Prevalence of myopia in Europe: a systematic review and meta-analysis of data from 14 countries.

Background: Although myopia prevalence increased in East Asian countries, the burden of myopia in Europe is less known. We performed a systematic review and meta-analysis to estimate the prevalence of myopia in Europe and at the country level.

Methods: We searched PubMed, Scopus and Web of Science to identify studies on myopia prevalence published until January 2024, regardless of language. We included European cross-sectional and cohort studies with defined sampling strategies and excluded clinical surveys, myopia registries, self-reported near-sightedness, and non-representative populations. Pooled prevalence was estimated using random-effects models. Heterogeneity was assessed using the Cochran's Q (χ test) and the I statistic. The study protocol was preregistered in PROSPERO (CRD42023471527).

Findings: We screened 2074 records and included 22 studies (from 14 European countries; n = 128,012) in the meta-analyses. The pooled prevalence of myopia was 23.5% (95% CI: 18.5-29.3; I = 99.7%), ranging from 11.9% in Finland to 49.7% in Sweden. In cycloplegic studies, myopia prevalence was 18.9% (95% CI: 13.2-26.5%; I = 99.7%) vs. 31.2% (95% CI: 24.9-38.3%; I = 99.3%) in non-cycloplegic studies. Subgroup and meta-regression analyses exploring sources of heterogeneity showed a lower prevalence in children (6-11 years; 5.5%) compared with adolescents (12-17 years; 25.2%) and adults (18-39 years; 24.3%) in cycloplegic studies. No significant differences in prevalence were observed between sexes. Myopia prevalence increased significantly between 2000-2010 and 2011-2022 (p = 0.040), although age-specific trends remained stable.

Interpretation: Myopia prevalence in Europe is lower than in Asia, with a less pronounced increase that disappears after stratifying by cycloplegic refraction and age. These findings highlight the need for age-specific data and cycloplegic refraction in future studies to reduce heterogeneity. Uneven country representation may limit the generalisability of these results.

Funding: The present publication was funded by Fundação Ciência e Tecnologia, IP national support through UID/04923-Comprehensive Health Research Centre.