Body mass index (BMI), waist circumference (WC), waist-to-stature ratio (WSR), and sum of skinfolds (SoS) are widely used proxy measures of adiposity with variable performance across populations. Given the relationship between body fat and cardiometabolic disease, BMI, WC, and WSR are often used in clinical and public health settings to screen for risk. Because body composition, body size, and fat distribution vary between populations, these measures are strengthened using population-specific morbidity risk to determine thresholds for obesity. Unfortunately, few references are available for Indigenous populations, particularly those undergoing rapid market integration and experiencing increased burdens of cardiovascular and metabolic disease. Here, we compare the accuracy of BMI, WC, WSR, and SoS in predicting metabolic health in Ecuadorian Shuar using biomarkers collected as part of the community-centered, long-term Shuar Health and Life History Project. We compare the results of the proxy measures and examine establishing population-specific cutoffs for clinical categories using metabolic syndrome to improve the accuracy of screening tools. Results show BMI has the highest correlation with metabolic markers in the population, and matching metabolic syndrome prevalence to U.S. references raises the overweight and obesity thresholds to 27kg/m2 and 34 kg/m2. These results reduce the prevalence of overweight and obesity in the Shuar and are distinct from other population-specific BMI thresholds. These findings are important for understanding how metabolic health and excess adiposity interact through population-level research, and for interpreting health information for participants and healthcare patients.