Flawed Meta‑Analyses Drive Regulation

Updated: 2026.05.07 26D ago 2 sources
When meta‑analyses mix inappropriate effect measures or selectively use adjusted statistics, they can produce large, misleading estimates of population health impact. Those inflated numbers can then be cited by regulators or media to justify costly bans or mandates that lack a solid causal basis. — Shows how technical epidemiological mistakes can have outsized political and economic consequences by creating a veneer of scientific certainty for regulatory action.

Sources

Do anti-cholinergic drugs cause dementia?
Sebastian Jensen 2026.05.07 72% relevant
The author scrutinizes the statistical interpretation of PET/tangle studies and the inference from biomarker associations to clinical dementia, mirroring the concern that meta‑analyses and aggregated observational evidence can mislead policy and clinical guidance when methodologic subtleties are ignored.
Lying for a Climate Crusade - Cremieux Recueil
2026.03.05 100% relevant
The MDPI paper (cited by a Los Angeles Times story and linked to CPSC discussion) estimated 12.7% of U.S. asthma attributable to gas stoves by feeding meta‑analytic results into an attributable‑fraction formula; the article argues this relied on misusing prevalence ratios as odds ratios and inconsistent adjusted versus raw estimates.
← Back to all ideas