Objective: The evidence for a prospective association of vitamin D deficiency with the occurrence of late-life depression is limited. We aimed to study the long-term association between vitamin D serum levels and depression in a large population-based study of older adults.
Method: We included 3251 participants from the Rotterdam Study, aged 55 and older with 32 400 person-years follow-up for depression. Baseline 25-hydroxyvitamin D (25(OH)D) serum levels were analyzed continuously and categorically. Repeated depressive symptoms' questionnaire assessments were used to assess the change of depressive symptoms. Semistructured psychiatric interviews, and GP records were used to assess incident major depressive disorder according to DSM-IV criteria.
Results: Low serum vitamin D levels were cross-sectionally associated with more depressive symptoms. However, low 25(OH)D serum levels were not prospectively associated with change of depressive symptoms (unstandardized beta = 0.02, 95% CI = −0.23; 0.26) or incident MDD (hazard ratio = 0.95, 95% CI = 0.86; 1.05).
Conclusion: We observed a cross-sectional but no prospective association between serum vitamin D levels and depression. A cross-sectional association in the absence of the longitudinal association can mostly be attributed to reverse causality or residual confounding. Probably, vitamin D deficiency is not an independent risk factor for depression but co-occurs with late-life depression.

Additional Metadata
Keywords depression, epidemiology, old age
Persistent URL dx.doi.org/10.1111/acps.12689, hdl.handle.net/1765/97846
Journal Acta Psychiatrica Scandinavica
Story-Jovanova, O, Aarts, N, Noordam, R, Zillikens, M.C, Hofman, A, & Tiemeier, H.W. (2017). Vitamin D serum levels are cross-sectionally but not prospectively associated with late-life depression. Acta Psychiatrica Scandinavica, 135(3), 185–194. doi:10.1111/acps.12689