Search

Objective: Most research on the association between sleep disturbances and depression has looked at cross-sectional data. The authors used two waves of data from a panel study of community residents aged 50 years or more to investigate this issue prospectively. Method: Data on symptoms of major depressive episodes and sleep problems were examined for a subgroup of the 1994 and 1995 surveys of the Alameda County (California) Study (N=2,370). The authors examined the effects of age, gender, education, marital status, social isolation, functional impairment, financial strain, and alcohol use. Depression was measured with 12 items that covered the DSM-IV diagnostic criteria for major depressive episodes, including insomnia and hypersomnia. Results: The prevalences were 23.1% for insomnia and 6.7% for hypersomnia in 1994. Sleep was a significant correlate of depression, as were being female, older age, social isolation, low education, financial strain, and functional impairment. When sleep problems and depression were examined prospectively, with controls for the effects of the other variables, sleep problems in 1994 predicted depression in 1995. However, other symptoms of major depressive episodes - anhedonia, feelings of worthlessness, psychomotor agitation/retardation, mood disturbance, thoughts of death - were much stronger predictors of future major depression. Conclusions: Sleep disturbance and other symptoms that are diagnostic for major depression are strongly associated with the risk of future depression. Sleep disturbance appears to be a less important predictor of depression. More epidemiologic research is needed on the relative contributions of the range of depressive symptoms to the risk of clinical depression.

Sleep disturbance is an important dimension of posttraumatic stress disorder (PTSD), but most of the limited available data were obtained years after the original traumatic event. This study provides information on sleep disturbance and its relationship to posttraumatic morbidity from evaluations done within a year after the trauma. Method: Sleep and psychiatric symptoms of 54 victims (12 men and 42 women) of Hurricane Andrew who had no psychiatric illness in the 6 months before the hurricane were evaluated. A subset of hurricane victims with active psychiatric morbidity (N=10) and nine comparison subjects who were unaffected by the hurricane were examined in a sleep laboratory. Results: A broad range of sleep-related complaints were rated as being greater after the hurricane, and psychiatric morbidity (which was most commonly PTSD, followed by depression) had a significant effect on most of the subjective sleep measures. In addition, subjects with active morbidity endorsed greater frequencies of 'bad dreams' and general sleep disturbances before the hurricane. Polysomnographic results for the hurricane victims revealed a greater number of arousals and entries into stage 1 sleep. REM density correlated positively with both the PTSD symptom of reexperiencing trauma and global distress. Conclusions: Subjects affected by Hurricane Andrew reported sleep disturbances, particularly those subjects with psychiatric morbidity. Tendencies to experience bad dreams and interrupted sleep before a trauma appear to mark vulnerability to posttraumatic morbidity. Results of sleep laboratory evaluations suggested brief shifts toward higher arousal levels during sleep for PTSD subjects and a relationship of REM phasic activity and symptom severity.

Objective: Disturbed sleep is a common complaint among patients with posttraumatic stress disorder (PTSD) that appears in the reexperiencing and hyperarousal symptom clusters in DSM-IV. The causal relationship between sleep complaints and PTSD is unclear. Method: Self-reported insomnia and excessive daytime sleepiness were assessed in 102 victims of motor vehicle accidents and 19 comparison subjects 1 week and 1, 3, 6, and 12 months after the trauma. At 12 months the Structured Clinical Interview for DSM-III-R was administered to determine diagnoses of PTSD. Results: Twenty-six of the accident victims but none of the comparison subjects met the criteria for PTSD. Logistic regression models indicated that sleep complaints from 1 month on were significant in predicting PTSD at 1 year. Conclusions: These results suggest that on the basis of sleep complaints as early as 1 month after the trauma, it is possible to detect subjects who will later develop chronic PTSD. (Am J Psychiatry 2002; 159:855-857)