You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Primary Care Corner with Geoffrey Modest MD: Central Obesity

2 Dec, 15 | by EBM

By Dr. Geoffrey Modest

There are a plethora of older studies showing that central obesity has a much stronger association with cardiovascular and mortality outcomes than BMI. The role of BMI as a predictor of events is largely explained by the concomitant risk factors of hypertension, hyperlipidemia, glucose intolerance/diabetes. Central obesity, on the other hand, is associated with the more metabolically active visceral fat. Not surprisingly, several studies have shown that there is a strong correlation between an increased BMI and central obesity for most individuals. However, little attention is paid to people with normal BMI but central obesity, and the 2013 Am Heart Assn Obesity Society guidelines on obesity management only recommends checking waist circumference in those with high BMIs. In this context, there was a study looking at the clinical consequences of central obesity in individuals with normal weight (see  doi:10.7326/M14-2525​), using the NHANES III database (Third National Health and Nutrition Examination Survey).


  • 15184 adults (mean of 40 yo, 52.3% women, mean BMI 27, mean waist circumference 94 cm men/87 cm women, mean waist-to-hip ratio 0.94 men/0.85 women, 85% white, 11% African-American, mean BP 120/75, hypertension in 29%, diabetes 7%, history of  MI 3%, A1C 5.2%, LDL 123, 50% physically active) had general obesity assessed by BMI and central obesity by waist-to-hip ratio (WHR)
  • Of those with normal BMI (<25), 322 men (11.0%) and 105 women (3.3%) had increased WHR (>1.0); of those who were overweight by BMI (25-30), 1064 men (37.0%) and 289 women (12.0%) had increased waist-to-hip ratio (WHR). Of those who were obese (BMI>30), 928 men (63.0%) and 336 women (14.0%) had increased WHR. Overall analysis showed that both waist circumference and WHR were strongly correlated with BMI, though a little less so with WHR than waist circumference
  • Primary outcome: total and cardiovascular mortality, after adjustment for confounding factors (age, sex, education level, smoking history)

Results, using a BMI of 22 to represent people with normal BMI:

  • Over a mean follow-up of 14.3 years, there were 322 deaths (1413 women), 1404 due to cardiovascular disease
  • For a man with normal BMI and central obesity vs similar BMI and no central obesity: mortality risk HR 1.87 (1.53-2.29) [ie, 87% increase]
  • For a man with normal BMI and central obesity vs overweight man by BMI: mortality risk HR 2.24 (1.52-3.32), or vs obese man HR 2.42 (1.30-4.53)
  • For a woman with normal BMI and central obesity vs similar BMI and no central obesity: mortality risk HR 1.48 (1.35-1.62)
  • For a woman with normal BMI and central obesity vs overweight woman by BMI: mortality risk HR 1.40, or vs obese woman by BMI HR 1.32 (1.15-1.51)
  • Overall, “WHR, but not BMI was associated with high mortality risk”

A few points:

  • There have been a slew of metabolic differences found between visceral and peripheral fat, perhaps related to the fact that visceral fat products enter the portal circulation, are accumulated in the liver, and affect hepatic processes. For example, central obesity is associated with systemic inflammation (perhaps the reason there are greater associations with MI, as well as with diabetes, Alzheimers, etc). Also with more small, dense and more atherogenic LDL particles, hypertriglyceridemia, thrombotic risk factors, insulin resistance.
  • The INTERHEART study of cardiac risk factors in 52 countries found that the population-attributable risk to MI was 24.5% for central obesity but only 7.7% for BMI (see Lancet 2005; 366: 1640)
  • Central obesity (as opposed to BMI) is part of the definition of Metabolic Syndrome, and in the US/Western Europe is defined as abdominal circumference >40 inches in men (102 cm) or >35 inches in women (88 cm), though there are international definitions which vary by ethnicity, per the International Diabetes Federation recommendations (e.g., central obesity in South Asians is 90 cm for men and 80 cm for women).
  • The association between central obesity and cardiovascular risk could not be explained by the standard cardiac risk factors

So, what is the utility of waist circumference or other measures of central obesity vs BMI? It is clear from almost all of the studies that central obesity is more important as a risk factor than BMI. And, I think we should be using some measure of central obesity in our assessment of patients. However, BMI is still useful in that it provides a very easy measurement for providers and patients to track: it is probably more helpful to look at weight and BMI changes when patients are trying to lose weight, since we can appreciate very small changes and use that for motivational interviewing/reinforcing diet and exercise. Waist circumference or WHR will change more slowly and might therefore prove to be discouraging to providers and patients alike, and is also probably harder to perform with consistent accuracy. So, I think we should measure both, be especially aggressive in management of other cardiac risk factors in those with central obesity, and use changes in weight/BMI (including in those with normal BMI but central obesity) to reinforce the important lifestyle changes.

By submitting your comment you agree to adhere to these terms and conditions
You can follow any responses to this entry through the RSS 2.0 feed.
EBM blog homepage

Evidence-Based Medicine blog

Analysis and discussion of developments in Evidence-Based Medicine Visit site

Creative Comms logo

Latest from Evidence-Based Medicine

Latest from EBM