Unhealthy lifestyle factors and the risk of incident depressive symptomatology in older adults: Findings from the Three-City study

BACKGROUND: The association between unhealthy lifestyle factors and incident depressive symptoms has seldom been studied in the specific context of late-life depression. METHODS: Data were drawn from the Montpellier and Dijon centres of the Three-City study with community-dwellers (aged ≥65 years) followed at five time-points over 12 years. Analysis was performed on 3495 dementia-free participants, with no history of major depression, no antidepressant use and a CES-D score < 16 at inclusion. Lifestyle factors at inclusion included alcohol consumption, smoking, sleep disturbances, diet, social activity engagement and physical activity level. A cumulative unhealthy lifestyle index was built with one-point allocated per unhealthy lifestyle factor. Incident depressive symptomatology (DepS) was defined as a CES-D score ≥ 16 or antidepressant use. RESULTS: Over the median 9.02-year follow-up (range: 1.56-12.75), 922 participants (26.4 %) had incident DepS. Regarding the cumulative index, the DepS risk increased by 18 % [95 % CI, 12 %-16 %], per additional UHL risk factor. In the multi-adjusted model including the 6 factors, smoking (HR, 1.70 [95 % CI, 1.33-2.17]), sleep disturbances (HR, 1.37 [95 % CI, 1.20-1.57]) and low social activity engagement (HR, 1.20 [95 % CI, 1.04-1.39]) were independently associated with DepS. The risk increased for all two-by-two combinations of these three factors, the highest being for sleep disturbances combined with smoking (HR, 2.84 [95 % CI, 1.82-2.45]). Of those occurring in isolation, only sleep disturbances reached significance (HR, 1.18 [95 % CI, 1.00-1.39]). CONCLUSION: The risk of DepS onset increased hand-in-hand with the number of unhealthy lifestyle factors. Interventions targeting unhealthy lifestyles, and in particular sleep disturbances, smoking and social isolation, could have a substantial public health impact on reducing late-onset depression.