Waist-to-height ratio as a measure of abdominal obesity in southern Chinese and European children and adolescents

BACKGROUND: Waist-to-height ratio (WHtR), with a 0.5 threshold (WHtR(0.5)), is regarded as a simple age-and gender-independent criterion of abdominal obesity (AO) and a better predictor than the 90th percentile of waist circumference (WCP90). OBJECTIVE: An analysis of gender and ethnic differences of WHtR and other AO indices between children and adolescents from southern China (HK: Hong-Kong, China) and Europe (LD: Lodz, Poland). SUBJECTS: Two large cross-sectional surveys of children and adolescents aged 7-19 years, one from LD (13 172) and one from HK (14 566). METHODS: The percentile and standardized values of WHtR and other parameters (WC, body mass index (BMI)) were assessed using the LMS method. The WHtR values corresponding to WCP90 and to the BMI definition of global obesity (BMIP95) were evaluated with the polynomial regression model. The compliance of the AO prevalence data, obtained with two criteria (WCP90 vs WHtR(0.5)) was analyzed using Cohen's kappa index (kappa C). RESULTS: The WHtR data of Polish subjects were generally higher than those of their HK peers, and the ethic differences increased with age. The WHtR values of HK boys showed a stronger relationship with BMI z-score. WHtR corresponding to WCP90 assumed values < 0.5. An application of Cohen's kappa coefficient (kappa C) to Polish subjects showed either 'substantial' (kappa C > 0.6) or 'almost perfect' (kappa C > 0.8) agreement in the AO prevalence for both criteria (WCP90 and WHtR(0.5)). For these criteria, either 'fair' (kappa C < 0.4) or 'moderate' (kappa C < 0.6) AO consistency ratings were observed among HK girls. In HK boys, a significant difference in the prevalence of AO was observed, independent of the criterion used. CONCLUSIONS: Our results provide further evidence of the need for developing ethnic-specific WC charts and for recommending that a WHtR cutoff of 0.5 may not be appropriate to predict cardiometabolic risk in children of different ethnic groups.