Lean Really Is Better

An Important, Hard Look at BMI and Your Health

Woman weighing herself
Flashpop/Getty Images

Intuition would likely suggest that a “normal” body mass index (BMI) would also be the healthiest. That, after all, is what normal in this context really ought to mean: the “right” range for healthy people.

Just such reasoning underlies the prevailing tiers of the BMI scale. The BMI, which is the weight of the body in kilograms divided by height (in meters) squared, is essentially a sophisticated weight-to-height ratio. Values from 18.5 to 24.9 are considered “normal” for adults; below 18.5 is underweight. From 25 to 29.9 is overweight, while values of 30 and higher constitute obesity in three tiers of successive severity. (N.B.—The BMI is intended to work for the population, and on average. Any given individual might have a low BMI because they are a high-endurance athlete, or a high BMI because they are a high-strength athlete. The BMI is blind to such distinctions and should be interpreted accordingly.)

The Stratification of Weight

To some extent, this stratification of weight is informed by intuition and observation. While there is some variation in normal BMI among populations around the world, the normal range of values is where the weight of reasonably healthy, reasonably active people eating a diet of reasonably wholesome foods almost always tend to settle. Because of variations in build, some ethnic groups settle near the higher end of the range, others nearer the low end. But the range is reasonable on the basis of global and time-honored norms.

But there was a stronger basis than that for these particular values. Studies going back decades have suggested that mortality and disease risk rise when weight is too low or too high. That begs the question: too low or too high compared to what? The answer is: compared to the range of weight (or BMI) associated with the lowest risk of chronic disease or premature death. It was from just such analyses that the current scheme was derived. In fact, the scale was revised less than 20 years ago to better reflect the available data.

Back to “Normal”

So far, this all sounds pretty straightforward. Actually, though, the topic has been quite controversial for much of those past 20 years. Some, fighting against obesity bias, have objected to the idea of a “normal” weight for fear of stigma. But while opposing obesity bias is very important, that argument is weak. Having a normal range for blood pressure, or blood sugar, does not stigmatize values outside of those ranges—it just helps to identify them and address health risks accordingly. Weight should be treated the same, even if we have work to do to make sure that it is.

Another argument has been that fitness matters more than fatness, and weight is relatively unimportant in people who are otherwise healthy. This argument is also valid, but weak for two reasons. First, most people who are truly fit tend not to be fat. Second, research shows that among people who are comparably fit, there is additional benefit to being lean rather than overweight.

The third argument, however, has been the most important: It contends that the ranges are simply wrong. Studies in older adults have often suggested that mortality risk is lowest not in the “normal” weight range, but in the “overweight” range—suggesting that the ranges themselves are misnamed. This contention underlies versions of the “obesity paradox” theory and suggests that for some at least, being overweight protects health.

Addressing a Blind Spot in Weight Research

Many, and I among them, have been concerned for years that studies suggesting advantages of overweight, while maybe giving people news they were happy to get, were missing something very important. Namely, sick people routinely lose weight. So, among older people, it stands to reason that those who remain somewhat overweight are likely to be better off than those who were once overweight, then lose weight and become “lean;” that weight loss might be inadvertent and due to the smoldering of an undiagnosed disease.

This has been an important blind spot in many studies, and despite some efforts to adjust for it, the problem and debate have persisted. But they may be over now at last, due to a large study published in the Annals of Internal Medicine.

This time, the investigators looked not just at weight, but at peak weight achieved, and weight change over time. What they found, in some 225,000 people followed for more than a decade, was both clear and compelling. Those adults whose peak weight was in the normal range and stayed there had the lowest mortality risk. A peak weight in the overweight range elevated that risk, whether you were still overweight now or currently lean. That last group—formerly overweight, now lean—is the very group many of us have been worrying about all along. If weight comes down because of a new commitment to eating well and being active, it’s a good thing; but when it comes down for other reasons, it is all too often an ominous sign.

Coming Full Circle

This study suggests that what we thought we knew about healthy weight ranges 20 years ago was right all along, and that the controversy—as is often true where public health science and intense popular interest collide—has involved a whole lot of heat and very little light. Under the bright light of the new research, it is affirmed that lean (i.e., BMI in the normal range) is indeed generally healthier.

So, now we know where “there” is. We should be redoubling our efforts to make it easier for more people to get there from here because we’ve got a long way to go.

To determine your BMI, enter your information into our calculator here: