Objective To analyze the factors influencing the cognitive function of community-dwelling elderly in Haishu district of Ningbo city, Zhejiang province, in order to provide a reference for the prevention and early intervention of dementia in the elderly.

Methods A total of 3 850 residents aged 60 years and older were recruited from urban/township communities in Haishu district of Ningbo city, Zhejiang province using a stratified proportional sampling method for a face-to-face survey from April to September 2022. The participants' information was collected using a self-designed questionnaire, and the Mini-Mental State Examination (MMSE) was used to assess the participants' cognitive function. The structural equation model was used to analyze the pathway and the effect of the influencing factors on the cognitive function of the participants.

Results The mean MMSE score of the participants was 26.38 ± 4.14 with a range of 0 to 30. The Spearman correlation analysis showed that MMSE score was positively correlated with educational level (r = 0.551), sleep quality (r = 0.242), family monthly income (r = 0.237), physical activity (r = 0.145), alcohol consumption (r = 0.110) and smoking (r = 0.064) (all P < 0.001) , but inversely correlated with age (r = – 0.379), hearing loss (r = – 0.378), being unmarried (r = – 0.224), hyposmia (r = – 0.199), being female (r = – 0.116), and being a farmer before retirement (r = – 0.089), (all P < 0.001). The constructed structural equation model was of good fitness to the data, with the χ²/df of 7.359, the goodness-of-fit index of 0.995, the adjusted goodness-of-fit index of 0.983, the normed fit index of 0.985, the Tacker-Lewis index of 0.965, the comparative fit index of 0.987, the incremental fit index of 0.987, and the root mean square error of approximation of 0.041, respectively. Direct effect analysis revealed that higher education (β = 0.355), good sleep quality (β = 0.108), participation in physical activity (β = 0.068), and higher monthly household income (β = 0.062) were significantly associated with higher MMSE scores (all P < 0.001), whereas hearing loss (β = – 0.227), older age (β = – 0.213), and hyposmia (β = – 0.112) were associated with lower MMSE scores (all P < 0.001). In addition, age, education level, family monthly income also indirectly affected MMSE scores, with indirect and total effect coefficients of – 0.257 and – 0.470, 0.056 and 0.411, and 0.012 and 0.074, respectively.

Conclusion The cognitive function of elderly people in Haishu district, Ningbo city is affected by various factors such as age, education level, family monthly income, sleep quality, physical activity, and hearing loss and hyposmia.