Comment by rootusrootus

It depends. Overall, lowest all-cause death indexed on BMI seems to be around 27 plus or minus [0]. Interestingly, for all subjects (not just healthy non-smokers), BMI 30-32 isn't really associated with a dramatic increase in all cause mortality.

Obviously life is more complicated than just one measure. I figure this is just another data point saying BMI is useful for population studies and not great for individual diagnosis.

[0] https://pmc.ncbi.nlm.nih.gov/articles/PMC10321632/