Background: Although obesity is still associated with excess mortality, recent studies have found that persons who are obese are healthier than in the past. However, it is unclear whether the health improvements are because of primary or secondary/tertiary prevention. This is important because secondary or tertiary prevention allows persons to live longer with disease. Although persons who are obese are living longer, they may also have a higher burden of disabling conditions. Alley and colleagues examined the association between obesity and disability for older Americans over two periods: 1988 to 1994 and 1999 to 2004.

The Study: The authors used data from National Health and Nutrition Examination Surveys (NHANES) that correspond with the two periods to examine the relationship between body mass index (BMI) and disability in persons 60 years or older. Disabilities were defined as functional disabilities involving restrictions of movement, or interference with activities of daily living (ADL). After exclusions for missing data, the final sample was comprised of 9,928 participants. The authors accounted for demographic covariates, as well as comorbid conditions. Obesity was divided into three weight categories: class I (BMI 30.0 to 34.9 kg per m²), class II (BMI 35.0 to 39.9 kg per m²), and class III (BMI 40.0 kg per m² or greater).

Results: From the first period to the second, obesity prevalence increased from 23.5 percent of the population to 31.7 percent. Functional impairment in these persons increased from 36.8 to 42.2 percent. The odds ratio (OR) of having a functional impairment increased 43 percent. There was no increase in functional impairment for nonobese persons. Compared with nonobese persons, persons who are obese had increased odds of having functional impairment in the first period (OR =1.78; 95% confidence interval [CI], 1.47 to 2.16) and the second period (OR = 2.75; 95% CI, 2.39 to 3.17). The association of obesity and functional impairment increased with BMI class, and after adjusting for chronic conditions, smoking status, and health insurance status, persisted in all except class I obesity.

ADL impairment decreased between the first and second periods in nonobese persons, but remained unchanged in persons who are obese. Compared with nonobese persons, the OR for ADL impairment in persons who are obese increased by a factor of 1.56 (95% CI, 1.03 to 2.36) from the first to the second period. Adjusting for BMI class and confounders attenuated these associations, but the relative risk of ADL impairment persisted in all but class I obesity.

Conclusion: The odds of functional impairment in persons who are obese relative to nonobese persons increased between NHANES 1988–1994 and NHANES 1999–2004. ADL impairment declined in nonobese persons but remained unchanged in persons who are obese, a change partially explained by the increased prevalence of severe obesity. Persons who are obese were more likely to have disability in the second period than in the first. The authors conclude that these findings suggest that the burden of disability will likely continue to increase as the population ages.

editor's note:In an accompanying editorial, Gregg and Guralnik emphasize that advances in managing cardiovascular risk factors are allowing persons to live longer, but less active lives, and that a public health approach to preventing obesity is likely to yield better results than treating its complications.¹ In the same issue of the Journal of the American Medical Association, Flegal and colleagues report on excess cause-specific mortality and BMI. They confirm in more detail a previous finding that persons who are overweight have a decreased all-cause mortality compared with persons who are normal weight. However, combined kidney disease and diabetes was associated with increased mortality in persons who are overweight and obese. Obesity was associated with excess cardiovascular deaths and specific, obesity-related cancers.²

The findings in this study and the study by Alley and colleagues suggest that the most aggressive weight loss counseling should be reserved for persons who are obese and especially those who are severely obese. Counseling on diet composition, rather than weight, may be preferable because a diet high in whole grains and monounsaturated fats is positively associated with cardiovascular outcomes regardless of whether weight loss occurs.³—c.w.