Abstract
INTRODUCTION: The study aimed to investigate the associations of changes in social isolation, loneliness, or both, with cognitive function. METHODS: Data were from 7299 older adults in the Chinese Longitudinal Healthy Longevity Survey. We defined four change patterns (no, incident, transient, and persistent) for social isolation and loneliness, and created nine-category variable to represent the joint changes. Tobit regression models and Cox models were performed. RESULTS: Incident, transient, and persistent social isolation or loneliness may accelerate cognitive decline (p < 0.05). Incident, transient, and persistent social isolation were associated with higher cognitive impairment risk, while only persistent loneliness was associated with higher cognitive impairment risk (p < 0.001). Notably, short-term or persistent social isolation was associated with accelerated cognitive decline and incident cognitive impairment, regardless of different loneliness change status (p < 0.05). DISCUSSION: Short-term or persistent social isolation and persistent loneliness may be a salient risk factor for cognitive decline and cognitive impairment. HIGHLIGHTS: Incident, transient, and persistent social isolation were associated with accelerated cognitive decline and higher cognitive impairment risk. Persistent loneliness was associated with accelerated cognitive decline and higher cognitive impairment risk. Short-term or persistent social isolation with concurrent different loneliness change status accelerated cognitive decline and higher cognitive impairment risk.