Nothing is plausible. People are different. Indexing them does not make them any more or less healthy. Rough guidelines is all you have, but actuarial science is fine by that, because they can come up with statistics based on the BMI, because it's reproducable. 10% in this range have claims, 20% in this range, 50% in this range, etc. So you can come up with a rate difference, but whether it actually will work is another thing.