Depression evolves dynamically in old age. Studies of natural history of major depression in older adults suggest that 19–34% recover, 27%–32% remain chronically ill, and approximately 40% experience a fluctuating course. Another way of approaching depression from a longitudinal point of view is by adopting a symptom-based approach, that in addition to the evolution of clinically manifested diagnostic entities, also focuses on transitions involving subclinical/subsyndromal states, although few studies have attempted it. We examined psychosocial, behavioral, and clinical determinants of transitions across states that include no depression, subsyndromal-, and clinical depression.

We used data on 3086 adults aged 60+ from the Swedish National Study on Aging and Care in Kungsholmen, followed for 15 years. Markov-state transition models were used to capture transition patterns, as well as their associated determinants. Death and dropout constituted absorbing states. Depression was diagnosed in accordance with DSM-5; SSD was based on having at least 2 symptoms in the absence of DSM diagnosis. Determinants of transition patterns included index of social connections and support (i.e., psychosocial determinants); smoking, alcohol consumption, and physical activity (behavioral determinants); somatic disease burden and history of depression (clinical determinants).

At baseline, 10% of the study population exhibited clinically relevant levels of depressive symptoms. Over a 15-year period, a total of 11,489 transitions were observed. Preliminary results indicate that behavioral factors (primarily smoking) were mostly associated with transitions from no depression to clinical depression, as well as from clinical depression to death. Mostly the same pattern was seen for clinical determinants, although higher burden of chronic diseases and previous depression also increased the likelihood of transition from no depression to SSD. Notably, of high baseline values of social connection and support were found to: 1) lower the likelihood of transitioning from no depression to either SSD or clinical depression; 2) lower the likelihood of transitioning from SSD to clinical depression; and 3) increase the likelihood of transitioning from clinical depression to no depression.

Clinical and behavioral factors are mostly implicated in lowering the occurrence of depression, whereas psychosocial factors may also be implicated in recovery.