Subgroups of mild cognitive impairment based on cognitive function among Chinese older adults: a latent classes analysis

Abstract

Abstract Background: Due to the heterogeneity of MCI, the key points of non-drug therapies cannot be determined. Exploring the characteristics among different MCI subgroups would therefore be beneficial in guiding the development of non-drug therapies. In the present study, weexplored subgrouping of older Chinese adults with MCI based on cognitive function and analyzed the characteristics of the resulting subgroups. Methods: A cross-sectional study was conducted between October 2018 and January 2022. The study enrolled a total of 1017 participants with MCI in Fujian, East China. Participants underwent a comprehensive neuropsychological test assessing memory, language, attention, and executive function and completed a questionnaire examining socioeconomic and clinical variables.Neuropsychological test scores were transformed to z-scores, which were then examined using latent class analysis. Correlations of socioeconomic and clinical variables with MCI subgroups were assessed using multinomial logistic regression analysis. Results: Three MCI subgroups were identified: (1) false-positive MCI (56.64%); (2) early MCI (36.28%); and (3) late MCI (7.08%). Age, data sources, marriage status, living status, and participation in intellectual activities differed significantly across the three MCI subgroups (P<0.05). Compared with the false-positive MCI group, participants with late MCI were less likely to participate in intellectual activity and more likely to live in a nursing home. Conclusions: Older Chinese adults with MCI can be classified into subgroups based on cognitive function. A large proportion of participants were classified in the false-positive MCI group, suggesting that conventional diagnostic criteria are susceptible to false-positive errors, resulting in patients being misdiagnosed as MCI in China. And further research to define the characteristics of different MCI subgroups could facilitate the development of non-drug therapies for MCI.