Is Weight Good for You?

For those interested in weight issues, new Canadian research shows, again, that being “overweight” — having a body mass index (BMI) of 25–29.9 — is associated with a longer lifespan than that of “normal” weight people (BMI 18.5–24.9). According to the study, people who are mildly “obese,” (obesity class I, BMI 30–34.9) have the same risk of death as normal weight people. Not until weight reaches obesity class II (BMI 35–39.9) does heaviness seem to increase mortality.


A BMI of 30 for a person who is 5′ 8″ would be a weight of roughly 200 pounds. The same person would reach BMI 35, where obesity class II starts, at 230 pounds.

The highest death rates in the study occurred for those who were “underweight,” with a BMI of less than 18.5. (For our 5′ 8″ sample person, that would be a weight of 122 pounds or less.) You can calculate your own BMI at this Web site.

So how does this fit in with the constant warnings that weight is killing us? It is widely accepted that obesity is strongly associated with, and probably helps cause, Type 2 diabetes, heart disease, and increased mortality, among other health issues. But at least in the case of mortality, this study and several others seem to contradict these assumptions.

The “Obesity Paradox”
The scientific establishment refers to this disconnect as “the obesity paradox.” “Paradox” is the term scientists use when the facts don’t fit their theories. The “obesity paradox” originally referred to the tendency of heavy people with chronic conditions like heart failure to live longer than lean people with the same diseases.

“Paradoxes” often reflect problems with the underlying theories, but experts hate to admit this. The “French Paradox” refers to the fact that the French have a fairly low incidence of heart disease, despite having a diet high in saturated fats. Red wine consumption, eating more slowly, and having more family support have all been suggested as reasons for the “paradox.” Wine, relaxation, and support probably do provide health benefits. But maybe the theory about fat consumption is also mistaken. Few experts want to consider that perhaps fats are not so bad for you after all.

In other “paradoxes,” it’s the new information that’s wrong, not the theory. One could call the recent claims that tight glucose control leads to higher death rates a “paradox.” But this would ignore the fact that drug side effects, not low glucose levels, most likely caused those excess deaths.

I wonder how many paradoxes it takes before the underlying theory is changed. Usually it takes quite a few, and in the case of the “weight kills” paradox, it may never change.

What the Study Showed
The study involved over 11,000 participants in a 1994/1995 health survey of Canadians. The researchers calculated BMIs from participants’ self-reported height and weight. They looked to see who had died by the end of 2007 and adjusted for various factors that could affect the risk of death, such as age, sex, smoking, physical activity frequency, and alcohol consumption. The study was published in Obesity, a peer-reviewed medical journal.

Within the studied sample, 1,929 deaths were observed during the 12-year follow-up period. Analysis found that underweight people had 1.73 times the risk of dying as “normal” weight people. Overweight people had only 0.83 times the risk of dying, the best result of any group, while even the mildly obese people’s average was 0.95, slightly less than that of the normal weight people. People whose weight fell into obesity class II had 1.36 times the risk of dying relative to those in the normal weight group.

This study appears to say significant amounts of extra weight are good for you. But is this true, or are the problems with the findings?

One problem is that BMI is not really a good measure of body fat. Some people are naturally wider, or have heavier bones, or more muscle, so they weigh more and therefore have higher BMIs. Muscle mass could explain a bit of the high-BMI/low-mortality connection, but probably not much.

The big problem with the study is that the follow-up time of 12 years is too short. As cause of death was not reported, perhaps the “normal weight” people died from risky behaviors, or infectious disease, and “overweight” people took fewer risks. With a high proportion of younger people in the survey, perhaps the chronic illness deaths hadn’t had time to kick in. Perhaps by the time the participants reach their 70s, the heavier people will be dropping like flies. (But personally, I doubt it.)

The National Institutes of Health report some information on aging and weight. Mortality rates naturally increase with age, but they seem to go up faster for heavy people.

True Up to a Point
Please note that the benign effect of overweight didn’t extend past BMI 35 in this study. The very heavy are still considered at risk. But it is also possible that the weight effect has been overstated, and it’s really an “inactivity effect.” Quite a bit of evidence indicates that physically active people live longer, regardless of weight.

Quite possibly, the real reason for the “paradox, ” is that “normal” values have been set too low. They’re not really normal. Perhaps the top of normal should be set at a BMI of about 32.

My main point here is how slowly science changes its mind when it’s wrong. People get attached to their ideas and won’t change them. Richard Bernstein, MD, CSW, author of Diabetes Solution, told me that this is why the ADA still recommends a high-carbohydrate diet. He told me the recommendations won’t change until “all the people who first put them out have died.”

Learn more about the health and medical experts who who provide you with the cutting-edge resources, tools, news, and more on Diabetes Self-Management.
About Our Experts >>

  • gys de Jongh

    A high BMI reduces your chance of survival and a BMI between 22.5 and 25 is still the best.

    The relation between BMI and survival is a complicated by confounding with smoking and thus can be found back only in very large, long lasting studies, because most studies agree that the effect of smoking cannot be compensated for just linearly.

    Here is a recent meta analysis of 57 prospective studies in The Lancet.The abstract is free, the article is free if you register :

    The Lancet, Volume 373, Issue 9669, Pages 1083 – 1096, 28 March 2009

    Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies.

    Editors’ note: Around the world, increasing body-mass index (BMI) is a major public concern. Rightly so, according to this international collaborative analysis of almost 1 million people, followed from middle age in 57 prospective studies. A J-shaped mortality curve is observed, with optimal survival at a BMI of 22•5–25 kg/m2. Above this range, mortality from several causes—especially vascular diseases—was increased. Moderate obesity (BMI 30–35) was associated with 3 years’ loss of life. People with extreme obesity (BMI 40–50) lost 10 years of life, equivalent to the years lost by lifetime smoking.

    In both sexes, mortality was lowest at about 22·5—25 kg/m2. Above this range, positive associations were recorded for several specific causes and inverse associations for none, the absolute excess risks for higher BMI and smoking were roughly additive, and each 5 kg/m2 higher BMI was on average associated with about 30% higher overall mortality

    Although other anthropometric measures (eg, waist circumference, waist-to-hip ratio) could well add extra information to BMI, and BMI to them, BMI is in itself a strong predictor of overall mortality both above and below the apparent optimum of about 22·5—25 kg/m2. The progressive excess mortality above this range is due mainly to vascular disease and is probably largely causal. At 30—35 kg/m2, median survival is reduced by 2—4 years; at 40—45 kg/m2, it is reduced by 8—10 years (which is comparable with the effects of smoking). The definite excess mortality below 22·5 kg/m2 is due mainly to smoking-related diseases, and is not fully explained.


  • mike

    i recently was taking in low carbs no sugar.then one after noon i consumed some very sweet tasting frozen alcohol drinks and a plate of french fries.i was was unusaly drunk by my standards i couldnt speak thought it was night time it was 6pm.passed out 3 hours was weak the next day and not really hung over..beat ..can a spike in sugar .any thoughts

  • David Spero RN

    Thanks for posting this Lancet article, Gys. I’m not sure it contradicts the finding that BMI 25–30 did not increase mortality, but decreased it for the 12-year follow-up. The large Canadian study I’m citing did control for smoking (self-reported smoking.)

    The problem with giant surveys like the Lancet one is that they can only tell you what’s true for the average. But people vary a lot — some people can be very heavy and still have normal glucose, cholesterol, and blood pressure. Others can’t. I think you have to look at those metabolic numbers before deciding if you’re really too heavy.

    Mike, that’s an interesting story. I wonder what was in those drinks! I know the combination of alcohol and out of control glucose can do things to your mind. That may have been what happened to you. But I’m no expert. Does anyone else have some insight for Mike?

  • Nuno


    I read the first paragraph and it’s all wrong. wrong I say! Then I looked up to the blog title “diabetes self-management”. Don’t fool yourselves: having high BMI is NOT good for your health.

  • skeptic

    There are a number of holes in your theory. The risks associated with the flawed BMI are not in adipose tissue, but found with lean body mass. Your belief in metabolic risk factors is equally flawed and unsupported in the research.

  • Israel

    1) Very thin people tend to be sicklier, often from young ages. Was that factored out?
    2) How were they figuring BMI for these studies: Off a chart, or by water displacement? Muscular people can be obese based on the chart, only overweight based on displacement.
    3) Skeptic: “Your belief in metabolic risk factors is equally flawed and unsupported in the research.” Yagottabekiddingme!