Waist circumference versus BMI: A cross-sectional study comparing Cardiometabolic Risk association in the Nepalese
Sushan Gautam1, Abhinav Bhattarai2, Simran K C1
1Institute of Medicine, Tribhuvan University, Kathmandu, Nepal.
Background:
Obesity is a major cardiometabolic risk factor. Although body mass index (BMI) and waist circumference (WC) are established predictors of cardiometabolic risk, their predictive value may vary across ethnic groups. South Asians, including Nepalese individuals, exhibit a distinctive "thin-fat" phenotype characterized by increased central adiposity and higher body fat despite relatively normal BMI levels. Therefore, internationally validated anthropometric cutoffs may not accurately reflect cardiometabolic risk in the Nepalese population. Whether WC better predicts cardiometabolic risk than BMI in Nepalese individuals remains unclear. We conducted a comparative analysis of WC and BMI to evaluate their predictive ability for clinically relevant cardiometabolic outcomes.
Methods:
In this cross-sectional study of 678 adults, BMI and waist circumference (WC) were evaluated as anthropometric indicators of cardiometabolic risk in the Nepalese population. Their associations were examined across six clinically relevant outcomes: hypertension, diabetes mellitus, thyroid disease, fatty liver disease (FLD), non-alcoholic fatty liver disease (NAFLD), and dyslipidemia. Comparative performance of BMI and WC was assessed using correlation analysis with biochemical markers, multivariable logistic regression adjusting for key covariates, and receiver operating characteristic (ROC) curve analysis to evaluate their relative predictive and discriminatory ability in this setting.
Results:
BMI and waist circumference (WC) were strongly correlated (ρ = 0.75) but demonstrated distinct patterns in cardiometabolic associations. WC showed stronger correlation with triglycerides (ρ = 0.35 vs 0.24 for BMI) and was the only measure significantly associated with hypertension. Both BMI and WC were significantly associated with fatty liver disease (FLD), non-alcoholic fatty liver disease (NAFLD), and dyslipidemia; however, BMI generally showed higher odds ratios per unit increase compared to WC. Predictive performance was comparable between BMI and WC, with similar AUC values in the fair-to-acceptable range (0.6-0.7) across most outcomes. Notably, discrimination improved substantially (>0.8) when BMI and WC were combined with age and sex, yielding good predictive accuracy.
Conclusion:
BMI and WC provide complementary information for cardiometabolic risk in the Nepalese population. Combined assessment with age and sex improves risk stratification, indicating that integrating both measures with calibrated thresholds enhances clinical prediction compared with using either alone.
