Mortality and work-related disability as long-term consequences of anxiety and depression: historical cohort designs based on the HUNT-2 study

Objective: Over the last decades, quality epidemiological studies have repeatedly shown that the prevalence of mental illness is substantial, but still largely under-recognised and undertreated. Under-recognition in general health care might also explain the modest attention to possible detrimental effects of such illness in terms of disability and mortality, and the mechanisms it operates through, compared to similar issues in other important fields of public health. This thesis is based on three papers, all focusing on grave outcomes of two of the most common mental illnesses; anxiety and depression. First, we investigated empirically the contribution of psychiatric morbidity to the award of disability pensions. With the purpose of finding out if anxiety and depression are under-recognized as risk factors for disability pension award, we examined if they had an effect independent of awards for physical diagnoses. Effects of mental disorders on disability pension award are generally estimated by aggregation of diagnoses reported in medical certificates underlying applications for disability pension, which is vulnerable for biases including the above mentioned under-recognition of mental disorders. There are numerous reports of effects of depression on general mortality, and specifically for cardiovascular diseases (CVD) and suicide. The second and third papers address six uncertainties in this literature concerning (i) residual confounding, (ii) doseresponse effect of severity of depression, (iii) mechanisms driving the association, (iv) whether there is an effect of mortality beyond CVD and suicide, (v) effect-moderation by gender, and (vi) effect of comorbid anxiety