Quality-of-life: a many-splendored thing? Belgian population norms and 34 potential determinants explored by beta regression

Abstract

To identify determinants of health-related quality-of-life in the Belgian population and to provide age-specific population norms of health-related quality-of-life. Between September 2010 and February 2011, a representative sample of 1774 persons (age 0-99) was surveyed using the standard Euroqol questionnaire (EQ-5D-3L) with a Visual Analogue Scale (VAS). Significant determinants were identified using multivariate beta (VAS) and one-inflated beta (EQ-5D) regression, the latter modelling the probability to be in perfect health separately from the average EQ-5D score if not in perfect health. Health-related quality-of-life depends largely on age and experience with severe disease. The probability to be in perfect health is highest for children. For 0-2 years children who are not in perfect health, proxies report EQ-5D and VAS scores as low as that of the elderly. Also smoking behaviour, educational attainment, pet ownership, working or having worked in health care, and potentially household size and 60+ living on their own (yes/no) are associated with health-related quality-of-life, whereas no association was found with gender, living in a single-parent home, educational attainment of mothers, alcohol consumption of 60+, having (grand-) children and the frequency of seeing them. The same determinants are significant for VAS and the probability to be in perfect health, but not for the average EQ-5D score if not in perfect health. The population norms provided can be used directly as input in health economic evaluations. Estimating health-related quality-of-life in children and developing statistical tools capturing the particular features of health-related quality-of-life measures are important areas for future research.