Premium
The prevalence of migraine in university students: a systematic review and meta‐analysis
Author(s) -
Wang X.,
Zhou H. B.,
Sun J. M.,
Xing Y. H.,
Zhu Y. L.,
Zhao Y. S.
Publication year - 2016
Publication title -
european journal of neurology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.881
H-Index - 124
eISSN - 1468-1331
pISSN - 1351-5101
DOI - 10.1111/ene.12784
Subject(s) - migraine , medicine , meta analysis , subgroup analysis , confidence interval , sample size determination , demography , pediatrics , statistics , mathematics , sociology
Our aim was to determine the prevalence of migraine amongst university students. Migraine is highly prevalent amongst university students, but the exact frequency remains inconsistent between studies. PubMed, Embase and Google Scholar databases were used to identify studies dealing with the prevalence of migraine amongst university students published between 1 January 1988 and 31 August 2014. The pooled migraine prevalence was calculated using DerSimonian and Laird's random‐effects model. Heterogeneity of the results was investigated using subgroup analysis and the trend of migraine prevalence according to the publication year and sample size was determined by cumulative analysis. Data were combined from 56 independent studies, analysing a total of 34 904 students. The pooled migraine prevalence was 16.1% [95% confidence interval ( CI ) 13.6%–18.9%]: 11.3% (95% CI 8.8%–14.4%) amongst male students and 21.7% (95% CI 18.0%–25.8%) amongst female students. Subgroup analysis revealed that diagnostic criteria ( P < 0.0001) and gender distribution ( P = 0.004) significantly affected migraine prevalence. Cumulative analysis found that the 95% CI became narrower with ascending publication year and sample size. Many studies agree that migraine is highly prevalent amongst university students, but diverse methodologies lead to substantial heterogeneity in the results. It is shown that gender and diagnostic criteria significantly influence the migraine prevalence and may partially explain the heterogeneity between studies.