Reply to: Do not underestimate the cognitive benefits of exercise

not read — see design_notes

Not applicable — the authors' reply to Dupuy et al. 2024, itself a commentary on Ciria et al. 2023

ACCESS FAILURE, STATED PLAINLY. Nature Human Behaviour deposits no abstract for this reply with Crossref, Europe PMC or Semantic Scholar, and the article page redirects to an authentication gateway; no preprint or repository copy was located. We could not read it. The content below is secondhand from Sanabria et al. 2026 — an overlapping author group, which is a reason for caution, not confidence — and is flagged unverified. Filed as a known unread lead, not cited as evidence from the topic page. The underlying materials the reply points to are public at osf.io/e9zqf and the original umbrella review's peer reviews are posted by the journal, so a future verification pass has a route in.

NOT VERIFIED AGAINST THE PAPER. Per Sanabria et al. 2026, the reply reports that re-running the umbrella analysis with all the moderators Dupuy et al. requested (exercise type, intensity, dose) left the original bias-corrected near-null unchanged, that multiple publication-bias correction methods and both frequentist and Bayesian analyses were applied, and that the transparency objection is answered by the public materials and open peer reviews. This reply has 8 citations against the commentary's 30.

Not applicable — a methodological reply, not an empirical study.