Abstracts & Commentary
With Comments by Jonathan Abrams, MD, Professor of Medicine, Division of Cardiology, University of New Mexico, Albuquerque. Dr. Abrams is on the Editorial Board of Clinical Cardiology Alert.
Synopsis: Overweight is associated with no excess mortality, particularly in the older age groups.
Sources: Flegal KM, et al. JAMA. 2005;293:1861; Gregg EW, et al. JAMA. 2005;293:1868-1874.
A highly publicized report from the CDCP and National Cancer Institute has concluded that being overweight (but not obese) does not appear to be associated with decreased total mortality. This is a complex assessment of 3 National Health and Nutrition Examination Surveys (NHANES I, II, III) which represent a snapshot of the American population during the years of each particular survey. These data reflect NHANES surveys from 1971-1975 and 1976-1980, with follow-up in 1992 and 1988-1994. The focus of this report is on BMI, or body mass index (weight in kg/height in meters). Multiple CAD risk factors were elevated. The end point was the relative risk of death attributable to various BMI levels, as derived from the 3 NHANES studies. “Estimates were made from the combined data. . . to represent the US population during the 20 year period year covered by these surveys.” The results reflect a number of important findings. Obesity, defined by a BMI > 30, was associated with an increased risk of death, particularly at younger ages and with a BMI > 35. However, the relative risk in the overweight cohorts (BMI 25 to < 30) was lower than unity. Of interest, the mortality risk of underweight (BMI < 18.5) was actually higher than unity. The number of excess deaths in 2000, related to all BMI categories (the total data and as well as each individual NHANES survey), demonstrated a consistent pattern, with excess deaths greater than unity in the underweight category, less than unity for the overweight category, and increasing considerably with obesity. Furthermore, the greater the BMI in the obese population, the greater the number of excess deaths, indicating that marked obesity is truly hazardous. On the other hand, “for overweight (BMI to < 30), the data consistently suggested no excess deaths overall.” Thus, excess mortality in the obese population was derived mostly from a BMI > 35, and particularly in individuals younger than 70.
Visual examination of the graphs dealing with overall mortality by BMI, NHANES year, and age, indicates an extremely small difference between the normal BMI population of 18.5 to < 25 and the overweight category of BMI 25 to < 30, the latter was slightly lower but statistically significant. In patients followed for greater than 10 years, the category of overweight was associated with no excess risk, whereas the underweight cohort continued to show an increased relative risk of mortality, particularly in the elderly. Flegal and colleagues point out that these results differ from many previous reports which have been widely publicized over the past several years. They also stress that overall mortality has decreased over time within each weight category of the NHANES reports, suggesting “that the improvements in medical care, particularly for cardiovascular disease, the leading cause of death among the obese, and its risk factors may have led to a decreased association of obesity with total mortality.” They point out that cardiovascular risk factors have declined in all BMI levels in the United States (except for diabetes). Age adjusted death rates for CVD continue to decline.
Another NHANES analysis in the same issue of JAMA (Gregg et al. JAMA. 2005;293:1868-1874) confirms in a separate analysis that indeed CV risk factors have declined considerably over the past 40 years in all BMI groups. This report also concludes that obese persons still have higher risk than those who are lean, but the overall levels of risk are lower than in past years.
Particularly, cholesterol levels, smoking rates, and hypertension all markedly decreased over the time span of the analysis. Total diabetes prevalence was stable within the BMI groups, with a small increase in the last 24 years; Flegal et al suggest that the improved life expectancy in the United States is related to decreases in CV death. A mildly elevated BMI was not per se associated with decreased survival from ischemic cardiac disease. Obesity (BMI > 30) “is associated with a modestly increased risk of mortality,” 1-2 fold. No increased mortality was noted in any of the 3 NHANES surveys for the BMI cohort of 25 to < 30. All analyses demonstrate a J shaped curve, with minimum mortality close to a BMI of 25. Other data also “have suggested that overweight is associated with no excess mortality, particularly in the older age groups.” Flegal et al stress that many studies confirm that there is a serious impact on health, morbidity, and disability in overweight and obese individuals, “disproportionately borne by younger adults.” Nevertheless, overall obesity risk of mortality has declined since NHANES.
The differences in survival between the first NHANES report and later surveys “suggest that the association of obesity with total mortality may have been decreased over time, perhaps because of improvements in public health or medical health or medical care for obesity-related conditions.”
Comments
These 2 reports, particularly the analysis of total mortality by Flegal et al, engendered enormous publicity. These are somewhat difficult studies to read, replete with a great deal of statistical methodology and discussion about the residual attributable fraction of relative risk and many other epidemiologic terms. For the practicing physician, the second report (Gregg et al) relating to secular trends in cardiovascular disease risk factors according to BMI is perhaps of greater importance, documenting substantial decreases in major CAD risk factors in a long term assessment of multiple data bases going back as far as 1960. Thus, smoking, elevated cholesterol, and hypertension are considerably decreased among all BMI groups, although diabetes has increased. For instance, the absolute reduction in high cholesterol between 1960-2000 was 20% lower in obese or overweight persons, with a higher percent decline than in lean persons. Comparable changes were found for hypertension, which “tended to decline more over time among obese and overweight persons than among lean persons.”
Nevertheless, hypertension, overall, remains at least twice as common in obese subjects than in the lean individuals. Smoking rates were higher in lean individuals than in obese persons, declining in all BMI groups. Total diabetes was stable over time and increasing only by 1-2% points between 1976-1980 and 1999-2000, although “the prevalence of diagnosed diabetes was about 2.5 to 3.5 times as high in 2000 as in 1960 among overweight and obese.” Hypertension and lipid lowering medications, not surprisingly, have shown a marked increased in recent years among the obese. Flegal et al conclude “thus, obese and overweight persons may be at lower risk of CVD now than in previous eras,” with the exception of diabetes. Undiagnosed diabetes increased in the obese. They point out that “BMI is only one among many determinates of CVD risk.” What is clearly missing from these analyses is the proportion of overweight and obese individuals who meet criteria for the metabolic syndrome. That particular cluster of abnormalities is associated with increased cardiovascular morbidity and probably mortality; many but certainly not all, individuals classified as overweight or obese have the metabolic syndrome. This, and the increased prevalence of diabetes in overweight and obese patients, indicates that one cannot and should not ignore these data. Careful search for components of the metabolic syndrome (dyslipidemia, waist circumstance, presence of hypertension, presence of elevated fasting blood sugar) is mandatory, with very aggressive risk factor modification in such individuals. Also, these data clearly confirm that obese individuals (BMI > 30), particularly those with marked obesity, continue to have a very high risk of total and CVD mortality. Aggressive risk reductions therapies, including bariatic interventions in selected individuals, are mandated. Overall, I conclude that these data are good news, removing some of the health stigma away from people who fit into the category of overweight (BMI 25-30), and documenting that in fact, we are and have been paying attention to CV risk factors, with improved treatment of hypertension and dyslipidemia, as well as lower smoking rates. There is positive information for many individuals in these reports, except for the truly obese.
Overweight is associated with no excess mortality, particularly in the older age groups.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content