But at a population level (as the article uses it) it’s mostly fine. Unless the sample skews tall or muscular like the Dutch or bodybuilders.