‘Plump is
good – obese is bad’: A misleading
conclusion
Last November (2007), a research article appeared that was excerpted and summarized in many popular health columns. The New York Times Op-Ed page reported, “Being obese is bad for you in pretty much every way. But being chubby may be a wash – or even better than being skinny – at least when it comes to longevity.”
Many people have responded with relief, and have eased up on their resolve to lose weight. Here are two reasons why that might not be such a good idea:
Individual weight is determined by genetic factors, as well as diet and exercise. The research study made no effort to separate these two. People who were genetically inclined to be obese were lumped together with people with the munching couch potatoes.
But in animal experiments, excess weight has opposite effects depending on whether it results from overeating or from genetic factors. When genetically obese mice are maintained on a restricted diet, they end up looking a little hefty, and yet they live longer than ‘normal’ mice. They even live longer than the mice that were genetically normal, but maintained on a restricted diet so that they look emaciated.
If the same physiological effect applies to humans, we’d expect people who are congenitally overweight to have a very different risk profile from people who eat to excess.
We know that humans vary greatly in the way they metabolize food. Some of us seem never to get fat, no matter how much we eat, while others have to diet constantly just to look ‘acceptable’.
The need to look ‘acceptable’ may be crucial here. There is a social stigma attached to being overweight. At any given time, 65% of American women and 59% of American men report that they are actively trying to lose weight. ¾ of overweight people report that they are dieting, exercising, or both. This may well account for the fact that people who appear moderately overweight aren’t at increased risk compared to people in the normal range.
People who are overweight are more likely to be dieting. But it is not your absolute weight that determines your mortality risk; it is your weight relative to your genetic type. In population statistics, it may appear that people who are underweight are not any better off than people who are normal weight or slightly overweight. But for any given individual, it is true that losing weight will have benefits for health and longevity. There is no contradiction between these two statements.
There is a second artifact in the study: Data was drawn from surveys over a 28-year period, and pooled together for analysis. During this time, people were getting fatter, even as medical technology was improving. People who entered the study later were more likely to be overweight, and they were also more likely to have benefited from advances in medicine.
There is a good reason to
believe that humans behave like mice (and other animals) in lab experiments,
where they live longer the less they are fed.
A research lab at
In summer of 2006, a study of cancer
survival made news headlines, demonstrating dramatic effect of weight on
survival statistics. After 7 years, 70%
of the patients in the thinnest group were still alive (BMI<18.5), while none of the patients in the heaviest
group (BMI>30) were alive. (So why are cancer patients routinely counseled to ‘keep your weight
up’?)
The bottom line is that if you are overweight, you will benefit most from trimming down. Your health should improve dramatically, and your life expectancy will jump from the lowest to the highest group. The more overweight you are, the more you benefit from each increment of weight loss.
But even if you are ‘normal weight’, your health and your life expectancy will both improve from eating less and exercising more.
On the other hand…
I am convinced that keeping your weight as low as you can tolerate is one of the best things you can do for your vitality and longevity, but here are two confounding facts that support skepticism of this view:
1. In studies of weight loss, it has been maddeningly difficult to demonstrate that people who lose weight do better than a comparable group that maintains their weight. Some studies seem to show this, while others seem to show that people who lose weight actually face a higher mortality risk! You can’t do the experiments you can with animals, starting with groups that are genetically homogeneous and randomly dividing them into two groups, feeding one a lot and the other a little. It may be that it is just too difficult to disentangle smoking, sociology, and all the confounding life-style variables.
2. My thesis above is that people who are overweight are likely to be dieting,
and this explains the fact that there is no apparent decrease in longevity as
we move from BMI=20 up through BMI=30.
But women are more likely to be dieting than are men, for social
reasons, while, in fact, it is for men that the relationship between longevity
and BMI fails more distinctly.
Two
articles were published in the Journal of the American Medical Association by
the
JAMA
298:2028-37 (2007) and JAMA 293:1861-1867 (2005), both by K.M. Flegal
et al.
The
data they analyzed came from NHANES I, II, and III (National Health and
Nutrition Examination Survey) spanning the years 1976-1994. BMI was determined from initial weight and
height measurements, and mortality data from many years later was correlated
with the historic BMI data. BMI<18.5
is defined as underweight, 18.5-25 as normal, 25-30 as overweight, and >30 as
obese. It is known that the way BMI is
calculated tends to be “unfair to tall people”, i.e., it assigns a higher
number to people who are taller, relative to people who are shorter and have
the same build.
Other
references: