Estimating Age- and Sex-Specific Utility Values from the CHU9D Associated with Child and Adolescent BMI z-Score
Killedar A., Lung T., Petrou S., Teixeira-Pinto A., Hayes A.
© 2019, Springer Nature Switzerland AG. Objective: Our objective was to identify age- and sex-specific utilities for children and adolescents by body mass index (BMI) z-score. Methods: We used data from 6822 participants and 12,094 observations from two cohorts and two waves of interviews from the Longitudinal Study of Australian Children. We fit linear models using generalised estimating equations to investigate associations between Child Health Utility 9D and BMI z-score in girls and boys aged 10–17 years. We initially fit models for each sex, fully adjusted for known predictors of health-related quality of life, including socioeconomic position, long-term medical condition and maternal smoking status and also included an interaction between age and BMI z-score to examine age-specific effects. Finally, we derived a minimal model for each sex by eliminating interaction terms with P > 0.01 and predictors with P > 0.05. Results: Our adjusted results show different utility patterns in girls and boys. In girls, utility decrements for each unit increase in BMI z-score changed with age (P < 0.01 for interaction between age and BMI z-score). At age 10 years, the mean utility decrement for each unit increase in BMI z-score was 0.002 (95% confidence interval [CI] 0.011 decrement to 0.006 increment), but, by age 17 years, this utility decrement was 0.023 (95% CI 0.013 to 0.032). In boys, small non-significant decrements were found in utility for each unit increase in BMI z-score, with no observable change with age. Conclusion: Our analyses demonstrated that age and sex should be considered when attributing utility values and decrements to BMI z-scores.