Body Mass Index Above Ideal Range Linked to Large Increase in Mortality Rate CME/CE

News Author: Laurie Barclay, MD
CME Author: Désirée Lie, MD, MSEd

Authors and Disclosures

CME/CE Released: 03/17/2009; Valid for credit through 03/17/2010

March 17, 2009 — Body-mass index (BMI) above the ideal range may cause a large increase in mortality rates, according to the results of a collaborative analysis of 57 prospective studies reported in the March 18 Online First issue of Lancet.

"The main associations of...BMI with overall and cause-specific mortality can best be assessed by long-term prospective follow-up of large numbers of people," write Gary Whitlock, and colleagues from The Prospective Studies Collaboration at the University of Oxford, Oxford, United Kingdom. "The Prospective Studies Collaboration aimed to investigate these associations by sharing data from many studies."

The investigators analyzed the association of baseline BMI with mortality in 57 prospective studies enrolling a total of 894,576 participants, mostly in western Europe and North America, with median recruitment year 1979. Mean age at recruitment was 46 ± 11 years, 61% were men, and mean BMI was 25 ± 4 kg/m². The analyses were adjusted for age, sex, smoking status, and study, and the first 5 years of follow-up were excluded.

During additional follow-up (mean, 8 ± 6 years), there were 66,552 deaths of known cause, with mean age at death of 67 ± 10 years. Cause of death was vascular in 30,416 patients; neoplastic in 22,592; respiratory in 3770; diabetic, renal, or hepatic in 2070; and other cause in 7704.

Mortality rate was lowest with BMI at approximately 22.5 to 25 kg/m² for both men and women. At higher baseline BMI, there were positive associations for several specific causes and inverse associations for none. Absolute excess risks for higher BMI and smoking were approximately additive.

On average, each 5-kg/m² higher BMI was associated with approximately 30% higher overall mortality rate (hazard ratio per 5 kg/m² [HR], 1.29; 95% confidence interval [CI], 1.27 -1.32): 40% for vascular mortality (HR, 1.41; 95% CI, 1.37 - 1.45); 60% to 120% for diabetic (HR, 2.16; 95% CI, 1.89 - 2.46), renal (HR, 1.59; 95% CI, 1·27 - 1·99), and hepatic (HR, 1.82; 95% CI, 1·59 - 2·09) mortality; 10% for neoplastic mortality (HR, 1.10; 95% CI, 1.06 - 1.15); and 20% for respiratory (HR, 1.20; 95% CI, 1.07 - 1.34) and all other mortality (HR, 1.20; 95% CI, 1.16 - 1.25).

BMI less than 22·5 to 25 kg/m² was inversely associated with overall mortality rate, primarily because of strong inverse associations with respiratory disease and lung cancer. Although cigarette consumption per smoker varied little with BMI, these inverse associations were much stronger for smokers vs nonsmokers.

"Although other anthropometric measures (eg, waist circumference, waist-to-hip ratio) could well add extra information to BMI, and BMI to them, BMI is in itself a strong predictor of overall mortality both above and below the apparent optimum of about 22.5–25 kg/m²," the study authors write. "The progressive excess mortality above this range is due mainly to vascular disease and is probably largely causal. At 30–35 kg/m², median survival is reduced by 2–4 years; at 40–45 kg/m², it is reduced by 8–10 years (which is comparable with the effects of smoking). The definite excess mortality below 22.5 kg/m² is due mainly to smoking-related diseases, and is not fully explained."

Limitations of this study include lack of data on possible intermediate factors such as fibrinogen, C-reactive protein, and obstructive sleep apnea; possible confounding by diet, physical activity, or socioeconomic status; and lack of measures of central obesity.

"In adult life, it may be easier to avoid substantial weight gain than to lose that weight once it has been gained," the study authors conclude. "By avoiding a further increase from 28 kg/m² to 32 kg/m², a typical person in early middle age would gain about 2 years of life expectancy. Alternatively, by avoiding an increase from 24 kg/m² to 32 kg/m² (ie, to a third above the apparent optimum), a young adult would on average gain about 3 extra years of life."

UK Medical Research Council, British Heart Foundation, Cancer Research United Kingdom, EU BIOMED programme, US National Institute on Aging, and Clinical Trial Service Unit (Oxford, United Kingdom) supported this study.

Gary Whitlock was supported by a Girdlers' Health Research Council of New Zealand Fellowship. Writing committee member Sarah Lewington had a British Heart Foundation Fellowship to coordinate the project. Writing committee member Sarah Parish supplied unpublished analyses of BMI and cotinine. Writing committee member Nawab Qizilbash is a former director of epidemiology at GlaxoSmithKline and now works in Oxon Epidemiology, which has undertaken consultancy work for several pharmaceutical companies (including GlaxoSmithKline, Pfizer, Lilly, Roche, Gilead, and Grunenthal). All other members of the writing committee work in the Clinical Trial Service Unit, which has a policy of staff not accepting fees, honoraria, or consultancies. The Clinical Trial Service Unit is involved in clinical trials with funding from the UK Medical Research Council, British Heart Foundation, and/or various companies (AstraZeneca, Bayer, Merck, Schering-Plough, Solvay) as research grants to (and administered by) the University of Oxford.

Lancet. Published online March 18, 2009.

Clinical Context

Study Highlights

• Median year of recruitment was 1979, and 61% of participants were men.
• Mean recruitment age was 46 years, and mean BMI was 24.8 kg/m², but it was lower in European, Israeli, and Japanese studies.
• BMI at baseline was self-reported by participants and defined by the World Health Organization convention of BMI of 30 kg/m² or higher for "obese."
• Excluded were participants with stroke or heart disease at baseline or BMI below 15 kg/m², leaving 894,576 for analysis.
• For men and women, BMI was greatest at baseline at ages 50 to 69 years, with an increase in early adult life and middle age.
• The greatest increase was in men younger than 40 years and women younger than 50 years.
• BMI was positively linked with increased systolic and diastolic blood pressure, inversely with high-density lipoprotein cholesterol levels and positively with non–high-density lipoprotein cholesterol levels.
• Mean BMI was slightly lower in smokers and alcohol drinkers at baseline.
• The cause of death was determined from death certificates, medical records, and autopsy findings.
• Cross-sectional associations between BMI and risk factors were estimated by linear or logistic regression.
• Deaths in the first 5 years of follow-up were excluded to avoid confounders.
• During 6.5 million person-years of follow-up, 72,749 deaths occurred.
• In both sexes, all-cause mortality rate was lowest at a BMI of 22.5 to 25 kg/m².
• Above this level, all-cause mortality rate was, on average, 30% higher for every increase of 5 units of BMI and was greater at younger ages.
• In the lower BMI range of 15 to 25 kg/m², there was an inverse association between BMI and mortality, but this association was less evident when analysis was for nonsmokers.
• Ischemic heart disease accounted for more than one quarter of all deaths, and each increase in BMI of 5 units in the range 20 to 40 kg/m² was associated with a 40% increase in ischemic heart disease death, stronger in middle age but still evident at ages 80 to 89 years.
• Stroke accounted for one third as many deaths as ischemic heart disease, and each increase of 5 units of BMI in the range of 25 to 50 kg/m² was associated with a 40% increased risk for stroke.
• The HR attributable to heart failure was 1.86; for hypertension, 2.03.
• In the BMI range of 25 to 50 kg/m², the HR for death from liver disease was 1.79.
• The HR for cancer was lower and the association with BMI was weaker, with a 10% increase in mortality for every 5 units of increase in BMI.
• Site-specific HRs included 1.23 for kidney, 1.47 for liver, 1.15 for breast, 1.38 for endometrium, 1.13 for prostate, and 1.20 for large intestine cancers at different ages.
• In the lower BMI range of 15 to 25 kg/m², cancer deaths were inversely linked with BMI, with an HR of 0.52 for upper gastrointestinal tract cancers.
• Respiratory tract disease was inversely linked with BMI, with 4 times increase in death from COPD for BMI of 5 units or lower (< 25 kg/m²; HR, 0.31).
• In the lower range of BMI, COPD and lung cancer deaths in smokers accounted for the inverse association between BMI and mortality.
• In the range of 25 to 50 kg/m², the effects of BMI and smoking were additive vs multiplicative for vascular and all-cause mortality.
• For those with a BMI of 30 to 35 kg/m², median survival rate was reduced by 2 to 4 years, and for BMI of 40 to 45 kg/m², survival rate was reduced by 8 to 10 years.
• Avoiding an increase in BMI from 28 to 32 kg/m² in adult life was linked with an added 2 years of life expectancy; avoiding an increase from a BMI of 24 to 32 kg/m² was linked with 3 extra years of life.

Pearls for Practice