Abstract
BACKGROUND: To analyze the strength of association between self-rated health and six anthropometric and body composition measures to explore the best indicator. METHODS: Analyses were based on the cross-sectional data from the China Kadoorie Biobank Study and approximately 300,000 adults were analyzed. Logistics regression was used to analyze the association between self-rated health (good or poor) and anthropometric and body composition measures (height, weight, body mass index (BMI), waist circumference (WC), hip circumference (HC) and body fat percentage, waist-to-hip ratio and waist-to-height ratio). Stratified analyses were undertaken to understand the effect modification of socioeconomic status on the association. RESULT: Odds ratio of self-rated better health had an inverted U-shape association with weight, BMI, WC and body fat, with weight levels increasing until around 73.8 and 65.7 kg for male and female, BMI around 26.8 kg/m(2), WC around 85.8 and 87.6 cm, body fat around 24.3 and 36.3%, and then declining thereafter. Height and HC also indicated a slightly inverted U-shape association. The strongest association was observed after adjustment was weight, with one standard deviation greater weight associated with 10.2% and 10.6% increased odds in male and female. CONCLUSIONS: Being underweight and overweight are both risk factors for poor self-rated health in males and females, and weight is the best indicator of self-rated health compared with other measures.