“…Results are from proportional odds models. Results are displayed as ORs of change in sleep duration (taking 7-9 hours as reference) per unit increase in the original scale of generalized trust, visits and interaction with friends, participation of organized social activities, emotional social support, or instrumental social support a Adjusted for gender, age, education, marital status, current residential area, household annual income, living alone, and pension status b Adjusted for gender, age, education, marital status, current residential area, household annual income, living alone, pension status, smoking, alcohol drinking, and physical activity c Adjusted for gender, age, education, marital status, current residential area, household annual income, living alone, pension status, smoking, alcohol drinking, physical activity, sleep quality, BMI, depression, ADL, cognitive function, NCD, and comorbidity *p < 0.05; **p < 0.01; ***p < 0.001 in rural areas would suffer more from depression [63], which was found associated with higher risk of longer sleep duration and daytime napping in previous research [64,65]. Differences in family responsibilities have led to significant gender differences in sleep outcomes, such as longer self-reported sleep in women [66].Women in Chinese society, especially in rural areas, always spend more time to give family support such as unpaid care responsibilities for grand-children and spouse [67], which makes their social capital more concentrated on their family and relatives [68].…”