Why I Stopped Trusting BMI (and What I Use Instead)

I want to tell you about the day I lost faith in BMI. It was, I think, October 2019, in a small gym off Colfax Avenue where I was doing intake measurements for new clients. A guy named Jorge (not his real name, but close enough) sat across from me. Former college rugby player, stocky, thighs like fire hydrants, about 14% body fat if I had to eyeball it. The number on my clipboard said his BMI was 31.2.

Thirty-one-point-two. That's "obese, class I." According to the chart his doctor used, Jorge was in the same risk category as someone who genuinely needs to lose 40 pounds of fat for their heart's sake. Jorge's problem was the opposite — he was trying to cut because he wanted to run a half-marathon and all his muscle was slowing him down.

I remember writing the number down and thinking, "this is stupid." Not in a dramatic, throw-my-clipboard way. More of a quiet, long-overdue realization. I had been using this metric on intake forms for seven years. I had told actual human beings, who had paid me actual money, that their body was in a category it did not belong in.

A very short history of a very weird number

Here's the thing nobody tells you in your cert course: BMI wasn't invented by a doctor. It was invented by a Belgian guy named Adolphe Quetelet in the 1830s, and he wasn't trying to measure health. He was doing population statistics. He wanted a quick way to describe the "average man" across groups — something a government census taker could do with a tape measure and a scale.

It's weight in kilograms divided by height in meters squared. That's it. No adjustment for muscle. No adjustment for bone density. No adjustment for where you carry fat. No adjustment for sex, age, ethnicity, or whether you spent the last four months training for a powerlifting meet.

The formula got picked up by life insurance companies in the 1940s because it was cheap, and then by the World Health Organization in the 1990s because — I'm not kidding — they needed something simple enough to standardize across 190 countries. Simplicity isn't the same thing as accuracy. I think we all forgot that somewhere along the way.

Where BMI actually breaks

I've been keeping a mental list for years. Here are the people it fails most consistently, based on my own client roster, not a meta-analysis:

Lifters and athletes

Jorge wasn't unusual. I've had three clients flagged as "overweight" or "obese" by BMI who had body-fat percentages under 15%. Two of them were women, which matters because women face a much bigger social penalty for appearing on the wrong side of that line. One of them — a CrossFit competitor — told me her gynecologist lectured her about weight loss. She could deadlift 315. Her gynecologist could not.

Older clients who've lost muscle

Flip side: I had a woman, probably 68 at the time, BMI of 23 — a number that in the little green zone on every chart. She couldn't stand up from a chair without using her hands. Her body fat was 38%. This is what researchers call "sarcopenic obesity," and BMI is completely blind to it. Her "normal weight" number was hiding a body composition that put her at real fall risk.

South Asian and East Asian clients

The research here is pretty damning and has been for a while. South Asian populations tend to develop metabolic disease (type 2 diabetes, fatty liver) at much lower BMIs than the thresholds assume. India's own health ministry has been using adjusted cutoffs (23 for overweight, 25 for obese) since 2009. The US and most of Europe still don't. If you're reading this and your doctor is still using standard cutoffs on a South Asian patient, please — push back.

Pregnant and postpartum women

Do I even need to explain this one? And yet. I've seen intake forms ask for BMI on postpartum clients six weeks after delivery. Come on.

The moment I actually stopped using it

After Jorge, I spent maybe a month in denial. I kept the BMI box on my form but I stopped mentioning the result out loud. I'd write it down, flip the page, and move on to something useful. Eventually my business partner — Ren, who owns the studio with me — said, "Why are we even collecting this?" And I didn't have a good answer. So we deleted the field.

The only reason I think BMI still belongs anywhere near a health conversation is as a rough screening tool at the population level. If you're a WHO researcher trying to compare obesity trends between Japan and Mexico over 30 years, sure, BMI is fine. It's cheap, it's standardized, it's good enough for a graph. For you, sitting in your kitchen, looking at a number your doctor wrote down? It tells you almost nothing.

(Honestly, I still let clients calculate it on our BMI calculator because they ask for it, and giving people a number they asked for is better than giving them a lecture about why they shouldn't have asked. But I follow it up with the next three numbers, which actually matter.)

What I measure instead

1. Body fat percentage (done honestly)

This is the big one. A 180-pound man at 10% body fat and a 180-pound man at 28% body fat are completely different bodies with completely different risks. I use calipers (Jackson-Pollock 7-site if I care about precision, 3-site if I'm in a hurry) and cross-check against a bioimpedance scale.

At home, most people can get a reasonable estimate from a body fat calculator that uses Navy tape measurements — neck, waist, hips. It's not DEXA-scan accurate, but it's accurate enough to watch trends, which is all you really need. Watch the trend, not the absolute number. A client going from 24% to 21% over three months is winning, even if their BMI didn't move.

2. Waist-to-height ratio

If I could only pick one metric and throw the rest away, this might be it. Your waist measurement divided by your height. The old rule of thumb is "keep your waist under half your height," and the research that's come out in the last decade basically supports it. It's correlated with visceral fat — the stuff around your organs, which is the bad stuff — in a way BMI just isn't.

You need a tape measure and 30 seconds. No scale, no math, no judgment about whether your bones are "big." I keep a soft tape in my gym bag and I use it more often than I use my scale.

3. Functional benchmarks

Can you do ten push-ups? Can you carry your own bodyweight in groceries up a flight of stairs without your heart doing something weird? Can you sit on the floor and stand back up without using your hands? (That last one — the "sit-rise test" — has some surprisingly good mortality-prediction data behind it.)

I'm half-joking about the groceries, but only half. What "healthy" looks like for a 42-year-old dad of two is not "BMI 22." It's "can pick his kid up off the ground ten years from now without throwing his back out." BMI can't measure that. Neither can a smart scale.

4. Bloodwork and resting heart rate

I'm not a doctor and I'm careful about this one. But I'll tell clients: your fasting glucose, your A1C, your lipid panel, and your resting heart rate will tell you more about your cardiovascular risk than your BMI ever will. If your doctor only mentions BMI and skips the bloodwork, ask for the bloodwork. I don't know why I have to say this in 2025, but there you go.

What about the "ideal weight" charts?

I get this question a lot, and I'll be honest: I'm conflicted about ideal weight calculators too. The classic Devine and Robinson formulas were designed for medication dosing, not aesthetics or health, and they can produce numbers that are wildly off for tall people, athletes, and basically anyone not built like a 1960s hospital patient.

I use them as a ballpark — a "are we in the neighborhood of reasonable" check — but never as a target. If a client tells me their "ideal weight" from some chart is 135 pounds, and they're currently 165 with 22% body fat and they squat their own bodyweight, I tell them the chart is lying. Sometimes they believe me. Sometimes they don't, and they go find another trainer who'll help them starve down to 135 and then ask me six months later why their hair is falling out.

Why the medical establishment still uses it

Short answer: inertia, insurance codes, and the fact that a nurse can calculate BMI in eight seconds. I don't want to be too cynical here — there are real reasons a clinic needs a fast screening tool. You can't run a DEXA scan on every person who walks in with a sore throat.

But the problem isn't that BMI exists. The problem is that it's become the answer instead of the first question. I've had clients denied knee replacements because their BMI was "too high." One of them was a former firefighter with a BMI of 33 and a body fat around 18%. His knee was destroyed from a fall in 2017. The surgeon wouldn't touch him until he "lost weight," which in his case would have meant losing muscle. He waited two years and traveled out of state for the surgery. Two years of limping.

That's the kind of thing that makes me crank-y on this topic. (I know "cranky" is how you spell it. I'm leaving it.)

What I'd tell you if you were sitting across from me

I'd tell you this. Calculate your BMI if you want — it's a free number, it takes ten seconds, and sometimes doctors ask. Then put it in the filing cabinet at the back of your brain and mostly forget about it.

Measure your waist. Measure your height. Divide. If that ratio is under 0.5, and you can do basic functional stuff without wheezing, and your bloodwork is clean, you are in fine shape regardless of what the green-yellow-red chart says.

And if a doctor ever tries to make a serious medical decision for you based on your BMI alone — not in combination with bloodwork, not in combination with an actual look at you as a human being — ask them why. Politely. I've seen plenty of good doctors update their thinking when patients push. I've also seen plenty who haven't. You know your body better than a 190-year-old formula does.

I don't know everything about nutrition or metabolism. I'm not going to pretend I do — there are still mornings I watch a client's body respond to a program in a way I can't fully explain. But one thing I'm sure of, after twelve years on the gym floor: the number of times BMI has actually helped someone I was training is approximately zero. The number of times it's gotten in the way is a lot more than zero.

That's the honest version. Take it for what it's worth.