Yes, it’s a pet peeve.
This is a comment from over at Instapundit, figured I’d copy-paste it so that it has a chance of being read.
Found the article.
Those who looked at the years and went “hm, first one is before the BMI change, second is after,” yes they used BMI studies for each span.
On the upside, they did at least pick a cohort that has almost entirely hit puberty, 16-19.
On the other hand, they were using a screening test to diagnose “overweight.”
It’s known bad:
To find out whether BMI correlated with actual markers of health, a team of UCLA researchers analyzed data from 40,420 individuals who participated in the 2005-2012 National Health and Nutrition Examination Survey. They looked at individuals’ blood pressure, triglycerides, cholesterol, glucose, insulin resistance and C-reactive protein data — markers that are linked to heart disease and inflammation, among other issues.
They found that nearly half (47.4%) of overweight people and 29% of obese people were, from a metabolic standpoint, quite healthy. On the flip side, more than 30% of individuals with “normal” weights were metabolically unhealthy.
And “overweight” isn’t significantly associated with bad health outcomes– the ‘risk’ there is that you MIGHT be headed to “obese.” The lowest health risk appears to be situated inside of “overweight.”
Which makes sense for a screening test– having more body fat improves the outcomes if you get sick, but you don’t want so much body fat that it makes you sick, so logically the healthiest folks would be inside of the false positive.