Disproportionality analysis of dietary supplement adverse events in the FDA CAERS database, 2004-2025

Dietary supplements are consumed by over half of US adults, yet post-market safety surveillance remains limited. We applied four disproportionality methods (PRR, ROR, GPS, BCPNN), CUSUM temporal detection, and demographic stratification to 48,840 dietary supplement adverse event reports from the FDA CAERS database (2004-2025). Analysis was conducted at the product-name level, not the ingredient level, and the signals reported below are product-name-level signals to be interpreted as flags for further investigation rather than confirmed ingredient-specific risks. Of 4779 product-adverse event pairs, 3017 (63.1%) were detected by three or more methods. After variant-name consolidation of the most heavily fragmented products (Section 4.4 and Supplementary Material 4), this resolves to an estimated 1800-2200 distinct product-level signals. Kratom dominated the critical risk tier, with Kratom-Death ranked highest (PRR 19.7, N = 178). Hepatotoxicity signals clustered in herbal/botanical and weight loss products. Herbal/botanical supplements carried the highest serious outcome rate among classifiable products (78.0%; adjusted OR 2.08, 95% CI 1.53-2.84); the 54% "Other"-category share, however, limits the generalisability of category-specific estimates to all supplements within each nominal category. CUSUM detected 451 temporally emerging signals including preliminary hepatic enzyme elevations for AG1 and Nutrafol (2023-2025). Of consensus signals, 97.3% survived false discovery rate correction; 81% of established international supplement safety signals were recovered. These findings support regulatory prioritisation of herbal/botanical and weight loss supplement categories and identify emerging signals warranting continued monitoring.