Course of Health-Related Quality of Life: Results of the BELLA Cohort Study

Health-related quality of life (HrQoL) is regarded as an important parameter for describing the subjective health of children and adolescents and therefore increasingly taken into account in health reporting systems. Population-based cross-sectional studies have shown that HrQoL in children and adolescents is correlated with mental and physical health as well as with sociodemographic characteristics. However, empirical studies on the longitudinal course of HrQoL in the general population are rare. The aim of the present study was to examine the course of self-reported HrQoL in children and adolescents in the general population and to predict this in relation to mental health problems, somatic problems, sociodemographic factors, as well as further predictors. The survey on mental well-being and behavior ("BEfragung zum seeLischen WohLbefinden und VerhAlten," BELLA) is a nationally representative and prospective cohort study of German children and adolescents between 7 and 17 years of age. Of four measurement points (t1 between 2003 and 2006, t2 and t3 1 year later each, t4 4 years later), the first three were used for the analyses presented here. At all three time points, the self-report of children and adolescents between 11 and 17 years was used (n = 1,597), and some predictors were obtained according to the parents' judgment. The data were investigated using a multilevel model for longitudinal data. Physical well-being decreased significantly with age; girls had lower overall values and a greater decrease with age than boys did. The course of psychological well-being corresponded with that of physical well-being, although with smaller effect sizes. The values for relationship with parents and autonomy and school environment decreased slightly over the years but did so regardless of gender. There were no age or gender effects for peers and social support. Significant predictors were mental health problems and physical problems of the child, mental health problems of the parents, as well as age, gender, migrational background, and socioeconomic status. A stepwise inclusion of predictors revealed that the most important predictor for the course of HrQoL was the mental health problems of the child, cross-sectional predictors had a rather low impact, and interactions hardly contributed to the variance explained. To understand the temporal dynamics of HRQOL properly, not only a variety of relevant indicators but also a developmental theory of HrQoL is required. The results provide guidance for assessing the patterns of individual patients and patient groups, and give indications of possible outcomes for improving HRQOL through preventive measures.