At some point, you’ll probably have worked out your BMI. Body mass index, calculated by dividing your weight in kg by your height in metres squared, is the formula used by the NHS to diagnose obesity, and by clinicians and researchers to predict a person’s risk of all manner of health conditions, from anorexia to diabetes and Covid.
But is it doing more harm than good? A large body of research now suggests it is an outdated and ineffective way to measure how your weight affects your health. In April, the Women and Equalities Committee called for BMI to be scrapped, describing it as a “dangerous” obesity strategy which triggers eating disorders, and has led to “a rise in body image anxiety”. Many experts argue that it is a blunt tool, a rough guide which is flawed by the fact it does not distinguish between fat and muscle – so while useful in epidemiological research, on an individual level it can see healthy people misdiagnosed as overweight or obese, and vice versa.
It’s now widely established that body shape, and particularly how big your waist is, is a strong predictor of health problems. A new metric to measure obesity called ABSI (a body shape index) – which takes into account age, sex, weight, height and waist circumference – is looking like a more effective tool. In May, a study by the Universities of Glasgow and Newcastle, found that measuring ABSI alongside BMI was a better predictor of people’s risk of bowel, lung and liver cancer.
I’ve long been suspicious of BMI, as someone whose weight dances a kilo either side of what it considers “normal” and “overweight”. Being in the overweight category never felt accurate to me: I’ve got a narrow-ish waist and an above average amount of muscle from vigorous exercise since my days of rowing at university.
My BMI is currently 25.1, which is classed as overweight, but according to ABSI I am in fine fettle. I type my numbers into an online calculator and score 0.067, which puts me in the healthiest category, with a “very low” risk of health complications resulting from excess body fat.
From the beginning, BMI was never intended to be a way to measure a healthy weight. It was invented by a mathematician in the mid-19th century as a way to describe the growth spurts that happen after birth and puberty.
ABSI, on the other hand, was designed from the outset as a way to predict disease risk. Its inclusion of waist circumference is important because while research suggests that fat on your bottom or thighs may be neutral or even beneficial for health, fat around your middle is more dangerous, and is strongly linked to type 2 diabetes, dementia and heart disease.
Studies have shown that ABSI is much better at predicting the risk of diseases and early death than either BMI or just waist circumference alone.
Including your waist circumference not only helps to estimate where you store fat, but also your ratio of fat to muscle. For example, if your weight remains the same, but you start lifting weights and building muscle, then your waist size is likely to drop as you lose fat from around your middle. However, your BMI would remain the same, despite you becoming healthier. “If a man has really strong muscle then they might have a high weight, so their BMI might say they are obese but actually they’re really healthy in the amount of fat [they carry] and where they carry it,” says Dr Mengmeng Ji, an obesity researcher at the University of Illinois at Urbana-Champaign.
A 2008 study by the Mayo Clinic looked at the height, weight and body fat percentages of a group of more than13,000 people, to determine how well BMI diagnosed obesity as defined by the World Health Organisation (WHO): having over 35 per cent body fat for women, or 25 per cent for men.
While 31 per cent of women and 21 per cent of men in the study were classed as obese by their BMI, more than twice that number were obese by their proportion of body fat. In other words: BMI only managed to spot half of the obese people in the study, and gave the other half a false sense of security about the state of their health. Researchers concluded that the accuracy of BMI is limited “particularly for individuals in the intermediate BMI ranges”. In short: it is very accurate for those at the high end of the spectrum, but in the middle it’s a lot patchier. “BMI is a crude measure for fat distribution, which is more related to actual obesity,” says Ji.
The study also found that the accuracy of BMI is worse as you get older – you may stay the same weight as when you were young, but with less muscle and more fat.
Another study published in 2016 compared 40,000 people’s BMIs with specific measurements of health such as insulin resistance, blood pressure, and cholesterol levels. Nearly half of those whose BMIs diagnosed them as overweight were metabolically healthy by those measures.
Critics also point out that BMI was designed and validated mainly on white men, and body compositions and corresponding thresholds for disease risk can vary according to gender and ethnicity. For example, for people with South Asian heritage, a healthy BMI is considered as between 18.5 and 23, rather than up to 25 for people with a European background.
There are problems, too, in the way that BMI is applied. The report from the Women and Equalities Committee found that health problems in overweight and obese people are often not investigated or diagnosed as well as the same complaint in a “normal” weight person, with doctors too often simply blaming problems on their size. They also found issues with how GPs use BMI to diagnose eating disorders, with some patients told that they were too heavy to get help – an illogical and dangerous thing to tell someone with anorexia or bulimia.
Ji says she expects BMI to dwindle and ABSI to get more popular, though she says that the latter is still not perfect: “If you really want to use a number to show health or excessive fat you still need to use a machine to detect fat percentage and specifically where it is,” she says.
Sir David Haslam, a GP, obesity expert, and former chair of the National Institute for Health and Care Excellence, agrees that any one index will never be enough to measure someone’s state of health. “ABSI is edging forward from just plain and simple BMI, but nevertheless I object to a number telling me how healthy my patient is when there’s so much more to consider,” he says.
He says that while a BMI of 24 might put someone in the “normal” category, he looks out for other symptoms like being pale or short of breath, which could be markers of poor metabolic health. “What you actually need to do is assess someone with your eyeballs,” he says.
“It’s the person in front of you that matters”, he says. “Patients object to just being a number.”
To calculate your ABSI, a good online calculator can be found at: fatcalc.com/absi
For a more basic reading, you can measure your waist with a tape measure. According to the British Heart Foundation, a healthy waist circumference is 31.5in or less for women or 37in or below for men, whatever your age. For men from South Asian, Chinese, Japanese or Afro-Caribbean heritage, this should be 35.4in or less. Be aware that this doesn’t take your height into account, so it won’t be as accurate if you are very tall or very short.