Population-wide changes in obesity : their causes, trends and implications

The prevalence of obesity is increasing worldwide. In Australia, for example, the prevalence of obesity doubled in two decades between 1980 and 2000, and continues to increase. Considering the substantial body of literature suggesting that obesity is the cause of much morbidity, premature mortality and an important component of increasing health care costs, it has become an issue high on the public agenda. But much remains unclear about what is commonly referred to as the ‘obesity epidemic’. This thesis examines trends in the whole population distribution of body weight and differences in trends based on two measures of excess body weight, using data from national population-based surveys from Australia and the United States, respectively. It also investigates the prevalence of obesity in Vietnam, a country with a historically low prevalence of obesity, using data from a population-based survey of northern Vietnam. Furthermore, this thesis examines individual-level predictors of weight gain, projects the likely future burden of obesity in Australia based on current incidence rates, and examines the impact of hypothetical obesity-reduction strategies, using data from Australian national population-based surveys, the National Death Index (NDI), and national mortality statistics. The findings suggest that in high-income countries such as Australia and the United States, the increase in obesity is the result of an increase in body weight across the whole population distribution, with much greater increases at the higher end of the distribution. There also appears to be a greater increase in abdominal weight than overall body mass index (BMI) across the whole population distribution of body weight, suggesting that in each of the BMI categories the level of metabolic risk has increased. In Australia, a third of the population underwent substantial abdominal weight gain between 2000 and 2005. The main individual-level behavioural predictor of such a gain was poor diet quality; however a commentary article included in the thesis discusses how the focus on individual-level determinants should not overlook the context in which such behaviours are made. Projecting recent measures of the incidence of overweight and obesity in Australia over the next two decades suggests that by 2025 approximately 34% of the population will be obese (compared to 21% in 2000). Of the normal-weight 25-29 year-olds in the year 2000, an estimated 36% will be overweight and 29% obese by age 60-64 years. Analysis of the effect of a series of hypothetical interventions showed that population strategies that de-skew the distribution of body weight would be the most safe and effective. In urban Hanoi, Vietnam, almost a third of adults were found to be overweight or obese in 2004, and one in ten were underweight, based on Asian-specific BMI cut-points. It has long been suggested that obesity places a heavy health burden on individuals and an encumbrance on the health care system. By analysing trends in the whole population distribution of body weight, their causes, and the likely trajectory of the future burden of obesity, this thesis contributes to a better understanding of the increasing prevalence of obesity.