The time trends and the determinants of adult obesity measured by a fatness-based index

Background and Objective: A rapid and sustained rise in the obesity rate during the past three decades has stimulated a considerable amount of literature, using the obesity index defined by a body mass index (BMI). However, the BMI has been criticized because it fails to distinguish body fat from lean body mass. The objectives of this study are to (1) measure the prevalence of adult obesity defined by the fatness-based percent body fat (PBF, the ratio of body fat to total weight multiplied by 100), and to (2) explore policy options to reduce the PBF-based obesity.

Methods: This study developed the PBF algorithm equation using the variables (such as weight, height and age by gender and race) from the dataset: the Third National Health and Nutrition Examination Survey (NHANES III, 1988-1994). Then, the PBF value was predicted for each adult subject enrolled in the Behavioral Risk Factor Surveillance System (BRFSS), ultimately estimating the time trends of obesity prevalence measured by PBF from 1984 to 2009.

The pooled cross-sectional Ordinary Least Squares (OLS) analysis was conducted, using BRFSS 1984~2009 data (N= 2,520,693) for 18~65 aged White and African Americans. The outcome variables were BMI, PBF, BMI-defined obesity and PBF-defined obesity. Covariates include the number of restaurants per 10,000 capita, the prices of a meal in fast-food and full-service restaurants, the price of food consumed at home, the price of cigarettes, gender, race, marital status, age, education level, income, region and the clean indoor air law by state.   

Results: The mean PBF is 31.33, being greater than 27.37 of the mean BMI. The average rate of PBF-defined obesity (68.8%) was 2.6 times higher than that of BMI-defined obesity (26.2%).

Out-of-home food prices were more sensitive to PBF or PBF-defined obesity than BMI or BMI-defined obesity. For instance, the PBF-defined obesity was 1.5 times more sensitive to the fast food price than the BMI-defined obesity. A dollar rise of fast-food price reduces 7% points of the PBF-defined obesity probability compared to 4.7% points reduction of the BMI-defined obesity probability (difference: 2.4% points, p<0.01).

In addition, the PBF-defined obesity was 4 times more sensitive to the price of full-service restaurants than the BMI-defined obesity. A dollar rise of the price of full-service restaurants reduces 1.3% of the PBF-defined obesity probability, compared to 0.3% point reduction of the BMI-defined obesity probability (difference: 1.0%, p<0.01). The PBF was 1.3 times more sensitive to the price of full-service restaurants than the BMI. A dollar rise of the price of full-service restaurants reduces 0.124 points of PBF, compared to 0.095 points reduction of BMI (difference: 0.029 points, p<0.01).

However, the number of restaurants, the price of home food and the cigarette price were less sensitive to PBF or PBF-defined obesity than BMI or BMI-defined obesity.

Conclusions: The adult obesity is more prevalent in the United States when measured by PBF, compared to BMI. Controlling the prices of fast food and full-service restaurants is more effective to manage the PBF-defined obesity than the BMI-defined obesity.