The association between age and depression in the general population: a multivariate examination

In a large general population study we found a close to linear rise with age in the mean score and prevalence of self-reported symptoms of depression. The aim of this study was to examine if this linear relation prevailed when controlled for multiple variables and to examine factors that eventually explained the association.

Among individuals aged 20-89 years living in Nord-Trøndelag County of Norway, 60 869 filled in valid ratings of the Hospital Anxiety and Depression Scale as well as many other variables. Covariates were grouped into a multivariate model with six blocks. Logistic regression was used to model the blocks and variables with caseness of depression as the dependent variable.

The model explains a considerable part of the age-related pattern on depression. The pattern became less distinct in the age groups above 50 years. Variables within the blocks of somatic diagnoses and symptoms, as well as impairment, had most explanatory power.

Because of our large sample we were able to control for more relevant variables than earlier studies. In contrast to most other studies, we found that an age-related increase of the prevalence of depression persisted after control for multiple variables.