According to this study, overweight people actually live longer than people with normal/below normal weight. Is it really true? What about Okinawa people for example? What about underweight people because of their high metabolism? They are slim and they live very long. I'm confused a little. So what is better for a longer life? Is there any causation between BMI and life expectancy?

  • 4
    "Because the study only examined death risk, and not disease incidence or quality of life, the risk vs. benefit profile of carrying a few extra pounds is unclear," Nov 19, 2011 at 21:52
  • 1
    The article doesn't have a reference to the study. In fact it's the second part of an article. Nov 19, 2011 at 22:00
  • 4
    Just to be clear, the study did NOT seem to imply causation in any way.
    – user5341
    Nov 19, 2011 at 22:05
  • @DVK, I agree with you. The linked study is a little unfortunate. But the point is I want to know if there is any causation between BMI and life expectancy. It is written almost everywhere that a "normal weight" is between 18.5 - 25 BMI and I'm not sure this is the best weight for longevity.
    – Lucasus
    Nov 21, 2011 at 14:55
  • I think you missed a big part of the article, overweight people live longer, but not obese people.
    – Sam I Am
    Nov 21, 2011 at 17:05

3 Answers 3


It rather depends on what you mean by normal weight. The standard definition is to use Body Mass Index (weight in kg divided by square of height in metres) and describe normal as something like 20-25, overweight as 25-30 and obese as 30+. There are issues about this simple rule of thumb, since it ignores body type, and there seems to be no reason to believe that the power of two in the denominator is the correct number (it certainly is not for children and may not be for tall people). But ignoring those questions and just looking at epidemiological evidence, this study from the Journal of the American Medical Association says

Underweight and obesity, particularly higher levels of obesity, were associated with increased mortality relative to the normal weight category. The impact of obesity on mortality may have decreased over time, perhaps because of improvements in public health and medical care

while this meta-study from the Lancet says

Patients with a low body-mass index had an increased relative risk for total mortality and cardiovascular mortality, overweight had the lowest risk for total mortality and cardiovascular mortality compared with those for people with a normal BMI. Obese patients had no increased risk for total mortality or cardiovascular mortality. Patients with severe obesity did not have increased total mortality but they had the highest risk for cardiovascular mortality.

suggesting that being moderately overweight is not in itself unhealthy, but being underweight may be.

  • 6
    BMI is actually a grossly incorrect metric (npr.org/templates/story/story.php?storyId=106268439). The guy who invented it said he was only using it to study populations, and should never be applied to individual people (reason #1 in the article there). Shaq would be obese ("At 7-foot-1 and 325 pounds, O’Neal had the NBA’s highest BMI, 31.6, in the AP analysis." nbcsports.msnbc.com/id/7129586), when I'm pretty sure that he's not the target for the label 'obese.'
    – mmr
    Nov 20, 2011 at 3:11
  • 3
    @mmr: Given Henry gave a similar disclaimer, and then used it as a coarse measure to study populations, are you saying he used it appropriately?
    – Oddthinking
    Nov 20, 2011 at 3:22
  • 5
    @Oddthinking, I'm saying that using BMI for anything is bad. The entire concept fails to capture what it sets out to do. There are other metrics (like Waist To Hip ratio), but there's still some discussion about whether or not there are strong correlative links there to obesity (msnbc.msn.com/id/14483512/ns/health-fitness/t/…). He is right that defining 'normal' weight is tricky, and as such, I'm leery of the rest of the answer if they all rely on BMI as their metric for obesity.
    – mmr
    Nov 20, 2011 at 3:36
  • 2
    I fully understand that formula for BMI is not a great proxy. But it is a very cheap proxy - measuring height and weight is cheap and inobstrusive, so I understand why they use it.
    – Oddthinking
    Nov 20, 2011 at 3:46
  • 2
    Measuring waist and hip is also cheap (just requires a measuring tape, and presumably some devices were used to get height and weight). The trick is to find a way to accurately determine health boundaries caused by weight changes (ie, if this number is above or below this threshold, badness is likely to result). BMI isn't it, and I don't know what is. Until that question is answered, in my mind, it's difficult to go further in making any scientific, quantifiable statements about weight.
    – mmr
    Nov 20, 2011 at 3:54

Moderately overweight people have significantly lower mortality than those with "normal" BMI's; the really thin and the really fat do worse

A recent meta-analysis in the Journal of the American Medical Association (JAMA) concluded that the moderately fat by current standards (BMI 25-30) had significantly lower mortality than people in other standard weight categories.

The meta-analysis was very careful to exclude the normal confounding factors (e.g. ill people may well have low BMI's because of their illness).

The key conclusions were:

Relative to normal weight, both obesity (all grades) and grades 2 and 3 obesity were associated with significantly higher all-cause mortality. Grade 1 obesity overall was not associated with higher mortality, and overweight was associated with significantly lower all-cause mortality.

The key numbers from the study are the definitions of the weight categories and the resulting mortality risk. These are given below:

  1. Normal, BMI 18.5-25, mortality risk 1
  2. overweight, BMI 25-30, mortality risk 0.94
  3. obese grade 1, BMI 30-35, mortality 0.95
  4. obese grade 2 and 3, BMI >35, mortality 1.29

So, moderately overweight seems to (slightly) associated with lower risk of death compared to the thin, but if you get too fat the risk of (technically association with as true causality is hard to prove) death goes up a lot.

  • I know this is old, but I was about to give a +1, but then I noticed the last sentence, which suggests causality, not just association, and I got antsy.
    – Oddthinking
    Jan 24, 2017 at 22:14

What makes people think skinny or weight loss is healthy? Weight loss is a common sign of advancing disease. A Harvard review was horrified by this and similar studies suggesting obesity increases life expectancy. Here is their primary objection to this and other studies. ... included many samples of people who were chronically ill, current smokers and elderly, according to Hu. These factors are associated with weight loss and increased mortality. http://news.harvard.edu/gazette/story/2013/02/weight-and-mortality/

I also recognize being overweight may mean well-fed and a fully function immune system, but that does not equate to the known risks of diabetes, hypertension and heart disease.


  • 1
    Also need to take in to account the fact that all weight is not the same. The physically active person whose 'excess' BMI weight is mostly muscle is very different from the sedentary one whose excess is in a beer gut.
    – jamesqf
    Jan 24, 2017 at 18:43
  • I would recommend quoting a sentence or two from your references. Not only does it help protect against link rot, it helps us understand why you have cited the study. Your second reference seems to directly contradict the sentence above it it is supposed to support.
    – Oddthinking
    Jan 24, 2017 at 22:18
  • I have a direct quote in my above comment and I am not sure why you think the second quote contradicts the first. "Underweight was associated with increased risk of all-cause mortality and obesity with increased risk of CVD mortality. Therefore, maintaining a normal BMI through leading an active lifestyle and healthy dietary habits should continue to be promoted". The second study is reporting what is called selection bias. The investigators lumped thin unhealthy adults with healthy ones skewing the data to show overweight people are healthier, Incredible. This is why Harvard was outraged. Jan 24, 2017 at 22:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .