Social Isolation and Depressive Symptoms Among Older Adults: A Multiple Bias Analysis Using a Longitudinal Study in Japan

Abstract

PurposeDespite the potential burden of social isolation on mental health, it is unclear which component of social isolation (living alone, lack of social support, and lack of social participation) is mostly associated with depression among older adults and whether it varies by gender. Moreover, previous observational studies have suffered from several systematic biases such as confounding, misclassification, and selection bias. We thus aimed to investigate the risk of social isolation on subsequent depressive symptoms among older men and women, fully taking account of such systematic biases simultaneously. MethodsThis study included 15,311 adults from a nationwide longitudinal cohort study of older adults aged ≥65 years in Japan. We employed modified Poisson regression models to estimate the adjusted risk ratio (aRR) of elevated depressive symptoms (the Geriatric Depression Scale: ≥5 vs. <5) in 2016 according to the social isolation index (SII) score (0-5 points) in 2013. Non-probabilistic multiple bias analyses with weighting approach and bounding factor were conducted to adjust for possible confounding, exposure misclassification, and selection biases. ResultsThe mean age (SD) of participants was 72.1 (4.9) years. Adults with the SII score ≥2 showed an increased risk of elevated depressive symptoms than those with the SII score <2 (aRR [95% CI] = 1.57 [1.41–1.76]). All components of SII were associated with elevated depressive symptoms except living alone for women. Our multiple bias analysis showed that an unmeasured confounder needed to have at least moderate relationship (RR=2.0-3.0) with both exposure and outcome to explain away the observed association particularly when social isolation status was independently and non-differentially misclassified. ConclusionAmong older adults in Japan, social isolation was associated with an increased risk of elevated depressive symptoms, even after taking account of systematic biases simultaneously. Increasing social connection may mitigate the risk of depression, which should be the subject of future research.