Types of Social Support and Depressive Symptoms in Older Adults: an IPD Meta‐analysis of Longitudinal Ageing Studies
Suraj Samtani,Gowsaly Mahalingam,Ben C. P. Lam,Darren M. Lipnicki,Erico Costa,Maria Fernanda Lima‐Costa,Shifu Xiao,Nikolaos Scarmeas,Ki Woong Kim,Katya T. Numbers,Mary Ganguli,Murali Krishna,Tze Pin Ng,Perminder S. Sachdev,Henry Brodaty,Ingmar Skoog,Steffi G. Riedel‐Heller,Suzana Shahar
DOI: https://doi.org/10.1002/alz.074525
2023-01-01
Abstract:AbstractBackgroundDepression is a known risk factor for dementia, and social support may lower the risk of depression in older adults. There is, however, a lack of longitudinal analyses examining the association between emotional versus instrumental support and depressive symptoms in older adults. We analyzed the association of emotional and instrumental support with depression symptom scores using data from eight international longitudinal cohorts. This is an individual participant data meta‐analysis on harmonized social support and depression data.MethodWe obtained data from eleven COSMIC (Cohort Studies of Memory in an International Consortium) international longitudinal cognitive ageing studies, including data from Australia, Brazil, China, Germany, Greece, India, Indonesia, South Korea, Singapore, Sweden, and the United States. Randomly sampled 27,058 adults in later life (M = 71.7 years), of whom 59.5% were females. Social support was measured using individual items of emotional support (i.e., having a confidante) or instrumental support (i.e., help with daily living activities). Depression was measured using validated scales at baseline and follow‐up (M = 2.02 years). Covariates included age, sex, education, living situation, smoking history, diabetes mellitus, cardiovascular risk, global cognitive function and (where available) history of depression. Depression scores were standardized while other variables were harmonized before analysis.ResultIn the fully adjusted cross‐sectional models, having emotional support (B = ‐0.40, 95%CI: ‐0.60,‐0.21), but not instrument support (B = 0.15, 95%CI: ‐0.29,0.58) was associated with lower depressive symptoms. In the fully adjusted longitudinal models, having emotional support (B = ‐0.36, 95%CI: ‐0.53,‐0.19) but not having instrumental support (B = 0.07, 95%CI: ‐0.33,0.46) was associated with lower depressive symptoms at follow‐up.ConclusionEmotional, but not instrumental support, was associated with lower depressive symptoms cross‐sectionally and at follow‐up in harmonised data from 8 longitudinal ageing studies. Psychosocial interventions targeted at reducing the risk of depression (and therefore dementia) should focus on improving access to emotional support for older adults.