Editing Openai/69177fbb-14c4-800f-bc57-ccb5a1d88c95 (section)

=== - Fit: If selection favored cognitive/learning traits (e.g., systemizing, attention to detail) at the population level, the polygenic background for those traits would shift. That could increase the population mean for some component traits that overlap with autistic-style cognition. (This is what you were intuiting.) ===
* Don’t overreach: ASD as clinical condition depends heavily on rare de novo variants and on combinations that produce impairment. Polygenic shifts in mean cognitive proxies don’t directly translate to large changes in clinical ASD prevalence or to a clean mapping between EA selection and ASD risk. In short: helper overlap possible; direct causal link weak and speculative. PubMed<ref>{{cite web|title=PubMed|url=https://pubmed.ncbi.nlm.nih.gov/35361970/|publisher=pubmed.ncbi.nlm.nih.gov|access-date=2025-11-17}}</ref>