Exploring the nonlinear relationship between body mass index and health-related quality of life among adults: a cross-sectional study in Shaanxi Province, China

Yongjian Xu, Zhongliang Zhou, Yanli Li, Jinjuan Yang, Xiaoyuan Guo, Jianmin Gao, Ju’e Yan, Gang Chen

China is a country facing the "double burden" of both obesity and underweight. The objective of this study was to explore the relationship between body mass index (BMI) and health-related quality of life (HRQOL) in adults from Shaanxi Province. The data were derived from the fifth Health Service Survey of Shaanxi Province, which was part of China's National Health Service Survey (NHSS), conducted in 2013. The HRQOL was assessed using the three-level EQ-5D questionnaire and scored based on a recently developed Chinese-specific tariff. Semiparametric regression models were adopted to explore the non-linear relationship between continuous BMI and overall HRQOL scores. Logistic regression models were further undertaken to assess the relationship between categorized BMI and five dimensions of HRQOL. Among the study sample (n = 37,902), 77 % of men and 75 % of women were assigned to normal weight, according to the WHO International classification. There were statistical significant nonlinear relationships between BMI and HRQOL, with optimal HRQOL achieved at a BMI of near 23 kg/m(2) for men and 24 kg/m(2) for women. Before BMI reached optimal HRQOL, the EQ-5D utility scores were increasing faster among men than the women, whilst after the BMI value reached the optimal utility scores, women showed a faster decline in utility scores than men. With adjustments for socio-demographic, physical activity and co-morbidities, obese respondents were more likely to suffer from physical rather than mental problems. Underweight respondents were significantly more likely to report having any problems in all five dimensions of the EQ-5D, whilst the magnitudes of odds ratios were consistently larger for men than women. There was an inverse U-shaped association between continuous BMI and overall HRQOL scores, meaning that both underweight and obesity were associated with lower HRQOL. The relationship between BMI and HRQOL varied between sexes. Underweight respondents had a higher risk of suffering from both physical and mental problems. Interventions aimed to tackle the prevalence of underweight should be put into action in Shaanxi Province.