Cancer and depression in middle-aged and elderly Chinese: The mediating role of sleep quality and the moderating role of social participation diversity – Evidence from CHARLS 2018

Depression is highly prevalent among individuals with cancer, yet the pathways remain incompletely understood. This study utilized cross-sectional data from the 2018 wave of the China Health and Retirement Longitudinal Study. Cancer was identified from harmonized and health files. Depression was assessed using the 10-item Center for Epidemiologic Studies Depression Scale. Sleep quality was proxied by a single item, which captures one dimension of sleep disturbance. Social participation diversity was constructed from 6 types of activities and summarized as a variety index. Covariates included age, sex, education, partnership, urban residence, retirement, drinking, smoking, and comorbidities. Mediation analysis was conducted using nonparametric bootstrap with 5000 resamples. Moderation was tested via linear models incorporating a sleep quality × social participation interaction term. The analytical sample included 17,550 adults aged 45 years and above. The total effect of cancer on depression was 2.157 (95% confidence interval CI: 0.458–3.856). The indirect effect mediated through poor sleep quality was 0.265 (95% CI: 0.028–0.587, P = .026), accounting for 12.29% of the total effect, indicating that sleep quality partially statistically mediated the relationship between cancer and depression. The direct effect of cancer on depression was 1.892 (95% CI: 0.246–3.538, P = .024). Social participation diversity attenuated the path from sleep to depression (sleep × participation interaction: β = −0.134, standard error = 0.043, P = .002). Among Chinese adults aged 45 and older, sleep quality statistically accounted for a modest portion of the cancer–depression association, and social participation diversity modified the sleep–depression association. Given the cross-sectional design and the single-item sleep measure, findings should be interpreted cautiously with respect to temporality and causality. Interventions addressing sleep problems and supporting social engagement warrant further evaluation in longitudinal and interventional studies.