ABSTRACT. OBJECTIVE: Pain and depression are both common in old age, but their (long-term) temporal relationship remains unknown. This study is designed to determine whether pain predicts the onset of depression and vice versa. METHODS: This is a prospective, population-based cohort study with 12-year follow-up and 3-year intervals in the Netherlands (Longitudinal Aging Study Amsterdam). At baseline, participants were aged 55 to 85 years (n = 2028). Main measurements outcomes were incident depression defined as crossing the cutoff of 16 and showing a relevant change (≥ 5 points) on the Center for Epidemiological Studies-Depression Scale among nondepressed participants and incident pain defined as a score of 2 or higher on the pain scale of the 5-item Nottingham Health Profile in pain-free participants. Multiple imputations were adopted to estimate missing values. RESULTS: In nondepressed participants (n = 1769), a higher level of pain was predictive of incident depression in multiple extended Cox regression analyses (hazard rate [HR] = 1.13 [95% confidence interval {CI}: 1.05-1.22], p = .001), which all remained significant after correction for sociodemographic characteristics, life-style characteristics, functional limitations, and chronic diseases (HR = 1.09 [95% CI = 1.01-1.18], p = .035). In the pain-free participants (n = 1420), depressive symptoms at baseline predicted incident pain (HR = 1.02 [95% CI: 1.01-1.04], p = .006). This depression measure did not independently predicted the onset of pain in the fully adjusted models. CONCLUSIONS: As pain precedes the onset of depression, strategies to prevent depression in chronic pain patients are warranted. In contrast, no effects of depression on the development of subsequent pain were found when adjusting for covariates.
