Normal BMI with a big belly 'deadlier than obesity'

Tuesday November 10 2015

"Slim adults with a 'spare tyre' of fat around their stomach have a twice as high mortality risk than those who are overweight," The Daily Telegraph reports.

A major new study tracked more than 15,000 adults to look at the effect of body size on mortality.

Researchers looked at two types of measurement:

  • body mass index (BMI) – which provides an assessment of overall body weight
  • waist-to-hip ratio (WHR) – which is calculated by dividing the circumference of the waist by the circumference of the hips; this can provide an assessment of abdominal fat (belly fat)

Researchers found that people with a normal BMI but a large WHR had increased risk of dying during follow-up compared to people with a smaller WHR. This included people of similar BMI, and also people who were overweight or obese, but with a smaller WHR. 

The risk increase was higher for men than for women. Men with normal BMI but a large WHR were roughly twice as likely to die within 5 or 10 years as other men.

One hypothesis is that having a big belly increases the amount of fat inside the abdomen (visceral fat). This may then cause inflammation to the vital organs stored inside the abdomen, which possibly makes people vulnerable to chronic diseases.

However, despite the large sample size, only a minority of people came into this high-risk category of normal BMI, but high WHR – 11% of men and 3% of women. Analyses based on small numbers have an increased chance of producing inaccurate risk estimates.   

Where did the story come from?

The study was carried out by researchers from the Mayo Clinic in the US and the University of Ottawa in Canada, and was funded by the National Institutes of Health, the American Heart Association, European Regional Development Fund and Czech Ministry of Health. The study was published in the peer-reviewed journal Annals of Internal Medicine.

The media accurately reported the results and quotes from the press release. However, we suspect that many of the reporters did not actually read the study. Most did not go into detail about the differing risks linked to different levels of weight and obesity, especially for women, or make it clear that this type of study cannot show whether central obesity directly causes early death.

What kind of research was this?

This was an observational study using survey data collected as part of a big, ongoing study in the US, which aimed to look at the relationship between central obesity in people with normal BMI and survival. Both BMI and central obesity – having a high WHR – have previously been associated with overall and cardiovascular mortality.

However, greater emphasis is often placed on using a person's BMI as an indicator of their health, overweight and obesity, rather than distribution of body fat. This study aimed to see whether central obesity carries a risk even in people with a BMI judged to be within healthy limits (18.5 to 24.9).

A study design such as this can find trends and links between different factors, but cannot prove that one thing directly causes another. Other factors could be having an influence.

What did the research involve?

Researchers took information about people's measurements and used them to see how BMI and WHR were linked to their chances of dying during study follow-up. They adjusted the figures to take account of age, sex, education level and smoking.

The information on body measures came from the US's third National Health and Nutrition Examination Survey, carried out from 1988 to 1994. The researchers were unable to use more recent figures, because the survey stopped measuring hip size, which is important for calculating the WHR.

Researchers looked at the National Death Index to identify any participants who had died from any cause up to the end of December 2006 (an average of 14 years follow-up per person). 

They then looked at the chances of having died within particular time scales (5 to 10 years), for people who had different combinations of BMI (normal, overweight or obese BMI) and WHR (either normal or showing what they described as "central obesity").

The researchers tested their results to see whether they were explained by other factors, such as whether people had diabetes. They analysed the figures separately for men and women, because they found that the effect of BMI and WHR differed between the sexes.

Finally, they calculated the relative chances of dying, based on these results, for people who were normal, overweight or obese in terms of BMI, with and without central obesity.

What were the basic results?

People with a normal BMI, but a high WHR (with central obesity) were more likely to have died during follow-up than people who had a similar BMI, but who did not have central obesity. More surprisingly, men with normal BMI but high WHR were more likely to have died than women who were overweight and obese, and also had a high WHR.

A man of normal weight with a high WHR was 87% more likely to die than a man of comparative BMI, but no central obesity (hazard ratio (HR) 1.87, 95% confidence interval (CI) 1.53 to 2.29). Surprisingly "he" was twice as likely to have died compared to a man who was overweight or obese by BMI, but had no central obesity (HR 2.24, CI 1.52 to 3.32).

At age 50, a man with a normal BMI and normal WHR had a 5.7% chance of dying within the next 10 years, but that rose to 10.3% chance for men with normal BMI, but a high WHR.

For women, the results were less striking. A woman of normal BMI but high WHR had an almost 50% increased risk of death compared to a woman of similar BMI without central obesity (HR 1.48, 95% CI 1.35 to 1.62), and a 33% increased risk compared to a woman with obese BMI, but no central obesity (HR 1.32, 95% CI 1.15 to 1.51).

A woman aged 50 of normal weight and normal WHR had a 3.3% chance of dying within 10 years, rising to 4.8% for women of the same weight, but a high WHR.

Men with normal BMI but a high WHR were more likely to have died than any other combination, including men who were obese and had a high WHR. 

The picture was more mixed for women. Women who were overweight or obese with a high WHR had about the same chance of death as women with a normal BMI but a high WHR.

How did the researchers interpret the results?

The researchers said: "Our findings suggest that persons with normal-weight central obesity ['belly fat'] may represent an important target population for lifestyle modification." 

They say we need more research on how central obesity develops in people with a normal BMI, and a better understanding of the effect of central obesity on health. They call for measures of central obesity to be used alongside BMI to calculate people's risk.


This study adds to previous research that it may not be just your weight that matters, but where you carry it. It found that – especially for men – those with a high WHR had a greater chance of dying from any cause during study follow-up than those without. The results were less strong for women.

A high WHR suggests excess fat around the waist, as muscle mass is unlikely to lead to greater waist circumference. Although this study does not explore why WHR may be linked to chances of dying earlier, other studies have shown that carrying excess fat around your waist may be more harmful than carrying it in other areas, such as your legs and hips. Fat around the waist has been linked to inflammation, increased risk of diabetes, and having raised cholesterol.

The study's strengths are its size and the fact that data was collected consistently over a long follow-up period. However, there are important limitations. One of these is that, despite the large overall sample size, the main study analyses for people with normal BMI but central obesity were based on a small number of people. Only 322 men (11.0%) and 105 women (3.3%) were in this risk group.

Analyses based on smaller numbers of people have an increased chance of producing inaccurate risk estimates. Therefore, we do not know that the risk figures obtained here – for example, the 50% risk increase – are completely accurate and would apply to all people in this category.

Also, as the researchers say, diseases such as diabetes and high blood pressure were self-reported by the participants. This means that some disease classifications may be inaccurate, and, overall, the full effect of all health and lifestyle factors that could be confounding the results have not been fully accounted for. Body fat measurements were also taken by hand rather than by imaging as recommended, so may be less reliable.

The main point to remember is that we don't know that high WHR directly led to a higher chance of dying, or know why this link may have been found – particularly for people with a normal BMI more than people with an overweight or obese BMI. We need to see more research to understand the results better.

Nevertheless, the study adds to the evidence on the importance of central abdominal fat as a health indicator. 

It shouldn't be taken that it is safe to be obese as long as your WHR is on the smaller side. While it could be the case that some types of excess fat are worse than others, all excess fat is bad for you.

Analysis by Bazian
Edited by NHS Choices