Obesity and Depression: Establishing Causality

Posted: 19 Jun 2007

See all articles by Greg Colman

Greg Colman

Pace University - Department of Economics

Inas Kelly

Loyola Marymount University; National Bureau of Economic Research

Abstract

In this paper we present an instrumental variables estimate of the effect of obesity on depression in females. Previous research has established a significant correlation between the two. The direction of causality, however, is unclear. We use employment data from the ES202 program and prices from the American Chamber of Commerce Researchers Association (ACCRA) as a vector of instruments for obesity to examine the independent effect of obesity on depression. In conducting this analysis, we employ micro-level data from the third National Health and Nutrition Examination Survey (NHANES III), covering years 1988 to 1994, as well as data from the 1979 cohort of the Child-Young Adult National Longitudinal Survey of Youth (NLSY79). We find marginal evidence that being obese actually causes levels of depression to increase. We conclude that unobservable characteristics influencing both obesity and depression are the likely causes of the strong correlation, and that factors such as perception of one's weight possibly play a greater role in determining levels of depression.

Keywords: obesity, depression

JEL Classification: I10, I12

Suggested Citation

Colman, Greg and Kelly, Inas, Obesity and Depression: Establishing Causality. iHEA 2007 6th World Congress: Explorations in Health Economics Paper, Available at SSRN: https://ssrn.com/abstract=995010

Greg Colman (Contact Author)

Pace University - Department of Economics ( email )

One Pace Plaza
New York, NY 10038
United States

Inas Kelly

Loyola Marymount University ( email )

7900 Loyola Boulevard
Los Angeles, CA 90045
United States

National Bureau of Economic Research ( email )

1050 Massachusetts Avenue
Cambridge, MA 02138
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
1,007
PlumX Metrics