The primary objective of this research is to integrate information on weight histories into estimates of prevalence, trends and mortality risks of obesity in the United States. Although histories are universally used in studies of smoking to differentiate between non-smokers who never smoked and those who smoked in the past and quit, the practice is uncommon in studies on obesity. In the obesity literature, most studies use a snapshot of weight at the time of survey for defining obesity status. As a consequence, there is generally no distinction made between non-obese individuals who were never obese versus non-obese individuals who were obese in the past and lost weight. However, this distinction would be important if individuals who were formerly obese are at higher risk than never obese individuals. Evidence from the prior literature suggests this may be the case for at least two reasons: first, the effects of past obesity may persist into the present even after weight is lost; second, some people who have experienced weight loss have done so as a result of a health condition, a source of bias commonly referred to as reverse causality. In this research, we will develop and validate a summary weight history measure in the National Health and Nutrition Examination Survey (NHANES). The focus of these efforts will be a question on lifetime maximum weight (max weight), which was first fielded in NHANES III (1988-1994) and has been included in every wave since (1999-2014). Second, we will use this measure to estimate prevalence and trends in the proportion of the population ever overweight or obese. These estimates will provide a more comprehensive understanding of the burden of obesity than estimates based on current obesity levels alone. Third, we will use max weight to obtain a new set of estimates of the mortality risks of obesity that both address reverse causality and capture the cumulative effects of past obesity status. Data will be drawn from the NHANES III, NHANES continuous waves (1999-2010) and NHANES linked mortality files (1988-2011) and Cox models will be used to estimate hazard ratios adjusted for multiple sources of confounding. This research is expected to generate new insight into the burden of obesity on the health of the US population as well as improved methods for measuring this burden. The project is also expected to enhance the public health utility of the NHANES weight history data and facilitate the use of lifetime measures of obesity in future research.
Weight histories are rarely used in national health surveillance, yet are vitally important to obtaining accurate estimates of obesity prevalence and health outcomes. This research will develop and validate a summary measure of weight history in the NHANES and use it to generate improved estimates of levels, trends, and mortality risks of obesity in the United States that address biases in prior estimates and capture cumulative effects of past weight. This research is expected to enhance the public health utility of the NHANES weight history data and facilitate the more widespread use of lifetime measures of obesity in future research.
