BMI is not a good predictor of cardiovascular risk

BMI or Body Mass Index is a measure of obesity that is used to approximate the health problems associated with being overweight. It is really easy to calculate. The formula for it is weight in kilograms divided by height in meters squared. Here are some calculators for those of your reluctant to mosey into the metric system. Anyway, after you calculate it values below 20 are generally considered underweight. Values between 20 and 25 are considered normal weight. Over 25 is considered overweight. Over 30 is considered obese (with various classes of obese determined by how large the value is).

However, there are several problems with BMI as a measure of obesity. The most notable is that you would really like a measure that fully accounts for your health risk. You don't want to be labelling people overweight only to discover that they actually are at lower risk for heart attacks than people who weigh less than them.

Unfortunately, this is precisely what happened. In a review of the studies published on the subject, Romero-Corral et al. show that BMI does not reflect an increased risk for Coronary Artery Disease (CAD) unless you are either very thin or severely obese. Here is a graph of the relative risk:

And a summary of the findings:

We found 40 studies with 250,152 patients that had a mean follow-up of 3·8 years. Patients with a low body-mass index (BMI) (ie, <20) had an increased relative risk (RR) for total mortality (RR=1.37 [95% CI 1.32-1.43), and cardiovascular mortality (1.45 [1.16-1.81]), overweight (BMI 25-29.9) had the lowest risk for total mortality (0.87 [0.81-0.94]) and cardiovascular mortality (0.88 [0.75-1.02]) compared with those for people with a normal BMI. Obese patients (BMI 30-35) had no increased risk for total mortality (0.93 [0.85-1.03]) or cardiovascular mortality (0.97 [0.82-1.15]). Patients with severe obesity (>35) did not have increased total mortality (1.10 [0.87-1.41]) but they had the highest risk for cardiovascular mortality (1.88 [1.05-3.34]).

This is probably profoundly counterinituitive for basically all of us because we have been told our entire lives that we need to lose weight or we are going to have a heart attack. But there are many reasons for this to be the case. Here is how they explain the discrepancy between BMI and the expected changes in risk:

An absence of association (or an inverse association) between obesity and death in patients with cardiovascular disease, previously coined "the obesity paradox", should be considered with caution before concluding that excess body fat is not a risk factor for coronary artery disease progression or complications once CAD has been manifest.

First, lower BMI values have been related to low lean mass, a condition also known as sarcopenia. Patients with sarcopenia or decreased muscle mass have a restricted exercise capacity and reduced mobility, both conditions associated with increased total mortality. Conversely, small raises of BMI as seen in overweight and mildly obese patients can be due to a preserved or increased lean mass and not to elevations in body fat. Preserved or increased lean mass have been associated with better fitness and exercise capacity, improvement in metabolic profiles and probably better prognosis in patients with cardiovascular diseases. Our results support this notion by showing that mildly elevated BMI was consistently associated with better outcomes.

Second, BMI might not adequately reflect adiposity. The inability of BMI to discriminate between an excess in body fat and increments in lean mass is certainly a plausible explanation of the better outcomes in overweight and mildly obese patients. Indeed, we have separately shown that a definition of being overweight as a BMI of 25 or greater has poor specificity (65%) to detect excess body fat in patients with CAD, and other studies in different patient populations have also shown similar results. Nonetheless, the higher the BMI, the better the discriminatory power for body fat and lean mass. In fact, the results of this meta-analysis show that severely obese patients had a significant 88% increase in the risk for cardiovascular mortality--probably due to the detrimental effect of excess body fat.

Third, low and normal BMI groups were consistently associated with a lower prevalence of established cardiovascular risk factors (with the exception of smoking). Therefore, these BMI groups were less likely to be a target of effective secondary prevention therapies--such as a healthy diet, exercise programmes, cholesterol reduction, or treatment for diabetes and hypertension. By contrast, obese and morbidly obese patients might receive more aggressive treatment for CAD because of their obesity itself. Practically all the studies included in this systematic review failed to adjust for type and dosage of medication received by BMI group, which can clearly affect the outcomes in this patient population.

...

Finally, extensive data have shown that central obesity poses a greater risk for cardiovascular disease than BMI alone. Nevertheless, most of the studies did not assess body fat distribution. In fact, two of the studies we identified in this study showed a direct association of central obesity but not BMI with an increased risk for total mortality. The INTERHEART study, a case-control study with data from 29 972 patients across 52 countries, showed a modest association between BMI and MI, whereas waist circumference and especially waist-to-hip ratio was a highly significant risk for MI.

Let me talk briefly about each one of the reasons. Each of them reflects a certain population of unhealthy or healthy people who are poorly evaluated under the BMI system.

• The Thin Elderly --The first one -- sarcopenia -- is a kind of wasting that is prevalent in old people. Basically, they are saying that the reason that people with very low BMI have high risk for CAD is that many of the individuals in this category have low levels of activity. This is reasonable because you very rarely see old ladies with osteoporosis running 5 miles a day. If exercise is necessary for good heart health, than these individuals may be unhealthy in a way that is not easily picked up by BMI.
• The Active Overweight or Moderately Obese -- The second one is related to the fact that there are a lot of active people out there who fall into the overweight category. I, for example, have a BMI of 27, and I like to run marathons. Being overweight does not necessarily suggest that you are in bad shape, and if the positive effects of regular exercise outweigh the negative effects of obesity -- and it looks like they do -- then many of these people will be fine. BMI being primarily a predictor of weight is not a good measure of muscularity or what kind of shape someone is in, and these things matter in determining risk.
• The Waifish and Unhealthy -- The third one is interesting. There they are basically saying that we assume thin people in this society are healthy when that is really not necessarily the case. For example, you could have a very thin person who smokes like a chimney. They are really raising their cardiovascular risk, but if you look at them -- and more importantly measure their BMI -- you would miss that. Because overweight people are more likely to have other risk factors, they have doctors riding them all the time to stop. Thin people get off easy.
• People Who Carry Their Weight Low -- Finally, we know that fat distribution matters. You remember all that stuff about apple shaped people and pear shaped people. Well, we know that where fat is on your body matters in determining your risk. If you for instance carry your fat in your butt, then the risk associated with being overweight is significantly less than if you carry it in your chest.

All of these are reasonable explanations for why BMI is a relatively blunt instrument for determining CAD risk. The paper they mention at the end -- the INTERHEART study -- suggests that waist circumference or waist-to-hip ratio might be more effective ones.

Just to be clear, the take home from this article is not that obesity is not a risk factor for cardiovascular disease. We know from the Framington Study that not only is it associated with a variety of other risk factors, obesity is also in itself a risk factor. In addition, there are numerous other ways that obesity increases morbidity and mortality.

This article is important, however, because it indicates that the common measure that we are using to evaluate obesity is not particularly helpful. If we want to save lives, we need to focus on those traits of obesity that most raise risk. We also need to confront people who are ostensibly not overweight, but still are at high risk.

Here is their take home:

From the clinical standpoint, a fundamental question is not whether BMI is associated with higher mortality and cardiovascular events in patients with coronary disease but whether weight loss or maintaining a normal weight can decrease cardiovascular events in patients with established CAD. Small trials in cardiac rehabilitation using exercise and diet as the main intervention, but not focused on weight loss, have shown a reduction in the rate of revascularisation.24 Because exercise and diet are the main components of a weight-loss programme, we can assume that such interventions in overweight patients with coronary disease will reduce cardiovascular outcomes, including mortality. However, no randomised controlled trials have yet been undertaken to address this relation.

...

Rather than proving that obesity is harmless, our data suggest that alternative methods might be needed to better characterise individuals who truly have excess body fat, compared with those in whom BMI is raised because of preserved muscle mass. We should remain concerned about patients with coronary artery disease who have a true excess of body fat, rather than a raised BMI per se. Pending further information, patients with coronary artery disease who are truly obese should be encouraged to pursue interventions that reduce body fat. Additional studies with different methods are needed, including randomised clinical trials with different weight-loss strategies, or using other ways to identify obesity, such as body composition techniques to measure body-fat percentage and distribution that also accounts for lean mass.