Claiming that BMI is flawed is a lazy half-argument. Finish the argument, what should be used instead of BMI? Look at the correlation plots and draw your own BMI line [0]. I think the BMI lines look pretty good for men, but for women who fall into the common BMI range of 25 to 30, the BMI formula tends to underestimate the amount of body fat.
It would be nice if everyone could get a DEXA scan regularly, but the data shows that 50% of people would find that they have even more body fat than the BMI charts predict. Again, it's worse for women, most women near the overweight threshold would find that their body fat is higher than expected.
This needs to become more widely known. BMI underestimates obesity, not the other way around. Studies that use better measures of body fat percentage consistently find even people who are "normal" weight are fatter than they think. It's why doctors are supposed to now use waist to height ratio as a secondary predictor.
I think due to its generalized nature there are exceptions where it doesn't even apply (like people who are weight training, pregnant women etc).
I also think it is becoming apparent that weight alone is not as useful of a metric as was once believed. It doesn't do a great job of identifying whether or not someone is healthy.
So maybe it is the best that institutions can do to gauge the health of a population with the data given, but I think it could be improved drastically with extra metrics (I've heard of measurements around the body being used).
There is a subset of people who are being misidentified as unhealthy or overweight. Idk how large it is but it seems significant.
Obviously there are edge cases when applying statistical tools at a population level. That's not the purpose of BMI though. If you're pregnant or weight training then OBVIOUSLY your BMI score is not comparable to others. Most people are neither pregnant nor weight training. I'd even bet that most people haven't performed a weighted squat.
>> I also think it is becoming apparent that weight alone is not as useful of a metric as was once believed.
Being fat is bad for your health. That is a fact. No amount of body positivity advertisements will change that fact. It is healthier to be thinner than obese. Look at the graph above to see BMIs impact on other health outcomes.
Also, if you think the obesity crisis is fake/blown up then simply go outside in most American cities and you'll see that many people are obese.
> There is a subset of people who are being misidentified as unhealthy or overweight. Idk how large it is but it seems significant.
Again, this is not relevant as the subset of people is so small compared with the whole population.
The "data given" is included in BMI. Your relative weight is an indicator of multiple other health outcomes.
> There is a healthy body fat % range but it is all very dependent on someone's activity and consumption.
The amount of body fat being detrimental doesn't change if you just exercise more or eat healthier. It's still bad by itself. You might have a better life expectancy than someone who is a sedentary lump at the same body fat, but that isn't saying much. Hence, again, BMI is a gut check. If you're at BMI=27, you almost certainly need to lose weight. Period.
> I do however think BMI has its flaws and that those flaws are pretty significant when it comes to assessing individual health.
While this is not untrue, I don't understand why people single out BMI as a measure to criticise. No measure is perfect, or even particularly good in many cases, and yet it's BMI that gets it.
What do you mean by "weight alone"? Even BMI is kg/m^2 rather than just weight. Do you mean total body fat percentage isn't a metric of health, or BMI doesn't closely track BF%, or literally that weight isn't a good metric?
Sorry if this is nitpicky, but responding to the claim is difficult without defining it precisely. For example, see this figure from a study comparing BMI and BF% in Greece: https://www.researchgate.net/figure/BMI-versus-body-fat-perc.... You can see that BMI incorrectly says a lot of people are healthy when they're obese in terms of body fat, but if you don't agree that body fat is a metric of health then I'd first need to convince you of that.
Edit: I'll also note that in that figure you can see the number of people who are incorrectly labelled as obese by BMI but not by BF%, and it's much smaller than the other way round. BMI tends to make your population look healthier than it really is, whereas most people who say BMI is flawed are making the opposite argument.
> I think due to its generalized nature there are exceptions where it doesn't even apply (like people who are weight training, pregnant women etc).
All models are false, some are useful. BMI is imperfect, but arguing we should ignore it because it has exceptions is like arguing we should ignore that the Earth is round because it isn't perfectly spherical. Looking at BMI alone is occasionally misleading, but it's a hell of a lot better than not looking at anything, which is the realistic alternative.
> So maybe it is the best that institutions can do to gauge the health of a population with the data given, but I think it could be improved drastically with extra metrics (I've heard of measurements around the body being used).
No-one has identified any such drastic improvements. Measuring body fat percentage gives you a small improvement over BMI alone, but it's hard to do at home and only really matters for people with very high muscle mass (who are most likely already using more advanced measurements than BMI alone).
It would be nice if everyone could get a DEXA scan regularly, but the data shows that 50% of people would find that they have even more body fat than the BMI charts predict. Again, it's worse for women, most women near the overweight threshold would find that their body fat is higher than expected.
[0]: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2877506/