Contents

Executive Summary

Section One

Section Two

Section Three

References

Appendix A

Appendix B

Appendix C

Credits

WV | DHHR | BPH | OCHSHP | HSC

Obesity:
Facts, Figures, Guidelines

Section One - Continued
OBESITY AND MORTALITY

According to the National Institutes of Health, obesity and overweight together are the second leading cause of preventable death in the United States, close behind tobacco use (3). An estimated 300,000 deaths per year are due to the obesity epidemic (57).

The results of two extensive studies examining obesity-attributable deaths in the United States were published in 1999. Allison, Fontaine, and Manson et al., reporting in the Journal of the American Medical Society, used data from a number of prospective cohort studies, including the Alameda Community Health Study, the Framingham Heart Study, the Tecumseh Community Health Study, the American Cancer Society’s Cancer Prevention Study I, the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study, and the Nurses’ Health Study, to estimate the number of deaths attributable to obesity in the United States on an annual basis (66). Their initial analyses, which examined deaths occurring among persons aged 18 and older in 1991, were adjusted only for age, sex, and smoking status. The weight categories used were overweight (BMI of 25-29.9), obese (BMI of 30-35), and severely obese (BMI >35).

Using data on all eligible subjects from all six studies, Allison et al. estimated that 280,184 obesity-attributable deaths occurred in the U.S. annually. When risk ratios calculated for nonsmokers and never-smokers were applied to the entire population (assuming these ratios to produce the best estimate for all subjects, regardless of smoking status, i.e., that obesity would exert the same deleterious effects across all smoking categories), the mean estimate for deaths due to obesity was 324,940.

Additional analyses were performed controlling for prevalent chronic disease at baseline using data from the CPS1 and NHS. After controlling for preexisting disease, the mean annual number of obesity-attributable deaths was estimated to be 374,239 (330,324 based on CPS1 data and 418,154 based on NHS data).

Calle, Thun et al. selected their study subjects from over one million participants in the Cancer Prevention Study II, a prospective study of mortality among adults in the U.S. begun by the American Cancer Society in 1982 (67). Calle et al. examined deaths occurring between 1982 and 1996 among four cohorts: (1) current or former smokers with no history of disease3, (2) current or former smokers with a history of disease, (3) nonsmokers with no history of disease, and (4) nonsmokers with a history of disease. Weight categories were normal range (18.5-24.9), grade 1 overweight (25.0-29.9), grade 2 overweight (30.0-39.9), and grade 3 overweight (40.0+). All cause mortality, cardiovascular disease (CVD) mortality, and cancer mortality were examined.

The lowest mortality rates from all causes were found among study subjects having a BMI of between 23.5-24.9 for men and 22.0-23.4 for women. The risk of mortality increased with increasing BMI at all ages and for all categories of death. The strongest association between obesity and death from all causes was found among study subjects who had never smoked and had no history of disease, with the highest rates among the heaviest men and women, i.e., those with a BMI of 40+. The relative risk (RR) was 2.68 among men and 1.89 among women, compared with the reference groups (a BMI of 23.5-24.9 among men and 22.0-23.4 among women). This association was stronger in whites than among blacks.

Obesity was associated with higher mortality rates for both cardiovascular disease and cancer. BMI was most strongly associated with cardiovascular disease mortality among men (RR=2.90), but significantly increased risks of CVD death were found at all BMIs of greater than 25.0 in women and 26.5 in men. The findings showed an increase of 40% to 80% in risk of dying from cancer among both men and women in the highest weight categories.

Calle et al.’s study supports the need for further research to ascertain the differences in the effect of obesity on mortality among the black population, especially among black women. Their data also support the use of a single recommended range of body weight throughout life.

An earlier (1995) study by Manson, Willett, and Stamfer et al. examined data from the Nurses’ Health Study, looking at 4,726 deaths occurring from 1976 through 1992, 881 from cardiovascular disease, 2,586 from cancer, and 1,259 from other causes (68). A direct association was observed between BMI and mortality among women who had never smoked. Using a BMI of <19.0 as the reference group (relative risk [RR]=1.0), women with BMIs of 19.0-21.9 and 22.0-24.9 had a RR of 1.2; women with a BMI of 25.0-26.9 had a RR of 1.3; women with a BMI of 27.0-28.9 had a RR of 1.6; those with a BMI of 29.0-31.9 had a RR of 2.1; and those with a BMI of >32.0 had a RR of 2.2. Among never smokers, women with a BMI of >32 had a RR of 4.1 of dying from cardiovascular disease and a RR of 2.1 of dying from cancer.

3Cancer (excluding nonmelanoma skin cancer), heart disease, stroke, respiratory disease, current illness of any type, or a weight loss of at least 10 pounds in the preceding year.

Continue with Section One:

WV | DHHR | BPH | OCHSHP | HSC


If you have questions concerning sections one and two of this report, contact the West Virginia Health Statistic Center (HSC) at (304) 558-9100 or dhhrvitalreg@wv.gov.
If you have questions concerning secton three of this report, contact the Division of Health Promotion at (304) 558-0644.
I