The differences observed are at the population level. That means there is HUGE overlap between the two populations and you can’t distinguish any data point as being a girl or a boy. Only that the populations have a different average or range. For a visual representation check out [1].

This kind of thing shows up all over the place in all fields of research. It’s why for example although you can see immune system differences at a population level between patients with disease X and healthy controls, you can’t in any way use it as a biomarker for an individual who might have disease X.

[1] https://link.springer.com/article/10.1186/s13293-026-00976-9...