Why do people want to lose weight? While this seems like an obvious question, it does merit answering. There are two major reasons: health concerns and appearance. Being obese increases the risk of suffering from many diseases (e.g.: diabetes). On the appearance side, individuals may experience social pressure to lose (or possibly gain) weight. Further, individuals may want to maintain a healthy body appearance to attract a mate.
One of the more interesting questions is how an individual’s marriage status affects obesity. It is generally found that–even controlling for age and other covariates–married individuals are more likely to be overweight than non-married individuals. Why is this the case. Sobal cites some studies which attempt to explain this.
- Entering into marriage results in higher food intake. Craig & Truswell (1988).
- Marriage may change the physical activity level of the individuals. Some studies find that married individuals are more active (King, Kiernan, Ahn, & Wilcox 1998) other that they are less active (Myers, Weigel, & Holliday 1989, Verhoef, Love & Rose 1992), and still others find no relationship between marriage and physical activity (Hirvensalo, Lintunen & Rantanen 2000, Umberson 1992, Van den Hombergh, Schouten, Van Staveren, van Amelsvoort, & Kok 1995.).
- Smoking has been found to be correlated with lower body weights. Two studies (Venters et al. 1989, Waldron & Lye 1989) have found that married individuals are less likely to smoke than single, divorced, or widowed individuals. Two other sutides found that people who enter marriage are more likely to stop smoking (Waldron & Lye 1989, Wee, Rigotti, Davis & Phillips 2001).
- Umberson 1992 found that after 3 years, Americans who exited marriage had lower weights than those who remained married. In a six year study, Finnish individuals who got married gained weight and those who were widowed lost weight (Rissanen, Heliovaara, Knekt, Reunanen & Aromaa, 1991). Three studies by Kahn and Williamson use a 10 year panel of U.S. residents and find individuals gained weight upon entering marriage and lose weight upon exiting marriage (Kahn & Williamson 1990, Kahn & Williamson 1991, and Kahn, Williamson and Stevens 1991). These findings hold for both men and women in these two studies.
- While most studies do find that married individuals weight more than nonmarried, there are some subtleties to this statement. A paper by French et al. 1993 using 2-year data on US men and women found that women gained weight after they married, but men did not.
After citing all this evidence, Sobal and co-authors state 4 hyptotheses to test:
- Marital trajectories that are stable are related to stable body weights,
- marital trajectories entering marriage are related to weight gain,
- marital trajectories dissolving marriage are related to weight loss,
- marital trajectories involving the death of a spouse are related to weight loss.
Sobal uses data from from the National Health and Nutritional Examination Survey (NHANES I). A 10 year follow up survey of the participants is collected in the National Health and Nutrition Epidemiological Followup Survey (NHEFS). The authors use an OLS specification with a lagged dependent variable (i.e., lagged BMI) in order to estimate the impact of marital status on weight. Sobal, Rauschenbach and Frongillo conclude the following:
- Stable marital trajectories were not associated with significant weight changes, except for weight loss among men who remained separated/divorced.
- Marital trajectories involving entry into marriage were associated with weight gain among women, but not among men.
- Marital trajectories involving dissolving marriages were associated with weight loss among men, but not women.
- Marital trajectories involving death of a spouse were associated with weight loss among men, but not women.
- Marital and other demographic characteristics were better predictors of weight loss than weight gain.
- Sobal J, Rauschenbach B, Frongillo EA., 2003. “Marital status changes and body weight changes: a US longitudinal analysis.” Social Science and Medicine, 56(7): 1543-55.