A cross-sectional study of the major risk factor at different levels of cognitive performance within Chinese-origin middle-aged and elderly individuals.

Senior citizens suffering from cognitive impairment (CI) are on the East Asia rise. Multiple variables could lead to inter-/intra-individual cognition effectiveness variations, though previous research efforts did not consider weighting issues. This study scrutinized 5639 participants meeting required inclusion criteria by the CHARLS. Cognitive capacity was evaluated through Mini-Mental State Examination (MMSE). Considering that MMSE scorings were not following normal distribution, a non-parametric test and multiple linear regression were performed to screen candidate variables linked to cognitive capacity. Such applicability of candidate factors in the cumulative effect and the weighting of the impact on cognitive performance were evaluated by random forest (RF) algorithm. Age, gender, education, marital status, residence, the type of residence, exercise, socialization level and drinking were correlated to MMSE scorings (p < 0.05). Among them, age, education, gender and sociality were correlated to individual MMSE items (p < 0.05). Regardless of MMSE scores and several MMSE items, age is always a prime factor. However, in the attention and computation item, education is better than age and ranks first. This preliminary study prompted age, education, gender, and sociality with varying weightings to be linked to cognitive capacity within a Chinese cohort by differing cognitive aspects. At different levels of cognitive performance, the main risk factors are basically similar, but there are still some differences. • The forecast accuracy between sociodemographic/lifestyle factors and MMSE score, together with weighting of each candidate factor in whole were initially analyzed for the first time. • Age, education, gender and social status were correlated to MMSE scores and individual MMSE items with differing weights. [ABSTRACT FROM AUTHOR]