Obesity, place and environment: the spatial distribution and correlates of weight status in South Australian preschool children.

The issue of overweight and obesity in childhood has received a great deal of recent attention in both the academic literature and popular media. These discussions have tended to concentrate on individual responses to behavioural and nutritional choices, with limited exploration of how the wider social and economic environment might influence weight outcomes. However there is a growing body of research which has identified area level effects on health outcomes, and this suggests that location should be an important consideration in obesity research. Currently, very little formal investigation of weight status has been conducted among children of preschool age and location is not routinely considered in obesity research, especially at the small area level and particularly with reference to children. Given that childhood overweight is known to persist into adulthood and that behavioural change may be easier to effect in preschoolers, it is appropriate to focus research attention on this age group. This study explores an administrative data set containing over 120 000 individual records collected over ten years and supplied by the South Australian Children, Youth and Women‟s Health Service. Geographical Information Systems (GIS) are used to determine the prevalence, distribution and area-level correlates of obesity in South Australian four year old children between 1995 and 2003. It aims to determine if there has been significant variation in the spatial distribution of obesity prevalence between different communities over this time period, and to detect relationships between weight status, socio-economic variables and environmental attributes at a small scale which may be able to explain some of the discrepancy. These are investigated in conjunction with the data items available for the individual children in this data set. A univariate analysis approach using cross-tabulation and chi square testing has been used to explore the relationships between the obesity prevalence of the study population and selected socio-demographic and environmental variables at a small area level. The Australian Census of Population and Housing is the primary source of socio-demographic data, but other variables including housing characteristics, proximity to fast food outlets, proximity to recreational areas and the walkability of neighbourhoods have also been examined. Analysis of this data set reveals an increase in obesity prevalence over time, in line with national and international trends. For individual children, birth weight, ethnicity and breastfeeding history appear to be particularly influential in the development of overweight at four years of age, but there is nevertheless a distinct spatial patterning of obesity prevalence throughout the state, and also within the metropolitan Adelaide area. While there is generally a positive association between socio-economic status and obesity, these relationships are not necessarily straightforward and the area-level physical and social environmental variables actually show a varying relationship with obesity prevalence in different communities. This study has clearly identified neighbourhood characteristics as an important component in the complex etiology of obesity development in even very young children. It has shown that aspects of environment such as ethnicity and disadvantage should be taken into account when targeting and tailoring public health initiatives to combat the development of obesity in these populations. The exploration of this unique, administrative data set with reference to location has illustrated the complexity of the relationship between biology and environment in the development of overweight and obesity in young children. This has implications for policy development across many spheres of government.Thesis (Ph.D.) -- University of Adelaide, School of Social Sciences, 201