Is it Better to be Overweight?
This is one study that everyone should be aware of, and this is especially true if you work in the fitness or health field. The study is a collection of studies and looked at the association between BMI and deaths (or mortality).
What is BMI or Body Mass Index?
BMI or Body Mass Index is commonly used as a measure of your body fat. It is calculated by having your weight divided by height square (kg/m-square). You can calculate your BMI using this calculator. Once you find your BMI, you can see from the chart below if you are normal weight, overweight or obese. And this is true no matter which country you live in. So what is your BMI?
What was the study?
The study published in 2013 in JAMA was done to see how BMI is associated or related to death (or mortality). Or to answer the question who lives the longest? Normal weight, overweight or obese people?
What is unique about this study is that that for the first time they looked at more than 7000 studies all over the globe. After this comprehensive analysis, 97 studies were included in the study which looked at more than 2.88 million individuals and 270,000 deaths. This is by far the largest meta-analysis done on this topic.
What were the results?
Now here is the fun part: If I plot the BMI ( 18-40) on the x axis and mortality on the y-axis, how would the curve look like? It should be an linear relationship suggesting as BMI goes up, mortality goes up, right?
WRONG.
As you can see from the graph below,
- Compared to normal weight, overweight has a 6% decrease in mortality. How is that for a surprise?
- The mortality rate for obese was similar to normal weight
- The mortality rate for obese (>35 BMI) was much higher than normal as expected.
- The mortality rate for underweight (<18.5) was much higher than normal as expected.
- The results are strikingly consistent across all ages, gender and population
Now what were the criticisms?
As expected, the article garnered a lot of opinion and criticism from both researchers and the press. For me the worst part was none of the experts were specifically responding to the criticisms; they were just beating around the bush.
1.The loudest and harshest of all of them were Dr. Walter Willett, a leading nutrition and epidemiology researcher and the chair of Nutrition department at the Harvard school. He wrote “This study is really a pile of rubbish, and no one should waste their time reading it”. Specifically, he wrote that the study is flawed because the normal weight group contains persons who are lean and active, heavy smokers, frail and elderly, and seriously ill with weight loss due to their disease so there were more deaths in the normal weight.
- My Comment: His criticism was picked by a lot of bloggers and even researchers and just ran with it. If people take 10 minutes to read the study than just blindly following what other prominent people say, this criticism would have never came up. From the original article, “The results presented herein provide little support for the suggestion that smoking and preexisting illness are important causes of bias. Most studies that addressed the issue found that adjustments or exclusions for these factors had little or no effect.” Basically, it means almost all the studies included in the analysis were adjusted for smoking, age related weight loss, smoking, and illness related weight loss..
- My Comment: This is not one study, but a collection of multiple studies (called meta-analysis)These results are only controversial, if you are unaware of the literature. These results are consistent with previous studies. For example, the same author published a meta-analysis in 2005 showing no increase in mortality in overweight people (2). This was repeatedly shown in other meta-analysis done by other authors too (3,4) . It is now pretty much established that in the presence of chronic conditions like heart disease or diabetes or older age, being overweight or slightly obese can be protective or the excess body weight makes them live longer. We even have a name for it called the ‘obesity paradox’.
- My comment: This is a valid point but is not a criticism of the study per se. The study aim was to look at mortality and not morbidity. It is true that as BMI increases there is an increase in disability, medication, health care costs, and unhealthy life years. And they do mention it in the study as a limitation.
- My Comment: The author is Katherine Flegal and works for the health statistics department at CDC (Center for Disease Control & Prevention). These people have the best bio statisticians in the country.What most people are unaware is that, she is the very same author who publishes all the ‘good’ news about the obesity prevalence in America. If there is any researcher in this world, who wants to see mortality goes up with BMI, I reckon it might be here since she is the one who showed/showing us how America is getting fatter.
On a personal note, she is one of those fantastic researchers that everyone can look upto. Even when she was pummelled with personal criticisms (and this dates back to 2005), she has responded very professionally everytime and left out the personal attacks and remarks that could have easily been justified. He letter responding to article comments in JAMA is just a beautiful example of how to respond to criticisms. I personally wrote a thank you letter which she gracefully replied and forwarded to her co-authors. In fact, the prestigious science journalNature wrote an extraordinary editorial literally scolding Dr.Walter Willet for taking his criticisms too far against fellow researchers who don’t agree with his conclusions.
What could be the explanations?
Now what could explain why being a little fat helps people live longer? Keep in mind that study shows only an association and the decrease in mortality may not have much to do with the extra weight. Here are some of the major reasons quoted (7):
- Deaths from obesity or being fat is mainly due to heart disease. But now with all the new prescription drugs available for blood pressure and cholesterol, there is a dramatic decrease in heart disease deaths. It has been shown that the levels of risk factors are now way lower than that of lean individuals 20-30 years ago. Further, it is been known that obese and overweight people are more likely to visit doctors, be screened for disease and be treated compared to lean folks.
- BMI is not an accurate indicator of body fat or the location of body fat. People with normal BMI can also have body fat percentage consistent with obese folks. Likewise, folks with higher BMI can have body fat levels comparable to a normal BMI individual. It is also now known that the location of body fat plays an important role in predicting health risks; abdominal fat posing the greater health risks This is why waist circumference is usually recommended as an additional marker among with BMI for health risks associated with excess body fat. BMI is also not an accurate indicator of the presence or absence of risk factors like cholesterol, blood pressure. and diabetes (8).
- Cardio respiratory fitness has now been shown to be an independent predictor of mortality. Studies have shown mortality risk of people who are fat, but fit to be twice lower than lean, but unfit folks. So you can be fat and still be fit and healthy, and live longer. This is the popular fit vs fat concept by Dr. Steven blair.
- As mentioned previously, presence of chronic conditions like heart disease or diabetes or older age, being overweight or slightly obese can be protective. This could be due toThe optimal BMI for people with chronic disease could be within overweight and obese level (5). As shown in this study, this maybe true for the population without any chronic disease, but needs to be further investigated.
Now what is this 6% decrease in mortality mean?
This study showed a hazard ratio of .96 for the overweight group or a 6% decrease in mortality in the overweight group. But what the heck does this 6% mean to you? This is one important question that almost all articles reporting this study have ignored or missed.
One way to communicate risk is in the form of life-expectancy. For example, a hazard ratio of .96 for a 35 year old means he may add 6 months to his life-expectancy by staying overweight or he may gain 15 min extra everyday. If you are obese (>35), you lose an hour and half everyday. So even if it shows a decrease in mortality with being overweight, it is just very a small effect. This is is the concept of microlives (1 microlife = 30 minutes of your life expectancy) developed by David Spiegelhalter (a very good biostatistician) to communicate health risks to the public (6).
Practical Recommendations
- BMI is just one risk factor and a minor one in the assessment of health risk factors. BMI should not be used as a sole determinant for health risk assessment. If you are overweight or obese (less than 35) you need to look at other risk factors such as waist circumference, cardiovascular fitness, blood profile (cholesterol, blood pressure,..) before you start panicking about your weight.
- Although, there is no increase in mortality with being overweight or obese (<35) BMI, there is an increase in morbidity or sickness with BMI ( decline in function, disease,long-term medication, medical costs and so forth). So your quality of life or health is better if you can stay around normal weight.
- As Dr. Flegal, the author of the study, puts it: We are absolutely not recommending people overeat. We intended our research to give a little perspective – to counter the view that if you weigh a bit less you will live forever or if you weigh more you are doomed. The relationship between fat and mortality is more complicated than we tend to think.