Association between Serum Creatinine and Cognitive Impairment in US Older Adults

Background: and aims: The global rise in the aging population has brought cognitive impairment to the forefront as a prevailing concern. Serum creatinine (Cre), an indicator of renal function, has been proposed as a potential predictor of cognitive impairment. However, the existing clinical literature lacks a unified conclusion on the association between serum Cre and cognitive impairment. Methods We utilized data from the National Health and Nutrition Examination Survey conducted from 2011 to 2014 to investigate the association between serum Cre and cognitive impairment. Cognitive function was assessed using three tests: Consortium to Establish a Registry for Alzheimer's Disease (CERAD), Animal Fluency (AF), and Digit Symbol Substitution Test (DSST). Cognitive impairment was defined based on cutoff values: CERAD < 5, AF < 14, and DSST < 34. A cross-sectional study employing a weighted multiple logistic regression model was conducted to demonstrate the association between serum Cre and cognitive impairment. To explore any potential nonlinear relationship, generalized additive models (GAM) and a smooth curve fit utilizing the penalty spline method were applied. Subgroup analysis and interaction tests were performed using the likelihood ratio test. Results A total of 2771 participants aged ≥ 60 years were included in the study. The data revealed an association between serum Cre levels and an increased risk of cognitive impairment, as measured by the DSST test [full adjustment, OR = 1.004, 95% CI: 1.001–1.006]. Additionally, the Q4 group in the AF test showed an elevated risk of cognitive impairment [full adjustment, OR = 1.445, 95% CI: 1.012–2.062]. To analyze the model, a two-piecewise linear regression model was employed. The calculated inflection point for serum Cre was 138.79 µmol/L. Participants falling on the left of this inflection point or with serum Cre levels less than 138.79 µmol/L, based on the AF test, exhibited an OR value of 1.007 (95% CI: 1.001, 1.013) and a p-value of 0.03, indicating a significant association with cognitive impairment. Furthermore, subgroup analysis and interaction tests demonstrated a significantly increased risk of cognitive impairment associated with serum Cre levels in female participants, as observed in both the AF and DSST tests. Conclusions The cross-sectional association between serum Cre and cognitive impairment was found to be non-linear, with a threshold level of 138.79 µmol/L. In order to further validate our findings and explore possible mechanisms, upcoming large-scale prospective clinical trials will employ a greater number of samples and utilize highly effective data collection methods.