RESUMO
OBJECTIVES: To explore the associations of social support, and cognitive activity with cognitive impairment incidence, and further examine the mediation effect of cognitive activity on the association between social support and cognitive impairment incidence based on a nationwide elderly Chinese cohort. METHODS: We collected the participants from an ongoing cohort of the Chinese Longitudinal Healthy Longevity Survey (CLHLS). A total of 9394 older adults aged 65 or more years and free of cognitive impairment who participated in the CLHLS between 2008 and 2018 were included. The information on social support and cognitive activity was collected through a questionnaire. The incident cognitive impairment cases were identified through the Mini-Mental State Examination scale (MMSE). Cox proportional hazard regression models were conducted to calculate the hazard ratios (HRs) and 95% confidence interval (CI) of social support and cognitive activity associated with cognitive impairment. We used casual mediation models to assess the indirect association of cognitive activities underlying the association between social support and cognitive impairment. RESULTS: The adjusted HRs (95% CI) of incident cognitive impairment were 0.956 (0.932 to 0.980), and 0.895 (0.859 to 0.933) associated with per 1 score increase in social support and cognitive activity score, respectively. Better adherence to social support was associated with a higher cognitive activity score (adjusted ß = 0.046, 95% CI[0.032-0.060]). The baseline cognitive activity, as well as the mean cognitive activity at baseline and during the first follow-up wave, mediate the association between social support and the incidence of cognitive impairment, accounting for 11.4% and 12.6% of the total association, respectively. The participants who were aged 80 years or older, or those with mild daily functional limitations gained more benefits in the development of cognitive activity related to social support, leading to a reduction in the risks of cognitive impairment. CONCLUSION: The results of this nationwide cohort provide consistent evidence linking social support, and cognitive activity to reduced risk of subsequent cognitive impairment incidence. These findings provide additional evidence to inform the social strategies to prevent cognitive impairment incidence in elderly people.