AUTHOR=Ballena-Caicedo Jhosmer , Zuzunaga-Montoya Fiorella E. , Loayza-Castro Joan A. , Bustamante-Rodríguez Juan Carlos , Vásquez Romero Luisa Erika Milagros , Tapia-Limonchi Rafael , De Carrillo Carmen Inés Gutierrez , Vera-Ponce Víctor Juan 

TITLE=Global prevalence of insulin resistance in the adult population: a systematic review and meta-analysis

JOURNAL=Frontiers in Endocrinology

VOLUME=Volume 16 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2025.1646258

DOI=10.3389/fendo.2025.1646258

ISSN=1664-2392

ABSTRACT=ObjectiveTo determine the global prevalence of IR, evaluating differences according to study designs and population characteristics.MethodologyA systematic review with meta-analysis was conducted. The search encompassed MEDLINE (PubMed), Scopus, Web of Science, and EMBASE, including observational studies that employed the HOMA-IR index to estimate IR and published adult prevalence data. Articles without clear IR definitions or with highly specific populations were excluded. The meta-analysis applied a random-effects model with proportion transformation (Freeman-Tukey), assessing heterogeneity with I² and Cochran’s Q test. Additionally, a meta-regression by publication year was conducted.ResultsEighty-seven studies were included, with 235,148 participants. The pooled prevalence of IR was estimated at 26.53% (95% CI: 24.10–29.03; I²=99%), with no statistically significant differences when comparing probabilistic versus non-probabilistic sampling or when stratifying by sex. The meta-regression revealed no clear variations according to publication year or other explored factors.ConclusionsThis systematic review demonstrates that IR reaches a global prevalence of 26.53%, with estimated differences between 26% and 30% across different populations and geographical regions. Despite the diversity in cut-off points employed for HOMA-IR, no statistically significant differences were observed when comparing sampling designs or stratifying by sex. Furthermore, no clear trend related to publication year was evidenced.