Social Isolation and Cognitive Functioning: A Quasi-Experimental Approach

Abstract

Abstract Objectives This study aimed to assess the relationship between social isolation and cognitive functioning. Method Data were retrieved from the National Survey of the Japanese Elderly, a nationally representative sample of Japanese adults, aged 60 years or older. We estimated a social isolation index to incorporate variables, such as social interactions, social engagement, and social support, with perceived social isolation, for a comprehensive measurement. The association of social isolation with cognitive functioning was assessed using a panel data fixed-effects model, controlling for age, socioeconomic status, health-related variables, and time-invariant heterogeneity. Moreover, we conducted analyses using the System Generalized Method of Moments (GMM) to address the dynamic relationship of cognitive functioning and potential endogeneity. Results For both men and women, the association between social isolation and cognitive functioning was significant, particularly among those aged 75 or older, as a 1% increase in social isolation was associated with decreased cognitive functioning (24% decrease for men and 20% decrease for women). However, this association was not confirmed by the System GMM, after addressing endogeneity. Discussion Our findings potentially suggest that the association between social relationship and cognitive functioning reported in previous studies was biased, due to endogeneity. Although we did not observe causal impacts, this does not necessarily mean that social isolation does not have a negative impact on health, as both positive and negative consequences of social relationships may exist. Further research is needed to reveal the causal relationship, as well as the detailed mechanisms of health effects of social relationships.