Why Does Everyone Hate BMI?
Everyone seems to hate BMI and yet none of the myriad alternatives get much traction. Two new studies look at one of the core problems — a “U-shaped” relationship between BMI and mortality. At the extremes, low BMI and high BMI are both associated with the risk of an early death. But in the middle, the relationship gets muddled.
Cynthia Thompson and colleagues analyzed data from the Women’s Health Initiative, comparing the ability of BMI to predict mortality with two other indices: a body shape index (ABSI) and body adiposity index (BAI). They found that ABSI might be a better predictor of mortality risk because the relationship is linear. ABSI seems to do a better job than BMI of factoring in the effect of visceral fat because it uses waist circumference in addition to height and weight. Thompson explains:
ABSI was developed to integrate central adiposity with the health risk assessment equation, knowing that there is a subgroup within the population who maintain a healthy BMI but disproportionately carry weight in the form of central adiposity.
In Annals of Internal Medicine, Raj Padwal and colleagues provide some new insights on the risk of visceral fat that is independent of a high BMI. They found that the combination of a low BMI and a high body fat percentage can be predictive of considerable mortality risk. It underscores the fact that obesity is a disease of unhealthy visceral adiposity and BMI is far from being a perfect proxy for it.
And that brings us back to the contempt for BMI. The problem and the virtue of BMI is that it’s a simple screening tool. Height and weight are easily measured and usually readily available in medical records. Direct measures of adiposity — like DXA scans — are not so readily available. We’re not even close to routinely measuring waist circumference. So BMI will continue to be useful for screening. And it’s useful for epidemiology.
When we see the prevalence of BMI in the range of obesity and severe obesity growing, there can be no doubt that it’s a problem. When someone has a BMI in the range of obesity, it’s a signal of something that is worth a closer look. But BMI by itself cannot diagnose the health of an individual.
If people hate BMI it’s because they are expecting too much from it.
Click here for the study from Thompson et al, here for the study by Padwal, and here for more perspective from Arya Sharma.
Yuck, photograph © Valerie / flickr
Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.
March 19, 2016
March 19, 2016 at 7:02 am, Al Lewis said:
HL Mencken said: “For every problem, no matter how complex, there is a simple solution…and it’s almost always wrong.” BMI fits that category. It was invented by a mathematician almost 200 years ago and was never intended to measure individuals but rather populations, and was intended to see if a country’s people were getting enough to eat. Obesity was considered a good thing back then.
March 19, 2016 at 7:31 am, Ted said:
Good point Al. Thanks!
March 19, 2016 at 8:13 am, Allen Browne said:
I like that, too: “If people hate BMI it’s because they are expecting too much from it.” My wife, Nancy, calls this “getting lost in the weeds”.
March 19, 2016 at 8:17 am, Ted said:
Nancy is wise. Best to listen to her.
March 19, 2016 at 10:05 am, Susan Burke March said:
I love that quote too! BMI can be a screening tool, but not a diagnostic tool. I always use the example of Serena Williams or other muscular athlete. If you’d assess her risk for disease using BMI, her “obesity” would mean she’s at high risk…not!
March 19, 2016 at 10:42 am, Ted said:
Good perspective, Susan.
March 19, 2016 at 1:54 pm, Amy Endrizal said:
So glad you commented on this–I just finished listening to an Australian radio interview with one of the authors of the Annals study, and your additional insights are spot-on. For those facing a paywall to the journal article (or podcast fans), here is a transcript of the interview, broadcast on ABC Radio National’s weekly Health Report: http://www.abc.net.au/radionational/programs/healthreport/putting-the-obesity-paradox-to-the-test/7245016
March 19, 2016 at 2:56 pm, Ted said:
Thanks, Amy!
March 20, 2016 at 7:09 pm, shelley kay said:
BMI obscures the other parts of the body that contribute to the kg numerator, it’s not just about fat weight. With disease and aging, there is typically a loss of muscle and bone weight and an increase in visceral adiposity. An individual with normal or low BMI may have a high waist circumference (predictive of the visceral compartment) and the secretory function of visceral adipocytes is inflammatory. The worst phenotype is a low BMI with high visceral adiposity because this indicates that disease and/or inactivity has resulted in the loss of muscle mass. Clinicians who use BMI in this demographic are likely to miss signs of deterioration.
On the other side, individuals who exercise may decrease waist circumference but have no weight change even with fat loss determined by DEXA, CT or MRI . This is the best outcome but typically interpreted as “not successful” if BMI is used as an outcome measure.
When people have advanced disease, they lose weight, that doesn’t mean losing weight causes death and disease. There are analyses that take disease into consideration in the obesity paradox that change the interpretation.
In Australian children, BMI has been levelling off, suggesting that obesity rates are slowing. However, waist to height ratio has increased and abdominal skinfold measurements in children have increased. So for the same BMI, children have more abdominal fat and probably less muscle – not good! BMI obscures body composition and changes in fat and muscle are more meaningful than BMI. Dr Shelley Kay
March 20, 2016 at 7:19 pm, Ted said:
Thanks, Shelly.