmare, six reported two nightmares, and two reported three nightmares, yielding a total of 39 nightmare reports. However, only 24 participants reliably pressed the event button when they had a nightmare allowing us to identify stage of sleep and associated sleep events, and of this group, only 21 recalled the content of one or more nightmares. As a result, we captured 24 nightmares with complete data—stage of sleep, associated sleep events, and nightmare content. Seven of the nightmares captured in the study were characterized as accurate trauma replays and 17 were characterized as non-replay, including eight with a mixture of replay and non-replay content. As shown in Table 1, replay, mixed, and non-replay nightmares differed in the expected direction on ratings of similarity to memory of the event, sense of reliving, and realism. There were large effect size (ES) differences for each of these variables using η2 (in which small ES = 0.01, medium ES = 0.06, large ES = 0.14), of 0.73, 0.67, and 0.41, respectively. The dream types also differed on vividness of images and strength of physical sensations, also with large ES differences of 0.32 and 0.29, respectively. However, there was no significant difference between dream types on how disturbing the nightmares were, and all were rated as more like a sensory experience (mean = 2.14) than a thought (mean = 0.70) or a feeling (mean = 1.10). Table 1 Phenomenological Features by Dream Type. Dream feature Dream type N Mean SD df F p-Value How similar? Replay 7 3.71 0.49 2.21 28.99 .000* Mixed 8 1.75 1.03 Non-replay 9 0.78 0.67 Total 24 1.96 1.43 How disturbing? Replay 7 3.43 0.53 2.21 2.94 .075 Mixed 8 3.13 0.99 Non-replay 9 2.44 0.88 Total 24 2.96 0.91 How realistic? Replay 7 4.00 0.00 2.21 7.23 .004* Mixed 8 2.38 1.41 Non-replay 9 1.67 1.50 Total 24 2.58 1.53 How vivid are images? Replay 7 3.86 0.38 2.21 4.95 .017* Mixed 8 3.75 0.46 Non-replay 9 2.56 1.42 Total 24 3.33 1.09 How vivid are other sensory details? Replay 6 2.83 1.60 2.20 2.39 .117 Mixed 8 2.50 1.41 Non-replay 9 1.33 1.32 Total 23 2.13 1.52 “Acting out” dream storyline Replay 7 1.00 1.73 2.20 0.30 .744 Mixed 7 0.86 1.46 Non-replay 9 0.44 1.33 Total 23 0.74 1.45 Strong physical sensations Replay 7 3.57 0.53 2.20 4.20 .030* Mixed 7 1.71 1.89 Non-replay 9 1.67 1.50 Total 23 2.26 1.63 Sense of reliving Replay 7 3.71 0.49 2.21 21.82 .000* Mixed 8 1.88 1.88 Non-replay 9 0.00 0.00 Total 24 1.71 1.88 Seemed like a thought Replay 6 0.83 1.60 2.17 0.41 .668 Mixed 6 1.00 1.67 Non-replay 8 0.38 0.74 TOTAL 20 0.70 1.30 Seemed like a feeling Replay 6 0.83 1.60 2.17 0.54 .590 Mixed 6 1.67 1.86 Non-replay 8 0.88 1.36 Total 20 1.10 1.55 Seemed like a sensory experience Replay 7 2.29 2.14 2.18 0.28 .758 Mixed 6 2.50 1.97 Non-replay 8 1.75 1.75 Total 21 2.14 1.88 Dream feature Dream type N Mean SD df F p-Value How similar? Replay 7 3.71 0.49 2.21 28.99 .000* Mixed 8 1.75 1.03 Non-replay 9 0.78 0.67 Total 24 1.96 1.43 How disturbing? Replay 7 3.43 0.53 2.21 2.94 .075 Mixed 8 3.13 0.99 Non-replay 9 2.44 0.88 Total 24 2.96 0.91 How realistic? Replay 7 4.00 0.00 2.21 7.23 .004* Mixed 8 2.38 1.41 Non-replay 9 1.67 1.50 Total 24 2.58 1.53 How vivid are images? Replay 7 3.86 0.38 2.21 4.95 .017* Mixed 8 3.75 0.46 Non-replay 9 2.56 1.42 Total 24 3.33 1.09 How vivid are other sensory details? Replay 6 2.83 1.60 2.20 2.39 .117 Mixed 8 2.50 1.41 Non-replay 9 1.33 1.32 Total 23 2.13 1.52 “Acting out” dream storyline Replay 7 1.00 1.73 2.20 0.30 .744 Mixed 7 0.86 1.46 Non-replay 9 0.44 1.33 Total 23 0.74 1.45 Strong physical sensations Replay 7 3.57 0.53 2.20 4.20 .030* Mixed 7 1.71 1.89 Non-replay 9 1.67 1.50 Total 23 2.26 1.63 Sense of reliving Replay 7 3.71 0.49 2.21 21.82 .000* Mixed 8 1.88 1.88 Non-replay 9 0.00 0.00 Total 24 1.71 1.88 Seemed like a thought Replay 6 0.83 1.60 2.17 0.41 .668 Mixed 6 1.00 1.67 Non-replay 8 0.38 0.74 TOTAL 20 0.70 1.30 Seemed like a feeling Replay 6 0.83 1.60 2.17 0.54 .590 Mixed 6 1.67 1.86 Non-replay 8 0.88 1.36 Total 20 1.10 1.55 Seemed like a sensory experience Replay 7 2.29 2.14 2.18 0.28 .758 Mixed 6 2.50 1.97 Non-replay 8 1.75 1.75 Total 21 2.14 1.88 * p < .05. View Large Summary sleep report and nightmare data for these 21 participants are presented in Table 2. The findings in relation to the study hypotheses are described below. Table 2 Sleep Factors Associated with Post-traumatic Nightmares. ID Age BMI Total sleep time (min) Total REM (min/%) REM latency (min) Waking with NM Sleep stage NM typea HR changes associated with nightmare Proximal sleep events Sleep disorderb 1 56 39.1 149.0 22.5 (15%) 124.5 03:45 REM 2 No change Nil OSA = 1 PLMs associated with respiratory events 2 37 29.3 140.5 1 (1%) 157.5 02:15 N1 1 Decrease by 10 bpm Awakening followed hypopnea OSA = 1 No PLMs 3 40 34.2 240.5 2.5 (1%) 266.0 03:40 REM 3 Increase by 52bpm Nil OSA=0 No PLMs 4 40 35.7 544.0 72.5 (13%) 321.5 01.56 REM 3 No change Awakening associated with hypopnea, preceding REM-featured hypopneas, sustained muscle activity and PLMs OSA = 0 5 38 24.4 366.0 38.0 (10%) 196.5 03:11 N2 2 Increase by 6 bpm Awakening followed RERA and sleep featuring limb movements OSA = 0 No diff. in EMG between REM and non-REM 6 45 38.5 238.5 18.0 (8%) 485.5 00.04 N1 1 Increase by 10 bpm Sleep-featured PLMs OSA = 1 7 65 30.4 139.5 33.5 (24%) 102.0 02.15 Fragmented N1/N2 1 Increase by 50 bpm Awakening followed hypopnea OSA = 2 No diff. in EMG between REM and non-REM 04.22 REM 3 Increase by 16bpm PLMs 8 46 33.2 242.0 0.0 — 04.52 N2 1 Increase by 40 bpm Nil No REM sleep 9 45 30.6 461.0 74 (16%) 157.0 00.26 Fragmented N1/N2 1 Increase by 10 bpm Sleep-featured cyclical hypopneas, awakening followed hypopnea OSA = 3 03.30 Fragmented N1/N2 1 Increase by 12bpm Sleep featured cyclical hypopnoeas and limb movements 10 57 27.2 405.5 51.5 (13%) 301.5 03.14 REM 2 Increase by 6 bpm Awakening followed hypopnea. Sleep-featured PLMs and sustained muscle activity OSA = 3 Leg movements associated with respiratory events 11 32 24 423.5 63.0 (15%) 120.00 22.24 N1/N2 2 Increase by 16 bpm Nil OSA = 0 No PLMs 04.21 N2 2 Increase by 44 bpm Awakening followed hypopnea 12 36 34.1 483.0 112.5 (23%) 137.0 05.21 REM 3 Increase by 16 bpm Sleep-featured periods of PLMs and isolated respiratory events OSA = 0 13 32 28.3 389.5 67.5 (17%) 198.5 02.09 N2 2 Increase by 22 bpm Sleep-featured PLMs OSA = 0 14 43 29.6 147.5 19.0 (13%) 161.0 01.19 REM 3 Increase by 32 bpm Awakening followed RERA respiratory event, sleep-featured PLMs, and sustained muscle activity OSA = 0 PLMs 15 48 25.2 427.5 122.5 (29%) 292.0 08.12 Fragmented NREM 2 Increase by 20bpm Sleep-featured non-PLMs OSA = 1 16 56 26.9 319.5 32.5 (10%) 182.5 05.48 REM 2 Increase by 10 bpm Sleep-featured PLMs and sustained muscle activity OSA = 3 17 43 35.4 399.0 42.0 (11%) 116.5 04.10 REM 1 Increase by 34 bpm Awakening followed hypopnea, sleep-featured respiratory events, and isolated leg movements OSA = 2 Leg movements associated with respiratory events 18 59 31.7 396.5 19.5 (5%) 305.0 06.56 Fragmented N1 2 Increase by 14 bpm Sleep-featured hypopneas and non-PLMs OSA = 1 19 33 34.9 273.0 1.5 (1%) 185.5 03.18 N1/N2 3 Increase by 26 bpm Nil OSA = 1 20 38 NA 434.5 90.0 (21%) 142.0 04.10 REM 3 Increase by 20 bpm Sleep-featured hypopneas and limb movements OSA = 2 21 37 32.2 234.0 25.5 (11%) 106.0 23.58 N1/N2 3 Increase by 26 bpm Awakening followed limb movements, sleep-featured snoring OSA = 0 ID Age BMI Total sleep time (min) Total REM (min/%) REM latency (min) Waking with NM Sleep stage NM typea HR changes associated with nightmare Proximal sleep events Sleep disorderb 1 56 39.1 149.0 22.5 (15%) 124.5 03:45 REM 2 No change Nil OSA = 1 PLMs associated with respiratory events 2 37 29.3 140.5 1 (1%) 157.5 02:15 N1 1 Decrease by 10 bpm Awakening followed hypopnea OSA = 1 No PLMs 3 40 34.2 240.5 2.5 (1%) 266.0 03:40 REM 3 Increase by 52bpm Nil OSA=0 No PLMs 4 40 35.7 544.0 72.5 (13%) 321.5 01.56 REM 3 No change Awakening associated with hypopnea, preceding REM-featured hypopneas, sustained muscle activity and PLMs OSA = 0 5 38 24.4 366.0 38.0 (10%) 196.5 03:11 N2 2 Increase by 6 bpm Awakening followed RERA and sleep featuring limb movements OSA = 0 No diff. in EMG between REM and non-REM 6 45 38.5 238.5 18.0 (8%) 485.5 00.04 N1 1 Increase by 10 bpm Sleep-featured PLMs OSA = 1 7 65 30.4 139.5 33.5 (24%) 102.0 02.15 Fragmented N1/N2 1 Increase by 50 bpm Awakening followed hypopnea OSA = 2 No diff. in EMG between REM and non-REM 04.22 REM 3 Increase by 16bpm PLMs 8 46 33.2 242.0 0.0 — 04.52 N2 1 Increase by 40 bpm Nil No REM sleep 9 45 30.6 461.0 74 (16%) 157.0 00.26 Fragmented N1/N2 1 Increase by 10 bpm Sleep-featured cyclical hypopneas, awakening followed hypopnea OSA = 3 03.30 Fragmented N1/N2 1 Increase by 12bpm Sleep featured cyclical hypopnoeas and limb movements 10 57 27.2 405.5 51.5 (13%) 301.5 03.14 REM 2 Increase by 6 bpm Awakening followed hypopnea. Sleep-featured PLMs and sustained muscle activity OSA = 3 Leg movements associated with respiratory events 11 32 24 423.5 63.0 (15%) 120.00 22.24 N1/N2 2 Increase by 16 bpm Nil OSA = 0 No PLMs 04.21 N2 2 Increase by 44 bpm Awakening followed hypopnea 12 36 34.1 483.0 112.5 (23%) 137.0 05.21 REM 3 Increase by 16 bpm Sleep-featured periods of PLMs and isolated respiratory events OSA = 0 13 32 28.3 389.5 67.5 (17%) 198.5 02.09 N2 2 Increase by 22 bpm Sleep-featured PLMs OSA = 0 14 43 29.6 147.5 19.0 (13%) 161.0 01.19 REM 3 Increase by 32 bpm Awakening followed RERA respiratory event, sleep-featured PLMs, and sustained muscle activity OSA = 0 PLMs 15 48 25.2 427.5 122.5 (29%) 292.0 08.12 Fragmented NREM 2 Increase by 20bpm Sleep-featured non-PLMs OSA = 1 16 56 26.9 319.5 32.5 (10%) 182.5 05.48 REM 2 Increase by 10 bpm Sleep-featured PLMs and sustained muscle activity OSA = 3 17 43 35.4 399.0 42.0 (11%) 116.5 04.10 REM 1 Increase by 34 bpm Awakening followed hypopnea, sleep-featured respiratory events, and isolated leg movements OSA = 2 Leg movements associated with respiratory events 18 59 31.7 396.5 19.5 (5%) 305.0 06.56 Fragmented N1 2 Increase by 14 bpm Sleep-featured hypopneas and non-PLMs OSA = 1 19 33 34.9 273.0 1.5 (1%) 185.5 03.18 N1/N2 3 Increase by 26 bpm Nil OSA = 1 20 38 NA 434.5 90.0 (21%) 142.0 04.10 REM 3 Increase by 20 bpm Sleep-featured hypopneas and limb movements OSA = 2 21 37 32.2 234.0 25.5 (11%) 106.0 23.58 N1/N2 3 Increase by 26 bpm Awakening followed limb movements, sleep-featured snoring OSA = 0 aDream type: 1 = replay; 2 = symbolic; 3 = mixed. bOSA: 0 = none/minimal: AHI < 5 per hour; 1 = mild: AHI ≥ 5, but <15 per hour; 2 = moderate: AHI ≥ 15, but <30 per hour; 3 = severe: AHI ≥ 30 per hour. View Large Hypothesis 1: Post-traumatic nightmares will occur across REM and NREM sleep Of the 24 nightmares captured in this study, 10 arose from REM sleep and 14 from NREM (stages N1 and N2). Of the three participants who recorded more than one nightmare during their sleep study, two had nightmares in NREM sleep only and the third had one nightmare arising from REM sleep and a second nightmare arising from NREM sleep. This hypothesis was therefore supported. Hypothesis 2: Post-traumatic nightmares that arise from REM sleep will have non-replay content while those that arise from NREM sleep will involve accurate replay of all or part of an actual traumatic event Seven of the nightmares captured in the study were characterized as accurate trauma replays and 17 were characterized as non-replay, including eight with a mixture of replay and non-replay content. The relationship between stage of sleep (REM vs NREM) and dream-type (Replay vs mixed and non-replay) was not significant, X2 (1, N = 24) = 0.172, p = .097 and the hypothesis was therefore not supported. Hypothesis 3: Post-traumatic nightmares that arise from REM sleep will be associated with a gradual increase in heart rate prior to waking while those arising from NREM will be associated with a sudden increase in heart rate on waking Increase in heart rate was associated with awakening in 21 of the 24 cases. This occurred during awakening in four cases, with a range of 4–26 beats per minute increase (mean = 13.50, SD = 9.29) and following awakening in 18 cases, with a range of 6–52 beats per minute increase (mean = 21.33, SD = 16.88). There was no indication of a gradual increase in heart rate in the preceding sleep period associated with nightmares arising from either REM or NREM sleep. This hypothesis was therefore not supported. As noted above, the sleep factors associated with nightmares were investigated in an exploratory way with no a priori hypotheses. The sleep record was used to determine the features of sleep and sleep disturbance associated with nightmare awakening. In addition, with the particular role of prazosin in the treatment of post-traumatic nightmares,28 differences in the nightmares of participants who were and were not taking prazosin were investigated in an exploratory way. Sleep Factors Associated With Nightmare Awakening Of the 21 participants who recorded a nightmare during their sleep study, 13 had obstructive sleep apnea (seven mild, three moderate, and three severe). Chi-square revealed no significant relationship between the presence and severity of sleep apnea, in participants who did and did not experience a nightmare during the study X2 (3, N = 35) = 0.1.04, p = .791. Among those who experienced a nightmare, chi-square revealed no significant relationship between the presence of sleep apnea and the sleep stage of nightmares (REM vs NREM) X2 (1, N = 21) = 0.269, p = .673 or dream type (replay vs non-replay) X2 (1, N = 21) = 1.64, p =.336. However, as shown in Table 2, the majority of awakenings with nightmares (n = 14) were associated with respiratory events in both REM and NREM sleep. Similarly, the majority of awakenings with nightmares (n = 15) were associated with leg movement across both REM and NREM sleep. In many cases, awakenings were associated with a combination of these sleep events. Although the lack of video recording in the ambulatory sleep study precludes the diagnosis of REM behavior disorder, the sleep record of six participants who experienced nightmares indicated sustained muscle activity during REM sleep, consistent with REM behavior disorder. Thus, overall, significant sleep disturbance was associated with the post-traumatic nightmares. Differences Between Those Taking and Not Taking Prazosin Chi-square revealed no difference in the likelihood of participants experiencing a nightmare during the study whether or not they were taking prazosin X2 (1, N = 35) = 0.326, p = .413. This is not surprising as participants were selected on the basis of their frequent nightmares. However, chi-square analysis also revealed no difference between those taking and not taking prazosin on the sleep stage of nightmares (REM vs NREM) X2 (1, N = 21) = 2.386, p =.140 or nightmare type (replay vs non-replay) X2 (1, N = 21) = 0.131, p =.557. Furthermore, ANOVA revealed no difference in the phenomenological features (including how disturbing, vividness of sensory detail, and strength of physical reactions) of nightmares between those taking and those not taking prazosin. DISCUSSION The key finding of this study was that the post-traumatic nightmares of PTSD occur across both REM and NREM sleep. This is contrary to our current understanding that nightmares arise primarily from REM sleep. There was no difference in the incidence of replay versus non-replay dreams arising from REM and NREM sleep. Of interest, the dreams arising from both REM and NREM were associated with increase in heart rate at the time of awakening or following awakening. This differs from the description of Fisher et al.18 of nightmares arising from REM sleep as involving a gradual increase in anxiety that eventually awoke the dreamer over the approximately 20-minute dreaming period, with heart rate returning to normal on awakening. The observed relationship between nightmare reports and increase in heart rate may be a spurious one. Arousals from sleep in healthy individuals are associated with vigorous cardiovascular activity.29 This phenomenon may be heightened in PTSD with its association with frequent arousals from sleep and an already elevated basal and reactive level of cardiovascular activity,30 and further compounded by hypopneas associated with obstructive sleep apnea. Thus, there are a number of reasons for increased heart rate associated with awakening but the question remains unanswered as to the relationship between increase in heart rate and the report of nightmares. Rather than the anxiety associated with nightmares leading to a (gradual or sudden) increase in heart rate, the reverse may be true; that is, the nightmares are triggered on awakening by the increase in heart rate. The dominant theoretical model of PTSD goes some way to account for this possibility. Foa and Rothbaum31,32 propose that in individuals who develop PTSD, the psychologically overwhelming nature of the experience means that memory for the traumatic event (sensory details, the individual’s response at the time, and meanings or interpretations of the event) is stored in an unprocessed way, encapsulated in what is termed the trauma memory network. It is proposed that connections between the elements of the network are made at the time of the trauma through classical conditioning, but, in the absence of emotional processing, the elements are fragmented in the individual’s awareness and deliberate recall. However, the entire trauma memory network can be activated by any one of the stimuli, responses, or meaning elements. In the case of post-traumatic nightmares, the increased heart rate associated with arousal from sleep may replicate the physiological arousal at the time of trauma, and thereby activate the entire trauma memory network, including sensory details of the event. In other words, the individual’s memory of the traumatic event may be triggered by the increased arousal on awakening via activation of the trauma memory network, and this memory is reported as a nightmare. Of course, this theory cannot account for non-replay nightmares. Another important finding of this study was that post-traumatic nightmares frequently occur in the context of other sleep disturbance, most commonly obstructive sleep apnea. This relationship has been observed in previous research. Most awakenings with nightmares were associated with a respiratory event and/or leg movements. A comparison of PTSD sufferers who do not experience post-traumatic nightmares would be needed to establish the extent to which the two aspects of sleep disturbance are related but, nevertheless, the high prevalence of undiagnosed sleep disorders in participants with post-traumatic nightmares, and the possibility that nightmares are triggered in some cases by arousal associated with sleep-disordered breathing or leg movements, highlights the need to diagnose and treat underlying sleep disorders as part of the approach to treatment of post-traumatic nightmares. The more a person is aroused from sleep the more likely that he or she will recall dreams, so by reducing arousals in OSA, it might be possible to reduce PTSD patients’ recall or experience of nightmares. Thus, eliminating OSA with targeted intervention such as continuous positive airways pressure (CPAP)33 may reduce the recollection of nightmares in PTSD. Indeed, a recent prospective study has found reductions in nightmare frequency and distress following 3 months of CPAP treatment.34 Limitations Study participants were almost all taking antidepressant medication that has REM suppressing properties. This may have impact on the proportion of events preceded by REM sleep. Study participants were inpatients. The extent to which results can be generalized to all PTSD sufferers is not known. It is unclear how sleep and nightmare content in an inpatient psychiatric facility might relate to sleep in the naturalistic setting. In line with generally accepted definitions of nightmares, the study only investigated nightmares that occurred around awakenings. It is important to acknowledge that some form of mental activity is assumed to be continuous during sleep,35 but this can only be captured by waking the person at set points during the sleep cycle. This was not done in the current study, limiting the findings to events associated with awakenings, rather than generalizing to all dream events. Study participants were recruited from a clinical treatment setting and had the range of comorbid mental health problems, typically associated with PTSD.36 Although not central to the diagnosis of these conditions, depression and anxiety have been associated with nightmares.37 In addition, participants were taking a range of psychotropic medications with potential impacts on sleep and nightmares,38 although this study identified no difference in the nightmares of those taking and not taking prazosin. Notwithstanding these limitations, this study was important in demonstrating that the post-traumatic nightmares of PTSD can be captured in sleep studies with the use of ambulatory PSG, and critically that post-traumatic nightmares can arise from both REM and NREM sleep. Future research should aim to replicate these findings with a larger number and diversity of participants and to further investigate the possibility that the post-traumatic nightmares of PTSD comprise a range of phenomena that vary across the stage of sleep and their association with underlying sleep disorders. Future research should also investigate the effect of treating underlying sleep disorders on the occurrence of post-traumatic nightmares. SUPPLEMENTARY MATERIAL Supplementary material is available at SLEEP online. FUNDING This study was funded by a research grant from the Defence Health Foundation in Australia. ACKNOWLEDGMENTS We gratefully acknowledge the assistance of staff from the Austin Health Sleep Clinic and the Heidelberg Repatriation Hospital Psychological Trauma and Recovery Service and study participants. REFERENCES 1. Ross RJ, Ball WA, Sullivan KA, Caroff SN. Sleep disturbance as the hallmark of posttraumatic stress disorder. Am J Psychiatry . 1989; 146( 6): 697– 707. Google Scholar CrossRef Search ADS PubMed 2. Spoormaker VI, Montgomery P. Disturbed sleep in post-traumatic stress disorder: secondary symptom or core feature? Sleep Med Rev . 2008; 12( 3): 169– 184. Google Scholar CrossRef Search ADS PubMed 3. Germain A, Buysse DJ, Nofzinger E. Sleep-specific mechanisms underlying posttraumatic stress disorder: integrative review and neurobiological hypotheses. Sleep Med Rev . 2008; 12( 3): 185– 195. Google Scholar CrossRef Search ADS PubMed 4. Mellman TA, Hipolito MM. Sleep disturbances in the aftermath of trauma and posttraumatic stress disorder. CNS Spectr . 2006; 11( 8): 611– 615. Google Scholar CrossRef Search ADS PubMed 5. Kobayashi I, Sledjeski EM, Spoonster E, Fallon WFJr, Delahanty DL. Effects of early nightmares on the development of sleep disturbances in motor vehicle accident victims. J Trauma Stress . 2008; 21( 6): 548– 555. Google Scholar CrossRef Search ADS PubMed 6. Koffel E, Khawaja IS, Germain A. Sleepdisturbances inposttraumaticstressdisorder:updatedreview andimplications fortreatment. Psychiatr Ann . 2016; 46( 3): 173– 176. Google Scholar CrossRef Search ADS PubMed 7. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders . 5th ed. Washington, DC. 2013. 8. Ehlers A, Hackmann A, Steil R, Clohessy S, Wenninger K, Winter H. The nature of intrusive memories after trauma: the warning signal hypothesis. Behav Res Ther . 2002; 40( 9): 995– 1002. Google Scholar CrossRef Search ADS PubMed 9. Krakow BJ, Ulibarri VA, Moore BA, McIver ND. Posttraumatic stress disorder and sleep-disordered breathing: a review of comorbidity research. Sleep Med Rev . 2015; 24C: 37– 45. Google Scholar CrossRef Search ADS 10. Krakow B, Melendrez D, Warner TD, Dorin R, Harper R, Hollifield M. To breathe, perchance to sleep: sleep-disordered breathing and chronic insomnia among trauma survivors. Sleep Breath . 2002; 6( 4): 189– 202. Google Scholar CrossRef Search ADS PubMed 11. Mellman TA, Kulick-Bell R, Ashlock LE, Nolan B. Sleep events among veterans with combat-related posttraumatic stress disorder. Am J Psychiatry . 1995; 152( 1): 110– 115. Google Scholar CrossRef Search ADS PubMed 12. Foulkes D. Dream reports from different stages of sleep. Journal of Abnormal and Social Psychology . 1962; 65( 1): 14– 25. Google Scholar CrossRef Search ADS PubMed 13. Phelps AJ, Forbes D, Creamer M. Understanding posttraumatic nightmares: an empirical and conceptual review. Clin Psychol Rev . 2008; 28( 2): 338– 355. Google Scholar CrossRef Search ADS PubMed 14. Phelps AJ, Forbes D, Hopwood M, Creamer M. Trauma-related dreams of Australian veterans with PTSD: content, affect and phenomenology. Aust N Z J Psychiatry . 2011; 45( 10): 853– 860. Google Scholar CrossRef Search ADS PubMed 15. Phelps AJ, Creamer M, Hopwood M, Forbes D. Features of posttraumatic dreams related to PTSD severity. J Trauma Stress Disord Treat . 2014; 3( 3): 1– 5. 16. van der Kolk B, Blitz R, Burr W, Sherry S, Hartmann E. Nightmares and trauma: a comparison of nightmares after combat with lifelong nightmares in veterans. Am J Psychiatry . 1984; 141( 2): 187– 190. Google Scholar CrossRef Search ADS PubMed 17. Medicine AAoS. International Classification of Sleep Disorders, Revised: Diagnostic and Coding Manual . Chicago, Illinois: American Academy of Sleep Medicine; 2001. 18. Fisher C, Byrne J, Edwards A, Kahn E. A psychophysiological study of nightmares. J Am Psychoanal Assoc . 1970; 18( 4): 747– 782. Google Scholar CrossRef Search ADS PubMed 19. Schlosberg A, Benjamin M. Sleep patterns in three acute combat fatigue cases. J Clin Psychiatry . 1978; 39( 6): 546– 549. Google Scholar PubMed 20. Kramer M, Schoen LS, Kinney L. Psychological and behavioral features of disturbed dreamers. Psychiatr J Univ Ott . 1984; 9( 3): 102– 106. Google Scholar PubMed 21. Germain A. Sleep disturbances as the hallmark of PTSD: where are we now? Am J Psychiatry . 2013; 170( 4): 372– 382. Google Scholar CrossRef Search ADS PubMed 22. Woodward SH, Arsenault NJ, Murray C, Bliwise DL. Laboratory sleep correlates of nightmare complaint in PTSD inpatients. Biol Psychiatry . 2000; 48( 11): 1081– 1087. Google Scholar CrossRef Search ADS PubMed 23. Weathers FW, Blake DD, Schnurr PP, Kaloupek DG, Marx BP, Keane TM. The Clinician-Administered PTSD Scale for DSM-5 (CAPS-5) . In: Affairs UDoV, ed. Vermont: National Center for PTSD; 2013. 24. Hensley MJ, Hillman DR, McEvoy RDet al. Guidelines for Sleep Studies in Adults . Sydney, NSW: Australasian Sleep Association and Thoracic Society of Australia and New Zealand. 2005. 25. Weathers FW, Litz BT, Keane TM, Palmieri PA, Marx BP, Schnurr PP. The PTSD Checklist for DSM-5 (PCL-5) . 2013. Scale available from the National Center for PTSD – mail web Site www.ptsd.va.gov Accessed January 29, 2015. 26. Babor TF, de la Fuente JR, Saunders J, Grant M. The Alcohol Use Disorders Identification Test. Guidelines for use in primary health care . Geneva: World Health Organization; 1992. 27. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand . 1983; 67( 6): 361– 370. Google Scholar CrossRef Search ADS PubMed 28. Aurora RN, Zac RS, Auerbach SHet al. Bestpracticeguide for the treatment ofnightmaredisorder inadults. J Clin Sleep Med . 2010; 6( 4): 389– 401. Google Scholar PubMed 29. Trinder J, Allen N, Kleiman Jet al. On the nature of cardiovascular activation at an arousal from sleep. Sleep . 2003; 26( 5): 543– 551. Google Scholar PubMed 30. Buckley TC, Holohan D, Greif JL, Bedard M, Suvak M. Twenty-four-hour ambulatory assessment of heart rate and blood pressure in chronic PTSD and non-PTSD veterans. J Trauma Stress . 2004; 17( 2): 163– 171. Google Scholar CrossRef Search ADS PubMed 31. Foa EB, Rothbaum BO. Behavioural psychotherapy for post-traumatic stress disorder. International Review of Psychiatry . 1989; 1( 3): 219– 226. Google Scholar CrossRef Search ADS 32. Foa EB, Rothbaum BO. Treating the Trauma of Rape: Cognitive-Behavioral Therapy for PTSD . New York: Guilford Press; 1998. 33. Loube DI, Gay PC, Strohl KP, Pack AI, White DP, Collop NA. Indications for positive airway pressure treatment of adult obstructive sleep apnea patients: a consensus statement. Chest . 1999; 115( 3): 863– 866. Google Scholar CrossRef Search ADS PubMed 34. El-Solh AA, Vermont L, Homish GG, Kufel T. The effect of continuous positive airway pressure on post-traumatic stress disorder symptoms in veterans with post-traumatic stress disorder and obstructive sleep apnea: a prospective study. Sleep Med . 2017; 33: 145– 150. Google Scholar CrossRef Search ADS PubMed 35. Schredl M, Wittmann L. Dreaming: a psychological view. Swiss Arch Neruol Psychiatr . 2005; 156: 484– 492. Google Scholar CrossRef Search ADS 36. Brady KT, Killeen TK, Brewerton T, Lucerini S. Comorbidity of psychiatric disorders and posttraumatic stress disorder. J Clin Psychiatry . 2000; 61 Suppl 7: 22– 32. Google Scholar PubMed 37. Ohayon MM, Morselli PL, Guilleminault C. Prevalence of nightmares and their relationship to psychopathology and daytime functioning in insomnia subjects. Sleep . 1997; 20( 5): 340– 348. Google Scholar CrossRef Search ADS PubMed 38. Pagel JF. Drugs, dreams and nightmares. Sleep Med Clin . 2010; 5: 277–287. © Sleep Research Society 2017. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail firstname.lastname@example.org.
SLEEP – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera