I did not see information on the ancestry of the 9-10 ASD toddlers or on the even fewer neurotypical controls. May have missed this. As for large putative effe

I did not see information on the ancestry of the 9-10 ASD toddlers or on the even fewer neurotypical controls. May have missed this. As for large putative effect sizes this is often due to subtle batch processing differences between cases and controls. Where they all processed by the same tech in an interleaved way? All stored in the same way over this very long duration study? Authors do discuss batch controls but with single digit sample sizes I regard statistics as fundamentally unreliable. I am also not convinced by the claim of any over-proliferation of neurons in autism during development. It is certainly a highly controversial result. See notes above by “subiculum…” and the Li et al paper he cites.