Understanding the Impact of Diabetes and Obesity on Direct Medical Expenditures of American Adults in 2011

Diabetes is among the top ten leading causes of death and disability in the United States. An important, known risk factor that can lead to diabetes is obesity, but understanding how obesity and diabetes interact to impact health expenditures is not fully understood. In 2011, there were 20.6 million non-institutionalized adults (age greater than 18) living with diabetes in the US. For this same year, we see that over 27% of adults in the US are obese.

This study attempts to quantify the amount of health expenditures adult Americans with diabetes spent in 2011 relative to adults without diabetes and how these expenses vary across person characteristics such as obesity. We extract health expenditure and personal characteristic data from the Medical Expenditures Panel Survey (MEPS) 2011 dataset.  MEPS is an annual survey conducted by the Agency for Healthcare Research and Quality. It collects self-reported patient- and family-level characteristics, medical expenditures, and causes of illness which are then cross-validated by the survey team from health facilities and insurance companies.

Using three different statistical models (Heckman selection two-step Tobit, Zero-inflated Negative Binomial, and Negative Binomial), this study determines the most appropriate model through goodness-of-fit tests. Using the Heckman selection model, we estimate the impact of diabetes on an individual’s health expenditures while holding all measured potential confounders at their means, such as income, smoking status, and number of comorbidities.

We found that in 2011, adult Americans with diabetes spent on average $1,222 more on health than adults without diabetes after controlling for potential confounders. Adults who are obese and diabetic are estimated to have incurred $1,000 more on healthcare than  adults who are diabetic but not obese, while adults who are only obese and not diabetic incurred $270 more on healthcare expenses than someone who is neither obese or diabetic. This finding indicates that diabetes and obesity interact in a way that incurs more health expenditures than if only one of these issues are present. For the first time since 1980, the Centers for Diseases Control estimates fewer adults living with diabetes in the United States. However, the prevalence of obesity continues to increase. Understanding how diabetes and obesity interact and contribute to health expenditures is essential for the healthcare system and its managers to best address these health issues while working to control and ultimately reduce health expenditures.