Absenteeism from work and inpatient treatment are well defined events in the course of any illness and may be indicators for the need of an intervention. The aim of this study was to estimate of the influence of these events on later permanent disability and death in a working population when caused by anxiety and/or depression.

Data of 128,001 members of a health insurance with a observation period of 6.4 years were used. Excess risks were calculated for patients suffering from anxiety, depression or both using Cox's proportional hazard models adjusted for age, gender, education and job classification.

Patients who received outpatient treatment had higher rates of permanent disability (hazard ratio (HR) 1.48 (Confidence interval 1.30, 1.69) for depression, 1.25 (1.07, 1.45) for anxiety) but lower premature mortality (HR 0.80(0.62, 1.03), 0.53(0.38, 0.73)) than controls. Inpatient treatment and co-morbidity were associated with further raised rates of permanent disability. Depressed inpatients also had higher premature mortality (HR 2.50(1.80, 3.48)).

The study shows a dose-responsive relationship between depression, anxiety, comorbidity and need for hospital treatment. It was was unexpected that some exposures seem to reduce mortality risk. Life style may partly explain this effect. However, these patients are subject to many other exposures, into which more research is needed to get a better understanding of the phenomenon.