Effectiveness and cost-effectiveness of early childhood obesity prevention

Abstract

Background: Due to individual health-related and social consequences, but also due to its immense cost burden, obesity remains a large public health challenge for which effective and cost-effective prevention is lacking. It has been suggested that primary preventive efforts operate optimally if started in early childhood, but scientific evidence is scarce. Optimal target groups and intervention programs are most likely to vary by structure of society, population characteristics, health care resources and many other factors.

Aims: The primary aim of this thesis is to evaluative the effectiveness and cost-effectiveness of a primary prevention programme addressing childhood obesity at Swedish child health care centres (CHCs). A secondary aim was to increase understanding of the association between health-related quality of life (HRQoL) and body mass index (BMI), which will be useful for future health economic evaluations.

Methods: Study I provides a description of the PRIMROSE intervention study, which includes a relative validation of a semi-quantitative food frequency questionnaire. Study II is based on repeated self-reported BMI and HRQoL measurements from the Stockholm Public Health Cohort. Study III is a main effect evaluation of the PRIMROSE cluster RCT, which includes 1148 children at follow-up. Study IV is a systematic narrative literature review that explores the methods and applications of economic evaluations within the field of childhood obesity. Study V is a trial-based economic evaluation of the PRIMROSE intervention from a societal perspective.

Results: Compared to normal weight, overweight and obesity was found to be associated with lower HRQoL, especially in the domains of mobility, self-care and usual activity. Heavy weight gain over eight years was also associated with lower HRQoL, whereas weight loss had no protective effect. The PRIMROSE intervention had no significant effect on BMI, waist circumference, obesity prevalence and objectively measured physical activity. Small intervention effects were found in dietary habits but should be interpreted with caution given the possibility of bias in assessments. Study IV showed that there were only a few health economic evaluations of childhood obesity, but these presented already small intervention effect to be worth the money. However, cost-effectiveness was largely dependent on model assumptions and, in particular, on decision-maker’s willingness to pay. The economic evaluation presented in Study V showed that PRIMROSE cannot be considered cost-effective given uncertainty around the effect measure.

Conclusion: There was no evidence for the effectiveness or cost-effectiveness of the population-based PRIMROSE RCT. Since the prevalence of childhood obesity remains high, further research is needed to disentangle the “failure” of intervention programs per se from “failure” of implementation and “failure” of evaluation.