Frontiers in Public Health · May 30, 2025

Association between atherogenic index of plasma and physical dysfunction: a cross-sectional study of middle-aged and older adults in China

By Zhou J, Chang Y, Shen H et al.

BACKGROUND: Physical dysfunction is common in older adults and increases disease risk. The atherogenic index of plasma (AIP) is a promising biomarker for this condition. This study explored the dose-response relationship between AIP and physical dysfunction. METHODS: Data from 11,369 CHARLS participants (aged ≥45 years) in 2015 and 2018 were analyzed using univariate and multivariate logistic regression, adjusting for demographics and lifestyle factors. The restricted cubic splines were used to examine possible non-linear associations and visualize the dose-response relationship between AIP and physical dysfunction. ROC curve analysis assessed AIP's predictive performance, and subgroup analyses evaluated interactions. RESULTS: Each interquartile range (IQR) increase in AIP was associated with a 13.4% higher odds of physical dysfunction (adjusted OR = 1.134, 95% CI: 1.066-1.207, p < 0.001), with a dose-response threshold identified at an AIP value of approximately 0.37. Beyond this threshold, the odds of physical dysfunction increased steadily, confirming a non-linear relationship. AIP exhibited moderate predictive accuracy for physical dysfunction (AUC = 0.748, 95% CI: 0.738-0.758). Stratified analysis showed AIP was significantly linked to higher physical dysfunction risk in subgroups including those aged <65, females, married individuals, high school or college-educated, rural residents, non-smokers, and non-drinkers (p < 0.05), with no significance in other subgroups. Interaction analysis identified marital status (p = 0.035) and education level (p = 0.034) as significant effect modifiers, where subgroup differences notably altered the AIP-dysfunction association, warranting further study. CONCLUSION: Elevated AIP is significantly associated with increased physical dysfunction risk, highlighting its potential as a simple, predictive biomarker.

doi.org/10.3389/fpubh.2025.1580340Open accessPubMed 40520296Biomarkers