Development of a prediction model for metabolic syndrome based on physical activity and fitness in individuals with physical disabilities.

Abstract

Purpose: The objective of this study was to develop a predictive model to estimate the number of risk factors associated with metabolic syndrome based on physical activity and fitness in individuals with physical disabilities.

Methods: A total of 134 adults aged ≥ 30 years with severe physical disabilities diagnosed over 1 year were enrolled in this study. Standardized procedures were used to collect anthropometric data, blood samples, and physical fitness measurements. Participants were randomly assigned to the derivation (70%) and validation (30%) sets. The derivation set was subjected to a stepwise multiple regression analysis to develop a predictive equation. Criteria and cross-validity were assessed using Bland-Altman plots, and the model's ability to identify metabolic syndrome was evaluated using receiver operating characteristic (ROC) analysis.

Results: The final model included neck circumference, the number of medications, leisure-time physical activity, and muscular strength, with an R² value of 0.397 and a standard error of the estimate of 1.019. The predicted values closely match the measured values for both sets. ROC analysis indicated good to excellent classification performance (derivation set: area under the curve [AUC], 0.867; 95% confidence interval [CI], 0.796-0.937; p < 0.001; validation set: AUC, 0.765; 95% CI, 0.617-0.913; p = 0.009).

Conclusion: A regression model based on physical activity and fitness could provide a simple, non-invasive approach to estimating the risk of metabolic syndrome in individuals with physical disabilities.