Objective: To find a suitable dividing value to classify cystatin C and evaluate the association between cognition and levels of cystatin C.
Methods: Using data from the China Health and Retirement Longitudinal Study, We conducted a longitudinal analysis of a prospective cohort of 6,869 middle-aged and older Chinese without cognitive impairment at baseline. Levels of cystatin C were categorized into 2 groups by method of decision tree. Logistic regression models evaluated whether cystatin C was related to cognitive impairment.
Results: Respondents were categorized as lower levels of cystatin C and higher levels of cystatin C, cut-point was 1.11 mg/L. Higher levels of cystatin C was associated with the odds of cognitive impairment (OR, 1.56; 95% CI, 1.10-2.22) after multivariable adjustment. Respondents with higher levels of cystatin C had worse cognition scores.
Conclusions: We found a suitable dividing value of cystatin C in middle-aged and older Chinese.
Keywords: cognitive impairment; cystatin C; decision tree; logistic regression models; longitudinal studies.