Exploring the social and spatial context of adult obesity in Aotearoa New Zealand : a spatial microsimulation approach.

Abstract

Obesity rates have risen substantially in recent decades. A large body of research links excess body fat to a variety of health conditions including cardio vascular disease (CVD), certain cancers, and non-insulin-dependent diabetes mellitus (NIDDM). Research also indicates that recent social and economic change are among the underlying causes of the ‘obesity epidemic.’ Aotearoa New Zealand exhibits one of the highest obesity rates in the Organisation for Economic Co-operation and Development (OECD). Reducing obesity in New Zealand is therefore a priority for policy makers.

Existing research demonstrates that the burden of obesity is not evenly distributed in Aotearoa New Zealand. Obesity rates are highest among Māori and Pacific Peoples, and those living in the most socially deprived areas, neither of which are evenly distributed spatially.

Sampling constraints mean that standard statistical methods are unable to estimate obesity rates at a spatial scale smaller than at the District Health Board (DHB) level. Yet fine-scale estimates of obesity would help to understand the distribution of obesity at neighbourhood level and thus provide policy makers with a tangible tool to target and combat obesity. Neighbourhoods in Aotearoa New Zealand vary across the regions of the country so to rely on large scale statistics for decision-making risks overlooking small pockets that would benefit from targeted assistance.

The aim of this thesis is to put population level adult obesity in New Zealand into a spatial context using spatial microsimulation modelling (SMSM). SMSM is a technique that combines detailed microdata from the New Zealand Health Survey (NZHS) with small area census data to generate obesity estimates at a neighbourhood level.

There are three key findings in this thesis. First, obesity is clustered into a spatially confined subset of areas, primarily associated with high deprivation mediated by age and ethnicity. Second, a broad range of obesity rates were estimated for small areas, varying from 15.3% to 67.2%; these estimates of obesity in 2013 are novel and not available through other sources.

Third, projections from the model for 2018 and 2023 predict only small changes in obesity rates, yet a widening of obesity related health inequities.

The SMSM outputs will be useful for operational policy decisions as well as informing policy more broadly. Collectively, the work presented here extends the understanding of the geography of obesity in Aotearoa New Zealand.