OBJECTIVES:

To map the Impact of Weight on Quality of Life-Lite (IWQOL-Lite) scores onto the EQ-5D-5L and SF-6Dv2 utility values in Chinese overweight and obese population.

METHODS:

The date were collected using online survey among a representative sample of the population stratified by BMI, age, gender, regions in China. The respondents were randomly divided into estimation (80%) and validation (20%) groups to develop and validate the mapping algorithms. The conceptual overlap between instruments was evaluated by Spearman's correlation coefficients (ρ). Three kinds of independent variables, the total scores, dimension scores, and item scores of IWQOL-Lite, were included to estimate the health state utilities of the EQ-5D-5L and SF-6Dv2, respectively. In each kind of independent variable, five models were conducted, including ordinary least squares (OLS), Tobit, censored least absolute deviations (CLAD), generalized linear model (GLM) and two-part model (PTM). The model performance was assessed using mean absolute error (MAE), root mean squared error (RMSE), Akaike information criterion (AIC) and Bayesian information criteria (BIC), where applicable.

RESULTS:

The mean IWQOL-Lite, EQ-5D-5L and SF-6Dv2 scores among the total sample (N = 1,000) were 78.5, 0.880, and 0.789, respectively. A high correlation between the EQ-5D-5L and SF-6Dv2 and the IWQOL-Lite was observed (ρ = 0.702 and 0.755). The CLAD model with total scores performed the best in both mapping algorithms for the EQ-5D-5L and SF-6Dv2 (MAE = 0.070 vs. 0.072-0.083 and 0.077 vs. 0.077-0.078; RMSE = 0.005 vs. 0.005-0.006 and 0.005 vs. 0.005).

CONCLUSIONS:

The CLAD model with total score were selected to map the IWQOL-Lite scores onto the EQ-5D-5L and SF-6Dv2 utility values among overweight and obese population in China.