Do depressive symptoms predict cancer incidence?: 17-year follow-up of the Whitehall II study.

OBJECTIVE To explore the association between depressive symptom history and cancer incidence. METHODS Affective/emotional depressive symptoms were assessed using the General Health Questionnaire (GHQ-30) depression sub-scale across phase 1 (1985-1988), phase 2 (1989-1990), and phase 3 (1991-1994) of the Whitehall II prospective cohort study; 'chronic'=depressive episode at phase 1, 2 and 3; 'new'=depressive episode at phase 3 only. Cancer incidence was obtained from the National Health Service Central Register with an average follow-up of 15.6 years (range 0.08-17.4). The study sample consisted of 6983 participants, aged 35-55 years at baseline. Results were adjusted for age, sex, socio-economic position, health behaviours, health status/conditions, medication, and social support. RESULTS Over a 17.4 year follow-up, chronic depressive symptoms did not increase the risk of cancer incidence compared to those who never experienced symptoms (hazard ratio (HR)=1.03, 95% confidence interval (CI): 0.71-1.49). Participants who experienced new depressive symptoms had an increased risk of cancer incidence in the first 9 years of follow-up (HR=1.89, 95% CI: 1.23-2.90) but no increased risk in later years (HR=0.84, 95% CI: 0.52-1.35). CONCLUSION Chronic depressive symptoms were not associated with cancer incidence. In contrast, new-onset symptoms were associated with a substantially increased risk, possibly due to reverse causality.