Influencing factors, gender differences and the decomposition of inequalities in cognitive function in Chinese older adults: a population-based cohort study

Abstract

Background Evidence remains limited and inconsistent for assessing cognitive function in Chinese older adults (CFCOA) and inequalities in cognitive function in Chinese older adults (ICFCOA) and exploring their influencing factors and gender differences. This study aimed to identify influencing factors and inequality in CFCOA to empirically explore the existence and sources of gender differences in such inequality and analyse their heterogeneous effects. Methods Based on data from the China Health and Retirement Longitudinal Study (CHARLS) for three periods from 2011 to 2015, recentered influence function unconditional quantile regression (RIF-UQR) and recentered influence function ordinary least squares (RIF-OLS) regression were applied to assess influencing factors of CFCOA, while grouped treatment effect estimation, Oaxaca-Blinder decomposition, and propensity score matching (PSM) methods were conducted to identify gender differences in ICFCOA and influencing factors, respectively. Results The results showed heterogeneous effects of gender, age, low BMI, subjective health, smoking, education, social interactions, physical activity, and household registration on CFCOA. Additionally, on average, ICFCOA was about 19.2–36.0% higher among elderly females than among elderly males, mainly due to differences in characteristic effects and coefficient effects of factors such as marital status and education. Conclusions Different factors have heterogeneous and gender-differenced effects on CFCOA and ICFCOA, while the formation and exacerbation of ICFCOA were allied to marital status and education. Considering the severe ageing and the increasing incidence of cognitive decline, there is an urgent need for the government and society to adopt a comprehensive approach to practically work for promoting CFCOA and reducing ICFCOA.