Changes in Obesity-Related Healthcare Spending 2006-2016 by Payer Type and Age Category

Monday, June 24, 2019: 10:00 AM
Taft - Mezzanine Level (Marriott Wardman Park Hotel)

Presenter: Eline van den Broek-Altenburg

Co-Author: Adam Atherly

Discussant: Adam I. Biener


Rising BMI has been suggested as a key reason for rising healthcare spending in the United States. Previous research has found that the obesity-related attributable fraction of healthcare spending has indeed been an important cause of healthcare spending. However, the role of obesity in increased healthcare spending in recent years is not established. Despite this lack of evidence, obesity has been repeatedly cited as a key reason for current increased in healthcare spending. This paper estimates the obesity attributable fraction of healthcare spending by payer and service type. It examines not only the role of obesity but also elevated BMI levels, often referred to as overweight, on spending. We analyze ten-year trends in obesity rates and obesity-induced spending and model the marginal effect of obesity prevalence and changes in the relationship between obesity and spending on healthcare expenditures.


Our data source is the Medical Expenditure Panel Survey (MEPS) Household Component, a nationally representative sample of families and individuals. We used the 2006, 2011 and 2016 Full Year Consolidated file for our analyses. To analyze the effect of obesity and overweight on healthcare spending, we looked at expenditures across service lines (total, inpatient, outpatient, ED and drugs), and stratified by payer (Medicare, Medicaid, private insurance, uninsured) and age. Expenditures were modelled using Generalized Linear Models (GLM) for total expenditures, pharmaceutical and ED spending and two-part models for inpatient spending. BMI between 25 and 30 was considered “overweight” and BMI was larger than 30 was considered “obese”. The models controlled for sociodemographic and health characteristics not in the causal pathway between obesity and healthcare spending. The attributable fraction is equal to the ratio of the change in spending with and without obesity divided by total spending, by category and overall, and represents the proportion of spending attributable to obesity and overweight, controlling for other variables in the model. Standard errors were calculated using a bootstrap method with 1,000 replications.


The total attributable fraction for obesity on total expenditures for insured population declined from 11.4 percent to 8.5 percent, although total obesity-related spending increased. This decline was found in privately insured (11.5% to 8.3%), and Medicaid populations (11.5% to 8.4%), but not Medicare (8.1% to 8.2%). Obesity-attributable spending peaked between ages 35 to 54. Outpatient obesity-attributable spending was relatively stable over ten years (a statistically insignificant change from 8.9% in 2006 to 9.0% in 2016), while inpatient obesity-attributable spending declined. Overweight was consistently insignificant in the expenditure models.


The preliminary results imply that obesity was not been an important driver in the increase of healthcare spending between 2006 and 2016. This was found across age, payer and service type. Models separating the effect of changes in obesity prevalence and the relationship of obesity and spending indicate that the latter is the reason for the moderation in effect. One possible explanation is that efforts to more effectively manage obese populations may be paying dividends.