August 28, 2006
Two Studies Find Being Overweight Shortens Life Expectancy

A pair of studies in the New England Journal of Medicine find that contrary to a controversial CDC study (more on it here) released last year being overweight (which is less heavy than being obese) really does shorten life expectancy.

Being overweight during midlife is associated with an increased risk of death, according to a new study conducted by the National Cancer Institute (NCI), part of the National Institutes of Health, in collaboration with AARP, the nation's leading organization for persons 50-Plus. Results of the study appear in the August 24, 2006, issue of the New England Journal of Medicine*.

Previous research had established a link between obesity and increased risk of death, but whether a relationship also existed between being overweight and increased risk of death remained uncertain. In 2004, the Centers for Disease Control and Prevention (CDC) reported that 34.1 percent of the U.S. adult population was overweight, but not obese**. Overweight and obesity are defined using a measurement called body-mass index (BMI), calculated as a person's weight divided by the square of their height. A BMI of 18.5 - 25.0 is considered normal, whereas people who have a BMI of 25.0 - 29.9 are considered overweight, and individuals with a BMI over 30.0 are regarded as obese. (Click to view BMI chart)

The NIH-AARP Diet and Health Study of 527,265 monitored the health status of Americans from 1995 through 2005 via mailed questionnaires and by surveying death records. When the analysis focused on BMI at age 50 among persons who had never smoked, the researchers, led by Kenneth F. Adams, Ph.D., of NCI's Division of Cancer Epidemiology and Genetics, found that the risk of mortality among participants who were overweight increased by 20 to 40 percent. Mortality risk among obese participants increased two to three-fold.

"BMI at age 50 gives a more accurate representation of the amount of excess fatness a person was exposed to over decades," said Michael F. Leitzmann, M.D., NCI, senior author of the study. "On the other hand, body weight reported at age 65, for example, might reflect a recent weight loss due to cancer or other disease. If that person then dies the next year, it would be inaccurate to classify that person in the normal BMI range when their pre-cancer BMI was actually overweight."

I think their focus on BMI before major illneseses develop is the correct approach. As people get older and sicker health problems start influencing body weight and weight becomes as much an effect as a cause of health status..

They also aimed at getting data on non-smokers in order to avoid weight reduction caused by smoking.

Some earlier studies have shown that being overweight was not associated with an increased risk of death. However, smoking and chronic illness are associated with a lower BMI and an increased risk of death, which may distort the relationship between BMI and mortality.

An advantage of the current study was the availability of data on more than 186,000 male and female participants who had never smoked. This allowed the researchers to untangle the complex relationships between body weight, smoking, existing disease, and risk of death. Other possible confounding factors that were accounted for included age, race or ethnic group, level of education, physical activity, and alcohol intake.

This result is not surprising given what we know about blood lipid levels and other aspects of metabolism among the overweight and obese.

A study on a Korean population found higher mortality with either high or low BMI.

While studies have linked being either underweight or overweight to poor health, the effect of being overweight or obese on the risk of dying has been a topic of recent controversy. Researchers have long used the body-mass index (BMI), weight in kilograms divided by the square of height in meters, as a measure of the appropriateness of weight in relation to height. Researchers from Yonsei University, in Seoul, South Korea, and the Johns Hopkins Bloomberg School of Public Health report in one of the largest studies to date (over 1.2 million study participants) that having either a high or low BMI increases risk of death. The researchers found that the effect of BMI on the risk of dying varied among major causes of death and that the risk of death from being overweight or obese was greater in younger people. The study is published in the Aug. 24, 2006, edition of the New England Journal of Medicine.

While higher BMI values was associated with less respiratory causes of death the higher BMI values were associated with higher rates of cancer and heart diease. Note that cancer and heart diseases are the two biggest killers.

The researchers found that the relationship of BMI with risk of dying varied among the major causes of death considered. The risk of death from cancer increased beginning at BMI levels of 26.0-28.0 and rose further at higher levels, according to the researchers. Risk for death from respiratory causes was highest at the lowest BMI values and decreased with higher BMI values, whereas the risk of death from atherosclerotic cardiovascular disease increased progressively with higher BMI values. Information on cardiovascular risk factors showed an increasingly unfavorable profile with increasing BMI values. Study participants younger than 50 years of age had the highest relative risk of death associated with a high BMI. The researchers report no evidence of an increased risk of death for 65-year-old and above, obese individuals.

Meir Stampfer at Harvard says this latest results are so strong we shouldn't continue to wonder whether being overweight increases your risk for dying.

Dr. Meir Stampfer, chairman of the epidemiology department at the Harvard School of Public Health who was not involved in either study, called both articles fine.

“They show quite convincingly, yet again, that overweight and, in particular, obesity, raise the risk of mortality,’’ Dr. Stampfer said. “It really should be the final word on this issue that’s arisen as to whether overweight is actually bad for you or not.”

For overweight people losing weight and keeping it off is hard. Drugs to suppress appetite will increase life expectancy.

Share |      Randall Parker, 2006 August 28 07:48 PM  Aging Studies

Alejandro Dubrovsky said at August 29, 2006 4:32 AM:

The Korean study seems to indicate that being overweight or obese is NOT a big problem. "Study participants younger than 50 years of age had the highest relative risk of death associated with a high BMI. The researchers report no evidence of an increased risk of death for 65-year-old and above, obese individuals." Very few people die younger than 50 compared to over 65s, so you'd be optimising the margins.

Also, your last comment, "Drugs to suppress appetite will increase life expectancy." does not follow even if the conclusions from the first study are correct. Being overweight might kill you. This does not mean that losing the weight will save you. There is a bit of evidence it might do the opposite. eg

Randall Parker said at August 29, 2006 4:51 PM:


But you are missing the point the researchers of the first study emphasise: Lower weight in the 65 and older starts getting caused by underlying diseases. The whole point of starting with 50 year olds and then tracking their life expectancies is that few of them have diseases that are causing them to lose weight that will eventually kill them. With 65 and up the fraction that is dying every year is much higher and the fraction with undiagnosed fatal diseaess that are causing weight loss is much higher.

These latest studies are seen by very sharp epidemiologists (and Meir Stampfer at Harvard is a very sharp epidemiologist) as addressing the underlying illness bias that probably caused misleading interpretations of the CDC study.

rsilvetz said at August 29, 2006 8:01 PM:

This is why I hate Harvard. Talk about debating angels on the head of pins.

Fat cells and macrophages share a common progenitor and thus fat cells actually make all the cytokines that macrophages do. The cytokine aspect multiplies inflammatory state and makes you sicker in terms of chronic inflammatory status which leads directly to heart disease/cancer. Fat can also be thought of as a benign tumor and shares every characteristic of benign tumors. This aspect is what makes fat an intractable problem. Here's a prediction for the FuturePundit: The exact same therapies that will end cancer/heart disease will end fat proliferation.

The true common conceptual denominators of chronic ageing disease are inflammation and glycation. Since BMI doesn't measure these except from a poor surrogate, fat, it gives essentially contradictory results. The contradictory results arise from the fact that humans in a non-inflammatory state and/or good glycation control, will not present with disease until late in life when the mitochondria fail from uncontrolled radical oxidation. The contrapositive is also true, inflammatory state humans will be sicker the fatter they are, since fat is an amplifier of the cytokine signal.

So, measure something non-essential (BMI), ignore the underlying actual state leading to disease (inflammatory/glycation status), pick the wrong measure since you should be looking at a r^3 measure or a surface area (taller people are always healthier by BMI), and get a result that is only valid for the inflammatory U.S. diet and not at all valid for the non-inflammatory South Korean diet. Forget that if you resampled the 1.7 million humans your correlations would goto hell, which should make Stampfer think twice about the meaning of fatness.

You want to do this right -- get a r^3-style measure, get a fat-percent, multiply by an inflammatory marker (amplification), stratify results by who dies, and voila! a real predictor. But no, let's use the BMI that has been discarded twice in the last 25 years...

In other words, if you are fat and inflamed and glycating you will die faster and sicker than the fat and non-inflammed and non-glycating. So, bottom line, get yourself a C-reactive protein and homocysteine and TNF/s-TNF-r levels and figure out if you are in a high-inflammatory state and act accordingly.

Lono said at August 31, 2006 10:18 AM:


That sounds reasonable - however is there really and difference in how I should act if I am Fat and inflammatory vs. Fat and non-inflammed?

Isn't fat reduction the same presription in both cases?

Also - I use a scale with a fat % meter - you are saying that is the better alternative to BMI measurements are you not?

rsilvetz said at September 1, 2006 12:55 PM:


Sure you should act differently. Fat and inflammed needs to radically alter diet to enter a non-inflammatory state (and a couple drugs would help). In fat and non-inflammed all you have to do is eat less and excersize!

Fat reduction is also a misnomer. Most fat people make their fat cells smaller but don't really drop the fat cell number. Which in turn means the blood receptor levels are almost unchanged. So it is of paramount importance to reach a non-inflammatory state more than just dropping fat.

Randall Parker said at September 1, 2006 6:15 PM:

Robert Silvetz,

Any specific dietary advice for reducing the level of inflammation in a body? I'm thinking omega 3 fatty acids, vegetables, fruits. I wonder if there are more specific things to eat. Are flavonoids particularly good at reducing inflammation? Or other compounds?

John Anderson said at September 1, 2006 8:12 PM:

Sorry, but I remain more skeptical of this than the earlier study. BMI is a lousy indicator (Marilyn Monroe was seriously underweight?!?) to begin with, there are others just as easy that are better. Overweight is admittedly bad but the BMI "standard" not only has problems but was also set too low to begin with, possibly by some ten pounds.

Now, I am some fifty pounds overweight, so this quibbling it may not matter that much to me. But I remember the "Food Pyramid", the salt/hypertension link (valid, but never mentioned was that it was valid for West African descent - not everybody), dentists advising us to only brush straight up-and-down, Alar, and some others: this study looks like it came to a pre-prepared conclusion.

rsilvetz said at September 1, 2006 11:04 PM:

Hi Randall,

Bottom line -- for the inflamed you have to go vegeterian. Heavy on the leafy greens. Fishes or Omega3 supplementation. Fruits, one has to avoid the damnable glycemic swings, so grapes, bananas, oranges are out. Apples are reasonably safe. Kill all the simple sugars. Whole wheat or Ezekiel-style bread are ok, in low quantity. No standard dairy cheeses. Go goat milk and derivatives if you need the diary-like. No beef or lamb unless it's old-style ranch grown organic. Fat composition of said meat is radically different than the average cut at the supermarket. And even so, not more than every 6 weeks or so. That's how long it usually takes for the enzyme systems to stand down. Basically, really avoid the meats.

If one is really in a inflammatory storm, I wouldn't hesitate to take half-a-Tylenol in the morn and afternoon, and if my TNF-alpha score in the upper third, and diet wasn't doing it, ultra-low-dose pentoxyfilline. Last but not least, curcumin, that started this whole thread -- e.g. indian curry dishes are really good.

Diabetics should be on a carnosine supplement to get in the way of glycation. I'm still not sure that it works, carnosinases in the blood should make quick work of the supplement, but PubMed says there's a benefit.

jessica said at October 16, 2006 7:54 PM:

you didnt help me at all thanks a lot
i wanna no the life expencancy of an obese person without the mumbojumbo!!!

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©