Unipolar depression, estimated to be the leading contributor to burden of disease in middle- and high-income countries, often has an onset in adolescence. The disorder is associated with substantial role impairment and is highly recurrent. This raises questions about both subsequent mental health and social outcome. In order to shed light on this, a community sample of adolescents with depression and non-depressed peers was followed-up after 15 years. In 1991-93, first-year students in upper secondary school (age 16-17) in the town of Uppsala, Sweden, were screened for depression. Adolescents with positive screening and selected peers with negative screening (n=631 in total) were assessed regarding mental health, social situation, and personality. At around age 31, the participants were followed-up in both national registers (n=609) and personal interviews (n=409). Outcome regarding social factors, mental health, and personality was assessed. At follow-up, the former depressed adolescents had completed higher education to a lesser extent than the former non-depressed adolescents. The females with adolescent depression were also at increased risk of subsequent abortion, divorce, single parenthood, and partner violence. Characteristics associated with depression in adolescence (such as poor school performance and disruptive disorders) seemed to contribute to the poor outcome in the social domain. Regarding adult mental health, long-term depression in adolescence was associated with a particularly poor outcome. Compared to adolescents with shorter episodes of depression, those with long-term depression were more likely to report recurrent depression, suicidal ideation, and a range of other mental disorders in adulthood. Measures of personality traits related to neuroticism (a tendency towards negative emotionality) were elevated during ongoing depression and anxiety disorders, but were normalized with remission. However, repeated depressive episodes seemed to leave the individual more vulnerable to stress. It is now important to assess if early treatment can alter the poor outcome depicted in this thesis. Since social adversity, educational difficulties, and interpersonal problems accompany the depressive disorder from adolescence onward, it should also be investigated if interventions aimed at such contextual factors can prevent recurrence and improve quality of life.

Bipolar disorders and recurrent depressions are two common psychiatric disorders with a life time prevalence of approximately 1% and 8%, respectively. Despite treatment these patients suffer from affective symptoms up to 50% of the time, resulting in lower well-being. The average life length is also reduced with 10-15 years, mainly attributable to suicide and cardiovascular disease. Increased stress is one of many factors that have been shown to be linked to an increased risk for developing affective disorders and some comorbid somatic conditions such as metabolic disturbances and cardiovascular disease. An increased stress level is known to cause hyperactivity of the hypothalamic-pituitary-adrenal-axis (HPA-axis) with increased cortisol secretion. Hyperactivity of the HPA-axis (or hypercortisolism) is one of the most replicated neurobiological finding in depression. In other stress related disorders it has however been shown that prolonged stress over long periods of time can lead to a state of low HPA-axis activity, hypocortisolism. Since persons with recurrent affective disorders such as bipolar disorder and recurrent depression are exposed to a high degree of recurrent and chronic stress it could be expected that in addition to hypercortisolism, a state of hypocortisolism could also develop in these disorders, potentially exerting an influence upon the psychological and somatic wellbeing among these patients. The major aim of this thesis was to evaluate whether hypocortisolism is related to relevant psychiatric and somatic phenotypes in recurrent affective disorders. In bipolar disorder, individuals with hypocortisolism exhibited a higher degree of depression and low quality of life compared to patients with normal HPA-axis activity. In recurrent depression, individuals with hypocortisolism exhibited shorter leukocyte telomere length than patients with normal or high HPA-axis activity, which is an indication of an accelerated aging process. In a sample of both bipolar and recurrent depression patients, hypocortisolism was associated with an increased proportion of obesity, dyslipidemia and metabolic syndrome compared with patients with normal or high HPA-axis activity. Patients with recurrent depression showed a higher occurrence of hypocortisolism than the control sample representative of the general population. Patients with bipolar disorder showed a similar occurrence of hypocortisolism as the control sample. Among bipolar disorder patients with a low degree of lifetime with lithium prophylaxis, there was an inverse correlation between age and HPA-axis activity. In contrast, among patients with a higher degree of lifetime with lithium prophylaxis as well as among the controls, there was no correlation between age and HPA-axis activity. Accordingly, hypocortisolism was most common among older patients with a low degree of lifetime with lithium prophylaxis. In conclusion, hypocortisolism in both recurrent depression and bipolar disorder was associated with multiple clinically-relevant phenotypes. Additionally it was shown for bipolar disorder patients that increasing age was a risk factor for hypocortisolism and that prophylactic lithium treatment was a protective factor. It is argued that the protective effect of lithium towards the HPA-axis is attributable to its mood-stabilizing effect, which in turn reduces the chronic stress level. These results provide new insight into the role of hypocortisolism and chronic stress in recurrent affective disorders warranting further studies and hopefully providing clues to improved treatment strategies.

Diabetes disproportionately affects minority populations. Social support, and more specifically marriage, has been found to buffer the negative effects of diabetes and depression. Data collected from African Americans with type 2 diabetes in Atlanta and NHANES data were compared to examine whether marriage affects health status and mental health. Approximately, 1742 African Americans aged 18-80 were included in this study. Chi square analysis revealed that married men had lower rates of depression (15.9% vs. 24.7%) compared to unmarried men (p < 0.05), but the same effect was not found in women. The findings show that marriage was not associated with HbA1c, but was associated with rates of depression. The lack of association of HbA1c with marriage may be because marriage may not be the best proxy of social support in the African American community. Future research should focus on alternative forms of social support such as cohabitation, extend family, and friend.