The authors tested the cognitive vulnerability hypotheses of depression with a retrospective behavioral high-risk design. Individuals without current Axis I diagnoses who exhibited either negative or positive cognitive styles were compared on lifetime prevalence of depressive and other disorders and the clinical parameters of depressive episodes. Consistent with predictions, cognitively high-risk participants had higher lifetime prevalence than low-risk participants of major and hopelessness depression and marginally higher prevalence of minor depression. These group differences were specific to depressive disorders. The high-risk group also had more severe depressions than the low-risk group, but not longer duration or earlier onset depressions. The risk group differences in prevalence of depressive disorders were not mediated by current depressive symptoms.

Consistent with the BAS hypersensitivity model of bipolar disorder, a highly responsive BAS provides vulnerability to onsets of (hypo)manic episodes. In addition, a highly sensitive BIS increases risk for major depressive episodes.

Do negative cognitive styles provide similar vulnerability to first onsets versus recurrences of depressive disorders, and are these associations specific to depression? The authors followed for 2.5 years prospectively college freshmen (N = 347) with no initial psychiatric disorders at high-risk (HR) versus low-risk (LR) for depression on the basis of their cognitive styles. HR participants had odds of major, minor, and hopelessness depression that were 3.5-6.8 times greater than the odds for LR individuals. Negative cognitive styles were similarly predictive of first onsets and recurrences of major depression and hopelessness depression but predicted first onsets of minor depression more strongly than recurrences. The risk groups did not differ in incidence of anxiety disorders not comorbid with depression or other disorders, but HR participants were more likely to have an onset of anxiety comorbid with depression.

The hypothesis that sleep deprivation depresses immune function was tested in 20 adults, selected on the basis of their normal blood chemistry, monitored in a laboratory for 7 d, and kept awake for 64 h. At 2200 h each day measurements were taken of total leukocytes (WBC), monocytes, granulocytes, lymphocytes, eosinophils, erythrocytes (RBC), B and T lymphocyte subsets, activated T cells, and natural killer (NK) subpopulations (CD56/CD8 dual-positive cells, CD16-positive cells, CD57-positive cells). Functional tests included NK cytotoxicity, lymphocyte stimulation with mitogens, and DNA analysis of cell cycle. Sleep loss was associated with leukocytosis and increased NK cell activity. At the maximum sleep deprivation, increases were observed in counts of WBC, granulocytes, monocytes, NK activity, and the proportion of lymphocytes in the S phase ofthe cell cycle. Changes in monocyte counts correlated with changes in other immune parameters. Counts of CD4, CD16, CD56, and CD57 lymphocytes declined after one night without sleep, whereas CD56 and CD57 counts increased after two nights. No changes were observed in other lymphocyte counts, in proliferative responses to mitogens, or in plasma levels of cortisol or adrenocorticotropin hormone. The physiologic leukocytosis and NK activity increases during deprivation were eliminated by recovery sleep in a manner parallel to neurobehavioral function, suggesting that the immune alterations may be associated with biological pressure for sleep. (J. Clin.

A prospective, behavioral high-risk design provided a theoretically guided examination of vulnerability to first onset of bipolar spectrum disorder based on the Behavioral Approach System (BAS) model. Adolescents (ages 14–19) at an “age of risk” for bipolar disorder onset were screened on BAS sensitivity by interviewers blind to current symptoms, lifetime history, and family history of psychopathology. Participants were selected with high versus moderate levels of BAS sensitivity and administered a lifetime diagnostic interview. Those with a bipolar spectrum disorder, psychosis, or hypomanic episode with onset prior to the BAS sensitivity assessment were excluded. High BAS (n = 171) and Moderate BAS (n = 119) sensitivity participants in the final sample completed baseline measures of symptoms, goal-setting, and reward responsiveness and were followed prospectively with semistructured diagnostic interviews every 6 months. Consistent with the vulnerability hypothesis of the BAS model of bipolar disorder, high BAS participants had a greater likelihood, and shorter time to onset, of bipolar spectrum disorder than moderate BAS participants across an average of 12.8 months of follow-up (12.9% vs. 4.2%), controlling for baseline depressive and hypomanic symptoms, and family history of bipolar disorder. High reward responsiveness on a behavioral task and ambitious goal-striving for popular fame and financial success (but not impulsivity) also predicted first onset of bipolar spectrum disorder controlling for the covariates and BAS risk group, and ambitious goal-striving partially mediated the BAS risk group effect. We discuss implications of the findings for the BAS model of bipolar disorder and early intervention efforts.