Fat people die sooner? Not according to their data!

back to healthread                                Obesity in the studies              back to Obesity page
We have by now, all heard about the latest study which is being shouted from all media sources.  The conclusion.... NOW WE HAVE PROOF that fat people die sooner.  The study was a joint effort of the AARP/ NIH.
 
The problem is, NONE of the data, except one small section of people at the age of 50, show any SIGNIFICANTLY greater relative risk factor even in people with a BMI over 40.  Remember that the differential in relative risk (RR) must be greater than 2.0 or it is considered insignificant.
 
So in all of the data except that small section, the relative risk factor was NOT significant:
 

BMI range <18.5 18.5–20.9 21.0–23.4 23.5–24.9 25.0–26.4 26.5–27.9 28.0–29.9 30.0–34.9 35.0–39.9 ≥40.0

No. of deaths               333 1513 5229 5768 6657 6021 6355 7488 2028 781

Age-standardized rate† 3520 2430 1520 1320 1270 1310 1440 1680 2260 3210

Multivariate relative risk 1.97 1.54 1.14 1.00‡ 0.95 0.95 1.00 1.10 1.35 1.83

 

As you may notice here, in the so called normal range, the RR is only .8 lower than in the severely obese range.  You might also notice that the RR in the "overweight" range is actually slightly lower than that in the normal range and that the RR in the underweight range is slightly higher than that of people with a BMI over 40.
 
OK, let's look at the deaths at the age of 50, one tiny subsection of data which did show a higher RR in people with a BMI over 40.
 
As you can see, they determined this from a small subset - these were people who were asked to remember what they weighed at age 50 (that's real accurate, right? *LOL*).  The differential RR is 2.2 (between those in normal range and those with BMI over 40).  Not exactly high but it was the only group which showed a significant differential in all the data.

<18.5 18.5–20.9 21.0–23.4 23.5–24.9 25.0–26.4 26.5–27.9 28.0–29.9 30.0–34.9 35.0–39.9 ≥40.0

No. of deaths                  133 796 3567 3812 4113 3262 2728 2910 627 212

Age-adjusted relative risk 1.81 1.36 1.06 1.00† 1.01 1.13 1.27 1.66 2.26 3.22

So for some reason, in the conclusion, they took THIS very small sub-group to conclude that ALL fat people die sooner and IGNORED what MOST of their data showed!  Only they didn't SAY it was only a very small section of their data that showed that.  They said the study was of a half million people and wrote their conclusion to suggest that ALL the data showed that fat people died sooner.

Scientists generally do not like to reveal to the public the uncertainties that are associated with their research. It might make them look less expert. Sometimes it's just too complicated to explain all the uncertainties. Whatever the reason, claims that scientists make are usually more dramatic and confident than can be defended with the science alone." (Dr Roy Spencer, former senior scientist climatologist with NASA in his book, CLIMATE CONFUSION, NY, 2008)

One other thing we should keep in mind about this study. According to the data, NONE OF THEM were exercising enough to affect their health.  None of the people exercised more than twice a week and those in most of the overweight ranges exercised an average of slightly over once a week. This was self reported data. And self reported data is notoriously inaccurate.

On "The Today Show", their medical advisor stated that women tend to lie to their doctors about several things - among which are how much they exercise (most people overstate their amount of exercise), whether they smoke (many women understate their smoking habits) and how much they eat in a day (few actually count calories and if you don't, it's easy to fool yourself about the amount of food you are consuming).  It can be assumed that when they fill out surveys for studies, they might be LESS honest than they are with their doctors and not MORE honest! :)

Several studies have suggested, among which are the studies out of the Cooper Institute, that people who exercise, have a lower risk of illness and death (cardio - 5 times a week, 30 minutes each time) than people who do not exercise, regardless of weight.

Another factor almost never taken in consideration in these studies is the weight cycling habits of many overweight people which can increase their risks of heart disease and other illness. But considering this factor in this study, those with a very high BMI only showed a slightly higher RR in the small subsection at the age of 50 and NO significantly higher relative risk in MOST of the data.

That suggests that being fat might actually have some health benefits which are lowering RR even though so many of these people are weight cycling....

It is important to read these studies because if we want to be healthy, dieting or calorie cutting or "losing weight" may not be the way to best achieve health.  Getting on a healthy lifestyle of exercise and good food choices will make us healthier at ANY weight and that is what clinical studies have suggested.  For example, see the HAES study, USC.   One of the researchers on this study, Dr Linda Bacon, wrote a book which is very informative.

As for these new studies which appeared in the NEJ this month, Milloy said it all:

"But who needs data when you can spoon-feed junk science to a gullible media?" 
    - Steven Milloy, Fox News


SueW

source: http://content.nejm.org/cgi/content/short/355/8/763