Preclinical evidence construction for resveratrol against pulmonary fibrosis: A meta-analysis and machine learning study

Resveratrol, a natural polyphenol, may serve as a potential drug for pulmonary fibrosis (PF). This study uses meta-analysis and machine learning to assess resveratrol's efficacy and optimal treatment on PF animals. Four databases were systematically searched. Eleven machine learning models were employed to predict the best dosage/medication time/frequency range to affect relationships. Eighteen articles were included, encompassing 545 animals. This study indicates that resveratrol significantly improved the indicators for PF, including histopathological scores, transforming growth factor-β, hydroxyproline, and tumor necrosis factor-α. Anti-inflammatory and anti-oxidative stress are the main mechanisms of resveratrol in the intervention of PF. Efficient dosage (over 22.10 mg/kg/day), appropriate timing (under 11.66 days or between 26.71 and 133.78 days), and frequency (over five times a week) enhance anti-PF effects. Resveratrol is effective in treating PF and is a promising drug. These findings support the potential of resveratrol in developing targeted preclinical strategies for animal models of PF. Created in BioRender. Wu, X.(2026) https://Biorender.com/htfdpss . • First machine learning-optimized dosing regimen for resveratrol against pulmonary fibrosis. • Resveratrol reduces key fibrosis markers mainly via anti-inflammatory and antioxidant mechanisms. • This study framework bridges preclinical data and dietary intervention design. • Validation of resveratrol as a high-potency nutraceutical for pulmonary fibrosis.