Background: Excess body fat leads to obesity-related morbidity and population/ethnicity-specific cut-off values of anthropometric measures are useful for better diagnosis. This study assesses the suitability of newly-developed Sri Lankan anthropometric cut-off values in the diagnosis of obesity in Sri Lankan children.
Methods: A cross-sectional study was conducted at University of Colombo, Sri Lanka involving 5-15 year old children. Height, weight, waist (WC), and hip (HC) circumferences were measured. Total body fat (FM) was measured using whole body BIA. WHR and WHtR were calculated. Validity of anthropometric measures in detecting childhood obesity (Sri Lankan BMI/WC; IOTF, WHO, British and CDC BMI and British WC cut-off values) were evaluated.
Results: Nine hundred and twenty children were assessed. FM showed significant associations with BMI (r = 0.92, p < 0.001), WC (r = 0.90, p < 0.001) and HC (r = 0.85, p < 0.001), but poor association with WHR (r = 0.17, p < 0.001). However, WHtR had a high association with FM (r = 0.75, p < 0.001) and %FM (r = 0.78, p < 0.001). Based on %FM cut-offs, 85 (22.8%) girls and 101 (18.5%) boys were obese. All international anthropometric cut-off values under-estimated obesity. Sri Lankan WC and BMI cut-off values over-estimated obesity. International BMI based cut-off values had high specificity (>99%) but a low sensitivity (∼12-33%), while Sri Lankan BMI cut-off values had high sensitivity (>93.1) but low specificity (>79.7).
Conclusions: Internationally available BMI cut-off values are poor in diagnosing obesity in Sri Lankan children. Newly developed Sri Lankan BMI cut-off values for children improved the diagnosis. WC can be used successfully as an alternative diagnostic tool of obesity.