Hacker News Reader

Comment by gizmo | Hacker News Reader

Why do people keep making this argument? It's a terrible argument.

BMI and body fat are extremely strongly correlated. People with a very high BMI are all obese or morbidly obese. And people with a very low BMI are almost all underweight. BMI is a very simple, easy to calculate and useful metric. It works across countries and across time.

BMI is just one metric. It doesn't tell you everything and knowing only the BMI of an individual isn't proof of anything. But average BMI is higher in Texas than Vietnam despite Texas having way more more bodybuilders. People in Texas have higher BMI because people in Texas are fatter. That's the truth. BMI denialism confounds me.

View on HN

> People with a very high BMI are all obese or morbidly obese.

That is tautological - obesity is defined in terms of BMI. BMI over 30 is obese.

Arnold Schwarzenegger is 6'2" and weighed 235lbs at contest (30.2 BMI) and 260lbs in the off season (33.4 BMI). So much for your "all" claim.

> Arnold Schwarzenegger is 6'2" and weighed 235lbs at contest (30.2 BMI) and 260lbs in the off season (33.4 BMI). So much for your "all" claim.

Most people aren't Arnold. Outliers don't count in studies, for very good reasons.

It is understood that when we say "all" we mean "close to 99%", and for 99% of the population with a BMI over 30, they are overweight.

I wouldn't make the all claim, but you must admit that a world renowned body builder is an extreme outlier. Obviously a DEXA scan would be far more accurate, but it seems clear from the literature I've read that BMI is a pretty good guide for most people.

Arnie is also called an outlier. You don’t dismiss BMI just because. A doctor will factor that in before declaring a patient unhealthy. Otherwise, it works for most people and might be appropriate in large studies.

> Arnold Schwarzenegger is 6'2" and weighed 235lbs at contest (30.2 BMI) and 260lbs in the off season (33.4 BMI). So much for your "all" claim.

Congratulations on discovering that tail ends of distributions exist.

It doesn't invalidate anything in the comment you are responding to.

Well, thankfully people in real life have eyes and brain and can seee that someone is just a heavy pile of muscle vs donut cosplayer, i.e can spot the 0.1% of muscle mass builders from the 99.9% of fatsos.

"A metric have 0.1% outliers so it is bad metric" is a horrible take

The model disregards ridiculous outliers like world class bodybuilders in order to be simple and useful for the other 99.999% of society.

According to BMI I spent years "obese" and am now "overweight" bordering on obese.

I'm about 16-18% body fat. I'm definitely muscular, but hardly in the .0001%. If I'm wearing a shirt, you can't really tell that I lift.

Hardly world class.

Is your sense that having excessive muscle weight is somehow healthy in a way that isn’t reflected in the BMI classification?

Tom Cruise is also overweight by BMI standards.

A lot of people are pointing out Arnold as your example, but fail to mention he first won the Mr Olympia more than 50 years ago! Our understanding of exercise and nutrition has improved since those times. You no longer have to be an outlier in genetics or dedication to achieve good results.

I'm far from a genetic outlier, but I go to the gym for 3-4 60-90 minute sessions a week. My program is efficient, I don't take steroids, and my nutrition is decent (but not great). I'm 5'10" and 202 pounds (29 BMI).

I find it disingenuous that people use the top performers as an excuse. There is a middle ground where a modest investment will reap significant returns.

If maintaining the healthy function of my body is skewing my BMI as to be useless, then perhaps BMI is a poor metric.

> Arnold Schwarzenegger is 6'2" and weighed 235lbs at contest (30.2 BMI) and 260lbs in the off season (33.4 BMI).

I understand what you're saying, but I think a person that wins Mr. Olympia seven times is clearly a severe, severe, severe outlier. For a decade he was THE most well-built body on the planet. Literally ONE in 5 billion.

Obviously no measurement is "perfect" for "everyone", but BMI is a very good indicator for an extremely high percentage of the population.

For the average person that isn't doing something radical with their body for a specific goal (powerlifer, bodybuilder, olympian, etc.) BMI is a fantastic first pass indicator of how things are going general health/weight wise.

Most people are fat. BMI is accurate for 99+% of people.

Severe obesity is 40+ BMI. About 1 in 16 children and adolescents ages 2 to 19 (6.1%) have severe obesity, according to niddk.nih.gov. It's a staggeringly high percentage of the population.

Even when you take a 1 in 5 billion outlier like Arnold he still doesn't come close to crossing this threshold, despite his high calorie body-builder diet.

Oh yeah that probably explains the US’ obesity epidemic. It’s in fact a body builder epidemic. Problem solved!

I thought this was well known. I’ve not known a gym that takes BMI seriously. People like Messi who are relatively lean and fit are overweight per the BMI scale.

There exist better metrics as well

It’s because people take their BMI and become shocked about where they land. It takes like a small beer gut to land in to obese.

BMI isn’t generally used as a picture of overall health anyway. More often, it’s used as a measure for health issue probability. As Texas BMI raises, it can be used to calculate the increasing chances of a huge variety of health problems.

Well, not so small unless you also exercise reasonable amount and have some muscles.

> BMI denialism

You've already pointed out that it depends on what you're using it for, and that's it's not especially useful on an individual basis. Why the weird Holocaust association?

BMI isn't a great measure for a lot of reasons.

There's way too much variation in height and muscle mass. There's variation based on race as well, for example CMS has different standards for DPP eligibility if you're asian [1].

BMI was created in 1830, and has always been of limited usefulness. There are much better methods for measuring health in relation to total body fat.

It is completely reasonable to be critical of it [2] when it's used improperly (which is very common).

[1] https://www.medicare.gov/coverage/medicare-diabetes-preventi...

[2] https://www.medicalnewstoday.com/articles/255712

Reply on news.ycombinator.com