Sleep Disturbances and Psychiatric Disorders Associated With Posttraumatic Stress Disorder in the General Population Maurice M. Ohayon and Colin M. Shapiro The aim of the study was to assess sleep disturbances in subjects with posttraumatic stress disorder (PTSD) from an urban general population and to identify associated psychiatric disorders in these subjects. The study was performed with a representative sample of 1,832 respondents aged 15 to 90 years living in the Metropolitan Toronto area who were surveyed by telephone (participation rate, 72.8%). Interviewers used Sleep-EVAL, an expert system specifically designed to conduct epidemiologic studies of sleep and mental disorders in the general population. Overall, 11.6% of the sample reported having experienced a traumatic event, with no difference in the proportion of men and women. Approximately 2% (1.8%) of the entire sample were diagnosed by the system as having PTSD at the time of interview. The rate was higher for women (2.6%) than for men (0.9%), which translated into an odds ratio (OR) of 2.8 (95% confidence interval [CI], 1.3 to 6.1). PTSD was strongly associated with other
mental disorders: 75.7% of respondents with PTSD received at least one other diagnosis. Most concurrent disorders (80.7%) appeared after exposure to the traumatic event. Sleep disturbances also affected about 70% of the PTSD subjects. Violent or injurious behaviors during sleep, sleep paralysis, sleep talking, and hypnagogic and hypnopompic hallucinations were more frequently reported in respondents with PTSD. Considering the relatively high prevalence of PTSD and its important comorbidity with other sleep and psychiatric disorders, an assessment of the history of traumatic events should be part of a clinician’s routine inquiry in order to limit chronicity and maladjustment following a traumatic exposure. Moreover, complaints of rapid eye movement (REM)-related sleep symptoms could be an indication of an underlying problem stemming from PTSD. Copyright r 2000 by W.B. Saunders Company
hallmark of PTSD. This conclusion fueled a growing interest in the study of sleep characteristics in subjects with PTSD. Polysomnographic studies have shown a disturbance in phasic REM sleep activity that could be compared with the hyperarousal manifestations observed in subjects with PTSD during the daytime. This dysregulation of REM activity manifests as recurrent awakenings that are often preceded by REM sleep24,25 and bad dreams or anxiety dreams occurring in both REM and non-REM sleep.26,27 High motor activity during the sleep of subjects with PTSD has also been reported during non-REM sleep. However, other polysomnographic studies failed to replicate these findings.28 The overwhelming majority of these aforementioned studies have been performed with war veterans, and one investigated the victims of a hurricane.
OSTTRAUMATIC STRESS disorder (PTSD) is a disorder caused by the experience of a traumatic event. It is characterized by a persistent reexperiencing of the traumatic event, persistent avoidance of stimuli associated with the trauma, numbing of general responsiveness, and persistent symptoms of increased arousal. It was first identified among veterans of World War II. In this specific population, the prevalence of PTSD has been estimated at 10% to 67%.1-4 PTSD has since been diagnosed in many individuals exposed to a variety of traumas. These include civilian victims of war,5,6 victims of natural disasters such as earthquakes and tornadoes,7,8 victims of sexual assault, aggression,9 or accidents,10-13 and persons exposed to suicide,14 severe injury to loved ones, and serious life-threatening disease.15-16 Epidemiologic studies in the general population have estimated the lifetime prevalence of PTSD to be 1% to 14% depending on the population sampled.3,17-21 The 1-month prevalence has been set at 1% to 2.3%.17-19,22 The persistent reexperience of the trauma can take the form of recurrent distressing dreams of the event. This has led researchers to study the sleep of subjects with PTSD. In 1989, Ross et al.23 proposed that dysfunctional rapid eye movement (REM) sleep mechanisms could be responsible for the distressing bad dreams reported in PTSD patients, and concluded that sleep disturbances were the
From the Sleep Disorders Center, Stanford University School of Medicine, Stanford, CA; and Department of Psychiatry, The Toronto Western Hospital, The University Health Network, Toronto, Ontario, Canada. Supported by the Fonds de la Recherche en Sante´ du Que´bec (971067). Address reprint requests to Maurice M. Ohayon, M.D., D.Sc., Ph.D., Stanford University School of Medicine, Sleep Disorders Center, 401 Quarry Rd, Suite 3301, Stanford, CA 94305. Copyright r 2000 by W.B. Saunders Company 0010-440X/00/4106-0011$10.00/0 doi:10.1053/comp.2000.16568
Comprehensive Psychiatry, Vol. 41, No. 6 (November/December), 2000: pp 469-478
OHAYON AND SHAPIRO
PTSD is very often associated with other mental disorders (depressive and anxiety disorders), reaching 80% of cases in some studies.3 Furthermore, exposure to trauma alone does not necessarily predict PTSD, nor does the severity of the traumatic exposure. Individual factors such as personality prior to the trauma,29 social support after exposure to trauma, and age at trauma,2,7,16,30 as well as the duration and intensity of exposure to trauma, are linked to its development.31 Certain traumas such as war and rape32 have a very high likelihood of precipitating PTSD. The present study in a representative sample of a general population examined to what extent sleep disorders (1) are specific in PTSD, (2) are exacerbated or triggered by PTSD, and (3) are related to the traumatic event, to PTSD, or to an associated mental disorder. METHOD This epidemiologic study was performed from March 1996 to January 1997 in the Metropolitan Toronto area (Ontario, Canada). Toronto is the largest city in Canada, with approximately 3,138,415 of the population aged 15 years or older. A representative sample of subjects aged 15 years or older was drawn using a two-stage sampling design. At the first stage, a random sample of telephone numbers was drawn based on the population distribution within the Metropolitan Toronto area. The first three digits in the telephone number served to identify the location of a target household. In the second stage, a controlled selection method was applied to limit the within–sampling unit noncoverage error. The Kish method33 was used to this end and served to select one respondent in the household. If the household member thus chosen refused to participate, the household was removed and replaced by another and the process was repeated. Interviewers explained the goals of the study to potential participants before requesting verbal consent. Excluded from the study were subjects who did not speak sufficient English or who had a hearing or speech impairment or an illness which precluded being interviewed. The ethics committee of the Montreal Philippe Pinel Center, a Canadian governmental and university ethical institution, approved the study. Subjects who refused to participate or who asked to stop before completing at least half of the interview were classified as refusals. Phone numbers were removed and replaced only after a minimum of 10 unsuccessful dial attempts were made at different times and on different days, including weekdays and weekends. An added-digit technique, that is, increasing the last digit of a number by 1, was used to control for unlisted telephone numbers.34 As a result, the final sample consisted of 13.8% unlisted numbers. The participation rate (72.8%) was calculated based on the number of completed interviews (n ⫽ 1,832) divided by the number of eligible telephone numbers, which included all residential numbers not meeting any of the exclusion criteria (n ⫽ 2,516). The final sample included 5% (n ⫽ 126) of sub-
jects who initially refused to participate but accepted upon a second request.
Interviewers Interviews were conducted from the Philippe Pinel Research Center by telephone using Sleep-EVAL. Interviews were performed by 30 interviewers, all university students inexperienced in psychiatric assessment but having received special training on how to use the Sleep-EVAL Knowledge-Based System. The training required between 6 and 18 hours and consisted mainly of role-playing, during which the interviewers practiced how to introduce the study and how to ask and to answer the questions. They were instructed to read all of the choices of answer and to never decide for the subject what is the most appropriate answer. The mean duration of the interviews was 40.4 ⫾ 20.0 minutes. The team of interviewers was monitored daily by two supervisors to ensure that questions were asked correctly and data entered properly.
Instrument The Sleep-EVAL system, intelligent computer software, is specifically designed to administer questionnaires and conduct epidemiologic studies on mental and sleep disorders in the general population.35 It includes a non-monotonic, level-2 inference engine endowed with a causal reasoning mode that simulates the reasoning process used in a psychiatric consultation. The causal reasoning mode enables the Sleep-EVAL system to formulate a series of diagnostic hypotheses based on the responses provided by a subject. The non-monotonic, level-2 inference engine examines these hypotheses and confirms or rejects them through further questions and deductions. The system renders its diagnoses using two classifications, the DSM-IV36 and the International Classification of Sleep Disorders (ICSD).37 The system formulates initial diagnostic hypotheses on the basis of responses to a standard set of questions posed to all subjects. Concurrent mental diagnoses are allowed in accordance with the DSM-IV. The system terminates the interview once all ICSD and DSM-IV diagnostic possibilities are exhausted. The system selects and phrases the questions to be used and provides examples and instructions on how to ask them. The interviewer simply reads out the questions as they appear on the monitor and enters the responses. Questions can be closedended (e.g., yes-no, present-absent-unknown, or five-point scale) or open-ended (e.g., name of illness and duration). The expert system’s questionnaire is designed in such a manner that the decision regarding the presence of a symptom is based on the interviewee’s responses rather than the interviewer’s judgment. This approach has been proven to yield better agreement between lay interviewers and psychiatrists for the diagnosis of minor psychiatric disorders.38 Concurrent diagnoses were explored as long as a symptom, criterion, or syndrome particular to that diagnosis was present. Elimination of a diagnosis was based on DSM-IV and ICSD criteria. The past history of the subject was used accordingly with the explored classification. The system has been validated in various contexts and has been demonstrated to be reliable and valid. In a study involving four psychiatrists and 114 patients from general practice, a of
SLEEP AND PSYCHIATRIC DISORDERS IN PTSD
.97 was found between the diagnosis of the system and the consensus diagnosis of psychiatrists. The between each psychiatrist (regardless of how psychiatrists agree with each other) and the system was .44 to .78.39,40 Another study compared the diagnoses of the expert system used by a psychologist against those of 10 psychiatrists in a forensic hospital and involved 91 patients.41 A of .48 was obtained for specific diagnoses of psychotic disorders, mostly schizophrenia, with a concordance for that diagnosis of 72.6% between the system and psychiatrists. In the general population, the diagnoses obtained by two lay interviewers using Sleep-EVAL were compared against those obtained by two clinical psychologists on 150 subjects. A of .85 was obtained for the recognition of sleep problems and .70 for insomnia disorders. A study with 105 patients of two sleep disorder centers (Stanford University and Regensburg University, Germany) reported a of .73 for any dyssomnia diagnosis. The most frequent diagnoses were obstructive sleep apnea syndrome ( ⫽ .94) and insomnia disorders ( ⫽ .78).42 As part of the data analysis process, we verified the reliability of the system’s reasoning by reconstructing each step of the decision-making process. The algorithm to achieve a diagnosis has been recreated from the responses provided to questions. Then, results from this algorithm were compared with the diagnoses of the expert system. All diagnoses used in this report have been verified prior to proceeding with the analysis.
Assessment PTSD was assessed through a series of 21 main questions that covered the entire description of the disorder.36 A probe question was first asked about experiences remembered as traumatic by the subject: ‘‘Did you experience an event that is outside the range of usual human experience [more than a sorrow or a chronic illness . . .]? For example, a serious threat to your life or your physical integrity such as war exposure, rape, robbery or attack; a serious threat or harm to your children, spouse, or other close relatives and friends such as kidnapping, torture, or murder; a sudden destruction of your home or your community such as a natural catastrophe or a fire, seeing another person being seriously injured or killed in an accident or as the result of physical violence?’’ Subsequently, they were asked to describe their traumatic experience. The questions assessing the symptoms associated with the traumatic experience were asked for the current time. We also assessed sleep disturbances associated with PTSD including sleep symptomatology (insomnia symptoms, hypersomnia symptoms, etc.) and sleep disorders according to DSM-IV and ICSD classifications.
Analyses A weighting procedure was applied to correct for disparities in the geographic, age, and gender distribution between the sample and the Toronto area population according to the Canadian census data. Results are based on weighted n values. Percentages for target variables are given with 95% confidence intervals (CIs). Bivariate analyses were performed using the chi-square (2) test with Yates’ correction or Fisher’s exact test when n values were smaller than five. Reported differences were significant at the .05 level or less. Logistic regression25 was used to compute the odds ratio (OR) associated with PTSD. Logistic regression was performed using the SUDAAN software (version
7.5.2; Research Triangle Institute, Research Triangle Park, NC) that allows an appropriate estimate of the standard errors from stratified samples by means of a Taylor series linearization method. The reported differences were significant at .05 or less.
The sample included 48.3% men and 51.7% women aged 15 to 90 years. The demographic characteristics of the sample are presented in Table 1. Exposure to a Traumatic Event Overall, 11.6% of the sample reported having experienced an event that had a traumatic impact on them (criterion A of PTSD diagnosis in DSMIV). This percentage contained an equal number of men and women (Table 2). Of the events reported, 32.4% represented some sort of aggression such as being attacked, stabbed, or robbed, marital abuse, or aggression by an animal. Exposure to war was reported by two subjects. Life-threatening illness or death of a family member was reported by nearly 20% of the sample. Close to half of the events occurred within 5 years prior to the interview. Fulfillment of DSM-IV Criteria The most frequently reported symptoms relative to each diagnostic criterion of the DSM-IV were investigated. For criterion B (i.e., the traumatic event is persistently reexperienced), the two most commonly reported symptoms were recurrent and intrusive distressing recollections of the event (33%) and recurrent distressing dreams of the event (24.8%). Respondents who met criterion B were queried regarding criterion C (i.e., persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness). The two most commonly reported symptoms were avoidance of thoughts, feelings, or conversations associated with the trauma (49.8%) and avoidance of activities, places, or people that arouse recollections of the trauma (44.7%). Subsequently, respondents who met criterion C were queried regarding criterion D (i.e., persistent symptoms of increased arousal). The two most commonly reported symptoms were hypervigilance (61.2%) and irritability or outbursts of anger (60.3%). Of the subjects who met all criteria for A to D, nine respondents did not meet the final criterion (criterion F), as they reported no impact on daily functioning. Respondents who met all
OHAYON AND SHAPIRO
Table 1. Demographic Characteristics of the Sample Characteristic
Gender Male Female Age group (yr) 15-29 30-44 45-59 60-74 ⱖ75 Marital status Single Married Separated/divorced Widowed Education (yr) ⬍11 11-13 ⬎13 Ethnic origin White Black Hispanic Asian Other
Prevalence of PTSD
505 637 343 267 79
458 683 372 247 72
27.6 34.8 18.7 14.6 4.3
2.0 2.9 1.5 0.4 —
0.8-3.2 1.6-4.2 0.2-2.8 0.0-1.2
650 915 153 115
617 943 160 112
35.5 49.9 8.3 6.3
2.8 1.0 4.6 —
1.5-4.1 0.4-1.6 1.3-7.9
378 969 480
367 971 489
20.7 53.1 26.3
1.9 2.1 1.4
0.5-3.3 1.2-3.0 0.3-2.5
1,346 103 33 143 200
1,352 102 33 143 196
73.7 5.7 1.8 7.8 11.0
1.8 3.1 7.4 0.8 1.5
1.8-1.8 0.0-6.4 0.0-16.3 0.0-2.3 0.0-3.2
*Weighted data for gender, age group, and geographic, distribution.
criteria from A through F were diagnosed with PTSD. The prevalence of each criterion by gender is listed in Table 2. As mentioned before, 11.6% of the sample reported experiencing a traumatic event (criterion A); 5.9% of the sample experienced a traumatic event and met criterion B; 2.4% met criteria A, B, and C; and 1.9% met criteria A, B, C, and D. The complete diagnostic criteria set was met by 1.8% of the sample (current prevalence). The prevalence was higher for women (2.7%) than for
men (0.9%), which translated into an OR of 2.8 (95% CI, 1.3 to 6.1). With regard to the type of traumatic event, PTSD was provoked by an event related to aggression in 24.8% of cases, illness of a family member in 15.1%, death of a family member in 5.4%, and illness of the respondent in 17.0%. For the sociodemographic variables, logistic regression analysis served to identify the following predictive factors for PTSD: female gender (OR ⫽ 3.0), being single (OR ⫽ 2.8), and being separated or divorced (OR ⫽ 3.8). Employment
Table 2. Prevalence of PTSD Criteria by Gender Total Criterion
A. Exposure to a traumatic event eliciting a response involving intense fear, helplessness, or horror B. The traumatic event is persistently reexperienced (1 symptom) C. Persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness (3 symptoms) D. Persistent symptoms of increased arousal (2 symptoms) *P ⬍ .05.
SLEEP AND PSYCHIATRIC DISORDERS IN PTSD
status, education level, and ethnic origin proved insignificant. Associated Sleep Disturbances Respondents with PTSD were more likely than non-PTSD respondents to report insomnia symptoms. The exception was the symptom of nonrestorative sleep (Table 3). Fulfillment of DSM criteria for insomnia disorders (i.e., complaints of insomnia lasting at least 1 month and accompanied by daytime repercussions) was observed in 39.6% of respondents with PTSD, as compared with 6.5% in non-PTSD respondents (P ⬍ .001). This finding indicates that nearly 60% of PTSD subjects with insomnia complaints had an insomnia disorder, while this rate was only 20% in non-PTSD insomnia complainers. The rates of obstructive sleep apnea syndrome and periodic limb movement disorder were similar for both groups. Nightmares occurring at least 1 night per month were more frequent among respondents with PTSD. The highest OR was observed for violent or dangerous behaviors during sleep (punching, kicking, running out of bed, etc.), with the likelihood of reporting such behavior being 10 times higher for respondents with PTSD. Violent or dangerous behaviors during sleep resulted in injuries to the subjects or to someone else in about half of the cases. In 70% of the cases, these behaviors occurred within the first half of the night, therefore reducing the likelihood of occurrence during REM sleep. Other parasomnias (sleep paralysis and sleep talking) were also more frequently reported among respondents with PTSD. Respondents with PTSD did not report episodes of sleepwalking.
The likelihood of reporting hypnagogic (occurring at sleep onset) and hypnopompic (occurring at awakening) hallucinations was greater for respondents with PTSD. About 70% of PTSD subjects were terrified by their hypnagogic hallucinations, compared with 40% of non-PTSD complainers (P ⬍ .05). We addressed the question of whether a particular type of traumatic event is more specifically associated with sleep disorders. Having been a victim of aggression, more so than any other type of event, was associated with nightmares (OR ⫽ 2.4). Are these sleep disturbances related to the PTSD or to the traumatic event? We then examined if subjects reporting the experience of a traumatic event but without the full-fledged manifestation of a PTSD also had high rates of sleep disturbances compared with individuals who had not reported such an experience. Disrupted sleep and early morning awakening had similar rates in both groups. Difficulties in initiating sleep and nonrestorative sleep were significantly greater among subjects who had experienced a traumatic event, with an OR of 1.5 each. The observed rates in this latter group were more than two times lower than in PTSD subjects. Sleep paralysis, sleep talking, and violent behavior during sleep were also significantly higher in the traumatic event group, with an OR of 1.4. Rates of hypnagogic and hypnopompic hallucinations were comparable between the two groups. Did these sleep disturbances exist prior to the PTSD? Subsequently, we examined whether the sleep disorders were present before the traumatic event or appeared afterward. Insomnia complaints existed prior to the traumatic event in 60.9% of
Table 3. Associated Sleep Disturbances Non-PTSD (n ⫽ 1,798)
PTSD (n ⫽ 34)
Difficulties initiating sleep Disrupted sleep Early morning awakening Nonrestorative sleep Sleep paralysis Sleep talking Nightmares Violent or dangerous behaviors during sleep Hypnagogic hallucinations Hypnopompic hallucinations
12.6 17.9 13.1 13.8 2.2 19.4 4.2 1.6 21.8 6.0
227 322 236 248 40 349 76 29 392 108
41.2 46.6 42.9 26.2 7.4 30.7 18.8 14.6 45.5 34.3
14 16 15 9 3 10 6 5 15 12
4.8 4.0 5.0
*P ⬍ .001, †P ⬍ .01, ‡P ⬍ .05 (2 test, df ⫽ 1).
2.4-9.7* 2.0-7.9* 2.5-9.9* NS
3.7 2.4 5.3 10.7 3.0 8.2
1.0-13.4‡ 1.2-4.9† 2.2-12.8† 3.9-29.8† 1.5-5.9* 4.0-17.1*
OHAYON AND SHAPIRO
PTSD subjects with such complaints. Excessive daytime sleepiness was present before the trauma in 71.4% of PTSD subjects with this complaint. Parasomnia symptoms (hypnagogic or hypnopompic hallucinations, violent behavior during sleep, and nightmares) appeared after the traumatic event in 60% of PTSD subjects with these complaints.
Perception of the Social Network Respondents with PTSD were more likely to be dissatisfied in their relationships with their spouse (OR ⫽ 18.8, 95% CI ⫽ 3.2 to 68.0) and friends (OR ⫽ 6.0, 95% CI ⫽ 2.7 to 13.5). They reported more frequently that, generally speaking, their social life was dissatisfying (OR ⫽ 4.4, 95% CI ⫽ 2.2 to 8.8). They spent their leisure hours alone more than half of the time in the 2 months prior to the interview (OR ⫽ 2.7, 95% CI ⫽ 1.3 to 5.7). They were also more likely to report having no confidence (OR ⫽ 3.0, 95% CI ⫽ 1.3 to 7.4) and feeling that nobody could help them (OR ⫽ 3.7, 95% CI ⫽ 1.6 to 8.7) when they had difficulties. More than half of the respondents with PTSD assessed their life as being stressful (OR ⫽ 4.4, 95% CI ⫽ 2.2 to 8.6).
Associated Mental Disorders The co-occurrence of PTSD with other mental disorders was assessed. The most frequently associated disorder was bipolar disorder (OR ⫽ 14.1). Depressive disorders (major depressive disorder and dysthymia) and panic disorder ranked second (Table 4). Overall comorbidity was very high, as 75.7% of respondents with PTSD received at least one other diagnosis of mental disorder (OR ⫽ 22.4). We also examined whether the associated mental disorder was present before the onset of PTSD. In 80.7% of cases, the associated mental disorder appeared after the traumatic event. In addition, we verified whether these respondents were greater consumers of health care. Both groups had consulted a physician at least once in the past year in more than 75% of cases. However, subjects diagnosed with PTSD in our study had more consultations (mean, 6.98 consultations in the past year) than the other subjects (2.9 consultations, P ⫽ .001). Respondents with PTSD also had a higher rate of hospitalization in the past year (22.5% v 8.8%, P ⫽ .01). We also verified whether respondents with PTSD were greater users of medication. We found that a higher proportion currently used antidepressants: 13.6% compared with 2.9% of the other respondents (P ⬍ .001).
Until recently, PTSD was a diagnosis reserved essentially for victims and veterans of war. However, it became evident from the literature of the last decade that a broader range of trauma can also provoke the disorder, including exposure to disasters, crime, bereavement, or witnessing or hearing accounts of death. We found in our study that traumas involving some kind of aggression were most likely to lead to PTSD, followed by lifethreatening disease experienced by the respondent or a family member. This last category of stressor is less commonly investigated. However, the body of research focusing on the development of PTSD following a life-threatening illness is growing. For example, Dew et al.16 found a 13.7% rate of PTSD in heart transplant patients in the year after surgery.
Table 4. Current Associated Mental Disorders Non-PTSD (n ⫽ 1,798) Disorder
Mood disorder Depressive disorder‡ Bipolar disorder Anxiety disorder Generalized anxiety Panic disorder Agoraphobia Simple phobia Non-PTSD mental disorder (total) *P ⬍ .001, †P ⬍ .01 (2 test, df ⫽ 1). ‡Major depressive disorders and dysthymia.
PTSD (n ⫽ 34) No.
1.7 2.8 1.9 1.3 12.2
31 50 34 23 219
13.7 22.1 16.0 8.0 75.7
5 8 5 3 26
9.4 9.7 9.8 6.5 22.4
3.3-26.4† 4.1-22.7* 3.7-26.1† 1.8-23.8† 10-49.5*
SLEEP AND PSYCHIATRIC DISORDERS IN PTSD
Alter et al.15 reported a PTSD lifetime rate of 22% among cancer survivors. This investigation is part of a broader study that explored several forms of disorders. As a result, the assessment of PTSD was selectively limited to curtail the length of the interview. Therefore, respondents who did not report a traumatic event or who failed to meet a diagnostic criterion were not further queried in this regard (e.g., respondents who presented no symptom relative to criterion B were not queried regarding symptoms associated with criteria C and D). Studies that used a checklist or scale of potential traumatic events, such as the Modified PTSD Symptom Scale,22 yielded lifetime rates of 39.1% to 80%.17,22 Obviously, the use of a checklist will yield a higher rate of exposure to several traumatic events but does not necessarily lead to a higher rate of PTSD. Indeed, the current prevalence of PTSD reported in these studies was similar to ours. Despite the obvious convenience of checklists or scales of potential traumatic events to explore the prevalence of certain traumatic events in the population, some events such as rape, spousal violence, and sexual abuse still remain underreported. In our study, the investigation of a traumatic event was based on its traumatic effect and not on a preestablished list of possible traumatic events: someone may have a traumatic event according to the checklist but it was not experienced as such. We think this approach is more pertinent because it is closer to the natural diagnostic process. A question could be raised about the reliability of data collected by telephone. Previous studies using this methodology for data collection indicate that, in general, telephone interviews are satisfactory, have good interrater reliability, and provide results comparable to other interview techniques.43,44 Furthermore, studies that compared the ability of telephone versus in-person interviews to investigate the prevalence of PTSD obtained comparable rates for the two methods.45 The National Institute of Mental Health Epidemiological Catchment Area study was one of the first community-based investigations to assess the prevalence of PTSD.3 The lifetime rate for the disorder was estimated at 1%. More recently, however, the National Comorbidity Survey set the lifetime prevalence at 7.8%.19 Most studies of PTSD have found that this condition becomes
chronic (i.e., symptoms last at least 1 year) in about half of the cases. Studies that examined the current prevalence of PTSD reported rates comparable to ours. For example, Stein et al.22 found the current prevalence of PTSD to be 2.7% among women and 1.2% among men in a sample of 1,002 persons interviewed in Winnipeg (Canada). These figures are close to the 2.7% in women and 1% in men obtained in our study, which was performed in Toronto, a nearby city. A high co-occurrence of PTSD with other mental disorders (75.7%) was observed in our study. Other studies that addressed this issue reported similar results.3,19 PTSD, it would seem, is rarely an isolated problem. Our study shows that the most frequently associated disorder was bipolar disorder, followed by depressive and panic disorders, with ORs of 8.4 to 14.1. To what extent concomitant mental disorders develop before or after PTSD was not documented. In our study, we found that in 80.7% of the cases, the concomitant mental disorder was reported to develop after PTSD. It is possible that the traumatic event served to trigger the full-fledged manifestation of the disorder. However, it should be kept in mind that this information is derived from cross-sectional data and therefore relies on the memory of the subjects. Poor social support after exposure to a traumatic event has been associated with a higher risk of developing PTSD.2,7,16,30 In this study, we found that many of our subjects had a poorer social network. However, this also could have been a consequence of PTSD in that it could have been a numbing of the general responsiveness. Consequently, it was not surprising to find a notable number of respondents with PTSD who had been treated with antidepressants (13.6%). Sleep disturbances affected approximately 70% of respondents with PTSD. These disturbances were mainly difficulties in initiating or maintaining sleep; the reporting of nonrestorative sleep did not significantly differ between PTSD and non-PTSD groups. Some researchers have proposed that insomnia complaints may be due to other sleep disorders, such as periodic limb movement disorder, rather than PTSD.32 Periodic limb movement disorder along with other sleep disorders such as obstructive sleep apnea syndrome or sleep disorders related to a medical condition can be associated with insomnia complaints. However, we did not find a significant association between obstructive sleep apnea
OHAYON AND SHAPIRO
syndrome or periodic limb movement disorder and PTSD in our sample, nor were reports of insomnia associated with other primary sleep disorders or medical illnesses in our PTSD group. However, the fact that about 66% of our PTSD sample expressed difficulties in initiating or maintaining sleep following the experience of the trauma suggests multiple causes of sleep disorders. Our findings are in contrast to the results of another study concluding that there was no specific sleep pattern accompanying PTSD.28 Studies using sleep monitoring have shown an increased phasic REM sleep activity as evidenced by REM behavior disorder–like signs24,26 and a higher REM density with a reduced amount of REM sleep46 and delayed REM onset. In our study, we found some suggestive evidence of disturbed REM activity such as hypnagogic and hypnopompic hallucinations and sleep paralysis, which are REM intrusions into the awake state. In our sample of PTSD subjects, the hallucination rates are higher than those observed in the general population47 and the hallucinations are terrifying for the majority of PTSD subjects (70%) and involve the presence of someone threatening in the room. Alternatively, violent or dangerous behaviors during sleep appear to occur mainly during nonREM sleep, since the overwhelming majority of respondents in our study reported that these behaviors occurred during the first half of sleep. Sleep talking can be present in all sleep stages and can be precipitated by emotional stress. Nightmares are a
REM parasomnia disorder. They have been frequently associated with PTSD and are considered characteristic of this disorder.48 In our sample, about one fifth of the respondents with PTSD reported nightmares. Some studies have shown that ‘‘nightmares’’ in PTSD subjects occurred outside REM sleep.49 Consequently, ‘‘traumatic mentation’’ or anxiety dreams would be a better term to describe the mental processes of individuals with PTSD. Since our results are based on self-report, it is not possible to confirm at which sleep stage these bad mentations occurred. However, we can hypothesize that they occurred in non-REM sleep, because many of our PTSD subjects reported that they occurred during the first hours of sleep. General Recommendations To prevent poor adjustment and chronicity, early detection and intervention in PTSD are important to consider. Recent research with civilians has shown that exposure to multiple traumas increases the risk of developing PTSD.22,50 As the prevalence of the disorder alone is relatively high and its comorbidity with other psychiatric disorders is considerable, clinicians should routinely inquire about the history of traumatic events and other experiences that people are often reluctant to talk about, such as sexual assault. Moreover, a sleep perturbation involving a disturbed REM activity could be the indication of an underlying problem and suggestive of a PTSD that deserves to be investigated.
REFERENCES 1. Long N, MacDonald C, Chamberlain K. Prevalence of posttraumatic stress disorder, depression and anxiety in a community sample of New Zealand Vietnam War veterans. Aust N Z J Psychiatry 1996;30:253-256. 2. Boscarino JA. Posttraumatic stress and associated disorders among Vietnam veterans: the significance of combat exposure and social support. J Trauma Stress 1995;8:317-336. 3. Helzer JE, Robins LN, McEvoy L. Posttraumatic stress disorder in the general population. Findings of the Epidemiologic Catchment Area Survey. N Engl J Med 1987;317:16301634. 4. Stretch RH. Posttraumatic stress disorder among U.S. Army Reserve Vietnam and Vietnam era veterans. J Consult Clin Psychol 1985;53:935-936. 5. Somasundaram DJ, Sivayokan S. War trauma in a civilian population. Br J Psychiatry 1994;165:524-527. 6. Brom D, Kleber RJ, Witztum E. The prevalence of posttraumatic psychopathology in the general and the clinical population. Isr J Psychiatry Relat Sci 1992;28:53-63. 7. Sakano Y, Shimada H, Tsujiuchi T, Ito K, Akabayashi A,
Yoshiuchi K, et al. Psychosomatic problems after the Great Hanshin earthquake in January 1995. I. Symptoms of posttraumatic stress disorders and psychological stress responses. Jpn J Psychosom Med 1996;36:649-656. 8. McFarlane AC, Papay P. Multiple diagnoses in posttraumatic stress disorder in the victims of a natural disaster. J Nerv Ment Dis 1992;180:498-504. 9. Astin MC, Lawrence KJ, Foy DW. Posttraumatic stress disorder among battered women: risk and resiliency factors. Violence Vict 1993;6:17-28. 10. Carlier IVE, Lamberts RD, Van Uchelen AJ, Gersons BPR. Clinical utility of a brief diagnostic test of posttraumatic stress disorder. Psychosom Med 1998;60:42-47. 11. Ohry A, Rattok J, Solomon Z. Posttraumatic stress disorder in brain injury patients. Brain Injury 1996;10:687-695. 12. Gregg W, Medley I, Fowler Dixon R, Curran P, Loughrey G, Bell P, et al. Psychological consequences of the Kegworth air disaster. Br J Psychiatry 1995;167:812-817. 13. Watts R. Posttraumatic stress disorder after a bus accident. Aust N Z J Psychiatry 1995;29:75-83.
SLEEP AND PSYCHIATRIC DISORDERS IN PTSD
14. Brent DA, Moritz G, Bridge J, Perper J, Canobbio R. Long-term impact of exposure to suicide: a three year controlled follow-up. J Am Acad Child Adolesc Psychiatry 1996;35:646653. 15. Alter CL, Pelcovitz D, Axelrod A, Goldenberg B, Harris H, Meyers B, et al. Identification of PTSD in cancer survivors. Psychosomatics 1996;37:137-143. 16. Dew MA, Roth LH, Schulberg HC, Simmons RG, Kormos RL, Trzepacz PT, et al. Prevalence and predictors of depression and anxiety related disorders during the year after heart transplantation. Gen Hosp Psychiatry 1996;18(Suppl):48S61S. 17. Breslau N, Davis GC, Andreski P, Peterson EL. Traumatic events and posttraumatic stress disorder in an urban population of young adults. Arch Gen Psychiatry 1991;48:216222. 18. Breslau N, Davis GC. Posttraumatic stress disorder in an urban population of young adults: risk factors for chronicity. Am J Psychiatry 1992;149:671-675. 19. Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB. Posttraumatic stress disorder in the National Comorbidity Survey. Arch Gen Psychiatry 1995;52:1048-1060. 20. Lindal E, Stefansson JG. The lifetime prevalence of anxiety disorders in Iceland as estimated by the US National Institute of Mental Health Diagnostic Interview Schedule. Acta Psychiatr Scand 1993;88:29-34. 21. Shore JH, Vollmer WM, Tatum EL. Community patterns of posttraumatic stress disorders. J Nerv Ment Dis 1989;177:681685. 22. Stein MB, Walker JR, Hazen AL, Ford DR. Full and partial posttraumatic stress disorder: findings from a community survey. Am J Psychiatry 1997;154:1114-1119. 23. Ross RJ, Ball WA, Sullivan KA, Caroff SN. Sleep disturbance as the hallmark of posttraumatic stress disorder. Am J Psychiatry 1989;146:697-707. 24. Ross RJ, Ball WA, Sanford LD, Morrison AR, Dinges DF, Silver SM, et al. Rapid eye movement sleep changes during the adaptation night in combat veterans with posttraumatic stress disorder. Biol Psychiatry 1999;45:938-941. 25. Ross RJ, Ball WA, Dinges DF, Kribbs NB, Morrison AR, Silver SM, et al. Rapid eye movement sleep disturbance in posttraumatic stress disorder. Biol Psychiatry 1994;35:195-202. 26. Mellman TA, Kulick-Bell R, Ashlock LE, Nolan B. Sleep events among veterans with combat-related posttraumatic stress disorder. Am J Psychiatry 1995;152:110-115. 27. Mellman TA, David D, Kulick-Bell R, Hebding J, Nolan B. Sleep disturbance and its relationship to psychiatric morbidity after Hurricane Andrew. Am J Psychiatry 1995;152:16591663. 28. Hurwitz TD, Mahowald MW, Kuskowski M, Engdahl BE. Polysomnographic sleep is not clinically impaired in Vietnam combat veterans with chronic posttraumatic stress disorder. Biol Psychiatry 1998;44:1066-1073. 29. Schnurr PP, Friedman MJ, Rosenberg SD. Premilitary MMPI scores as predictors of combat-related PTSD symptoms. Am J Psychiatry 1993;150:479-483. 30. Speed N, Engdahl BE, Schwartz J, Eberly RE. Posttraumatic stress disorder as a consequence of the POW experience. J Nerv Ment Dis 1989;177:147-153. 31. Rosenheck R, Fontana A. Long-term sequelae of combat in World War II, Korea and Vietnam: a comparative study. In:
Ursana RJ, McCaughey BG, Fullerton CS (eds): Individual and Community Responses to Trauma and Disasters. The Structure of Human Chaos. Cambridge, England: Cambridge University Press, 1994:330-359. 32. Brown TM, Boudewyns PA. Periodic limb movements of sleep in combat veterans with posttraumatic stress disorder. J Trauma Stress 1996;9:129-136. 33. Kish L. Survey Sampling. New York, NY: Wiley, 1965. 34. Landon EL, Banks SK. Relative efficiency and bias of plus-one telephone sampling. J Market Res 1977;14:294-299. 35. Ohayon MM, Guilleminault C, Paiva T, Priest RG, Rapoport DM, Sagales T, et al. An international study on sleep disorders in the general population: methodological aspects. Sleep 1997;20:1086-1092. 36. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Ed. 4. Washington, DC: American Psychiatric Association, 1994. 37. ICSD Diagnostic Classification Steering Committee. International Classification of Sleep Disorders: Diagnostic and Coding Manual (ICSD). Rochester, MN: American Sleep Disorders Association, 1990. 38. Lewis G, Pelosi AJ, Araya RC, Dunn G. Measuring psychiatric disorder in the community: a standardized assessment for use by lay interviewers. Psychol Med 1992;22:465486. 39. Ohayon M. Validation of a knowledge based system (ADINFER) versus human experts. In: Barahona P, Veloso M, Bryant J (eds): Proceedings of the Twelfth International Congress on Medical Informatics. Lisbon, Portugal: MIE, 1994: 90-95. 40. Ohayon M. Validation of expert systems: examples and considerations. Med Info 1995;8:1071-1075. 41. St-Onge B, Ohayon M. L’utilisation du syste`me Expertal dans un milieu de psychiatrie le´gale. Abre´ge´s du Congre`s de Psychiatrie et de Neurologie de Langue Franc¸aise. Paris, France: CPNLF, 1994:112. 42. Ohayon MM, Guilleminault C, Zulley J, Palombini L, Raab H. Validation of the Sleep-EVAL system against clinical assessments of sleep disorders and polysomnographic data. Sleep 1999;22:925-930. 43. Rohde P, Lewinsohn PM, Seeley JR. Comparability of telephone and face-to-face interviews in assessing axis I and II disorders. Am J Psychiatry 1997;144:1593-1598. 44. Slutske WS, True WR, Scherrer JF, Goldberg J, Bucholz KK, Heath AC, et al. Long-term reliability and validity of alcoholism diagnoses and symptoms in a large national telephone interview survey. Alcohol Clin Exp Res 1998;22:553558. 45. Dansky BS, Saladin ME, Brady KT, Kilpatrick DG, Resnick HS. Prevalence of victimization and posttraumatic stress disorder among women with substance use disorders: comparison of telephone and in-person assessment samples. Int J Addict 1995;30:1079-1099. 46. Mellman TA, Nolan B, Hebding J, Kulick-Bell R, Dominguez R. A polysomnographic comparison of veterans with combat-related PTSD, depressed men, and non-ill controls. Sleep 1997;20:46-51. 47. Ohayon MM, Priest RG, Caulet M, Guilleminault C. Hypnagogic and hypnopompic hallucinations: pathological phenomena? Br J Psychiatry 1996;169:459-467. 48. Neylan TC, Marmar CR, Metzler TJ, Weiss DS, Zatzick
DF, Delucchi KL, et al. Sleep disturbances in the Vietnam generation: findings from a nationally representative sample of male Vietnam veterans. Am J Psychiatry 1998;155:929-933. 49. Van Der Kolk BA, Blitz P, Burr W, Sherry S, Hartmann E. Nightmares and trauma: a comparison of nightmares after
OHAYON AND SHAPIRO
combat with lifelong nightmares in veterans. Am J Psychiatry 1984;141:187-190. 50. Breslau N, Davis GC, Andreski P, Peterson EL. Traumatic events and posttraumatic stress disorder in an urban population of young adults. Arch Gen Psychiatry 1991;48: 216-222.