Modern humans arrived in Europe around 40,000 years ago (in very round numbers), replacing Neanderthals by 30,000 years ago. Then, they were expelled from all but a few refugia in Southern Europe, by a catastrophic ice age which peaked at the Last Glacial Maximum (LGM) ca. 20,000 years ago, when most of Northern Europe was covered by glaciers because the average global temperature was about 4.5 degrees colder than it is today. As the glaciers receded, over a period of several thousand years after the LGM, modern humans repopulated Europe from the refugia and from outside of Europe entirely.
An analysis of skull shapes shows that pre-LGM Cro-Magnon modern humans had significantly different skull shapes than post-LGM Paleolithic modern humans, than early Neolithic era modern humans, and than modern humans in approximately their modern form a few thousand years after that.
A new study of these skull shapes concludes that about two-thirds of the differences between pre-LGM and modern modern human skull shapes comes from the gap between pre-LGM modern humans and immediately post-LGM Paleolithic modern humans. The next biggest gap was between post-LGM Paleolithic modern humans and early Neolithic modern humans. The smallest gap was between early and late Neolithic and later era modern humans.
The study offers no opinion concerning whether this reflects a different genetic makeup or other environmental or epigenetic factors.
The study is notable, because it is the first physical anthropology evidence of which I am aware that demonstrates substantial physical anthropology differences between modern humans in Europe before and after the LGM that must have had some cause, even if the exact cause is not known.
The data from mtDNA samples, in contrast, have tended to show more continuity between pre-LGM and post-LGM hunter-gatherer populations, which is a bit surprising itself given that we known for sure that there was a total replacement of the human population of most of Europe as a result of the LGM. Autosomal genetic data on pre-LGM modern humans in Europe is so scarce and so preliminary that it is too early to know if it differed materially from post-LGM Paleolithic populations unless you are actually in the labs doing the cutting edge work and have access to inside information as a result.
While analysis of skull shape is largely out of favor in anthropological and social scientific circles, for the very good reason that this data was grossly misapplied and misused in the 19th century hey day of this kind of scientific investigation in physical anthropology, that doesn't mean that modern studies taking this approach in the far more restrained context of modern scientific knowledge, are worthless. They tell us something, just not as much as 19th century proto-scientists thought that they did.