Comparer des méthodes

	Méta-analyse dose-réponse	Méta-régression
Domaine	Synthèse des données probantes	Méta-analyse
Famille	Process / pipeline	Regression model
Année d'origine	1992	2002
Auteur d'origine	Greenland & Longnecker (1992), Advanced by Orsini et al. (2012)	Simon Thompson & Julian Higgins
Type	Method	Weighted regression for effect-size heterogeneity
Source fondatrice	Greenland, S., & Longnecker, M. P. (1992). Methods for trend estimation of environmental health risks, with application to exposure to contaminated groundwater. Statistics in Medicine, 11(14‐15), 1837–1847.	Thompson, S. G., & Higgins, J. P. T. (2002). How should meta-regression analyses be undertaken and interpreted? Statistics in Medicine, 21(11), 1559–1573.
Alias	Dose-Response Relationship, Non-Linear Meta-Analysis, Dose-Effect Synthesis	Meta-Analytic Regression, Weighted Regression in Meta-Analysis, Moderator Analysis, Meta-regresyon
Apparentées	1	2
Résumé	Dose-response meta-analysis is a specialized evidence synthesis method that models the relationship between exposure dose (or intensity, duration, quantity) and health outcome across multiple studies, assessing whether effects follow a linear trend, nonlinear curve, or threshold pattern. Pioneered by Greenland and Longnecker (1992) and refined by Orsini et al. (2012), dose-response meta-analysis answers critical questions like 'Does cardiovascular disease risk increase consistently with salt intake, or is there a threshold above which risk plateaus?' and 'Does the benefit of physical activity increase linearly with exercise duration, or do diminishing returns occur at high doses?' This method is essential for risk assessment, policy-setting on safe exposure limits, and optimizing treatment dosing.	Meta-regression is a statistical technique that extends conventional meta-analysis by regressing study-level effect sizes on one or more study characteristics (moderators) to explain between-study heterogeneity. Formalized by Thompson and Higgins in 2002, it uses weighted least squares — weighting each study by the inverse of its variance — within a mixed-effects framework, allowing researchers to identify which study features systematically account for variation in observed effects across the literature.