Psychological predictors and outcomes of childhood overweight and obesity

Abstract

Obesity has clear and serious consequences for physical health, many of which emerge in adulthood. For obese children, poor psychological health may be the most obvious and immediate implication of their excess adiposity but the evidence to support this is mixed, particularly for community-based (i.e. non-clinical) groups. Given that associations between psychological health and body mass index (BMI) are likely to be bidirectional, longitudinal methodologies seem best suited to clarify the nature of these relationships. Furthermore, research conducted with community-based samples may be generalizable to the wider population of overweight and obese children. However, most research to date has been cross-sectional and conducted with clinical populations of obese children. In light of these evidence gaps, the first objective of this thesis was to determine the longitudinal relationship between psychological well-being and subsequent excess adiposity gain and obesity onset across childhood and adolescence. This was addressed in a systematic review (Paper 1) and a research paper (Paper 2). The systematic review, which selected population-based studies for maximum generalizability, suggested that poorer psychological well-being may increase the incidence of obesity and contribute to obesity persistence across adolescence. However, major limitations and inconsistencies were identified in the literature. Paper 2 of this thesis addressed a gap identified in the systematic literature review- the need for further good-quality research examining individual psychological predictors of adiposity change in the overweight or obese subgroup. It employed a longitudinal cohort of 5-9 year old children who were all initially overweight or mildly obese when they presented to primary care 4 years earlier. Parallel parent proxy- and child self- reported psychosocial measures were assessed as predictors of changes in body mass index (BMI; kg/m2) standard deviation scores (i.e. BMI z-scores). Results revealed little evidence that initial psychosocial functioning impacted on subsequent BMI z-score change. However, changes in several domains of psychosocial well-being, especially in relation to body-image and appearance, were associated with concomitant BMI z-score change. A strong degree of corroboration between parents and children strengthened the validity of the findings. The second objective of the thesis was to investigate whether and how BMI was associated with psychological outcomes among non-clinical overweight or obese children. Consistent with Paper 2, Paper 3 achieved this within a cohort of 5-9 year olds recruited from primary care for their overweight or mild obesity. Results confirmed that overall, psychological well-being at 4-year follow-up was weakly predicted by concurrent BMI and when impairments were found, they were most likely for peer and eating-behaviour domains. This study extended on existing knowledge by demonstrating that changes in BMI categories also contributed little variance to the psychological outcomes of initially overweight children. The implications of this research focus on informing effective prevention strategies to reverse the current trends in youth obesity. Recommendations include targeting poor psychological well-being prior to adolescence to yield the most benefit for preventing the onset of obesity. Intervention strategies for the quarter of children in the community who are overweight or obese could incorporate modules that target peer relationship and eating problems, the most relevant psychological comorbidities of excess adiposity. Enhancing well-being in these domains would be beneficial for immediate quality of life, future mental health and potentially initiate flow-on effects that improve physical health.