Gender Trends in Authorship Across Neuroradiology Journals (2016–2025)

BACKGROUND AND PURPOSE: Trends in neuroradiology authorship are not well characterized. This study performed a bibliometric analysis to assess gender trends in first and last authorship across four neuroradiology journals from 2016 to 2025. MATERIALS AND METHODS: A cross-sectional bibliometric analysis of four journals: American Journal of Neuroradiology, Neuroradiology, Journal of Neuroradiology and Journal of neuroimaging was performed for publications indexed in MEDLINE from January 1, 2016, to December 31, 2025. A previously validated natural language processing pipeline was used to infer author gender and country. Cochran-Armitage test and logistic regression was used to assess temporal trends. Multivariable logistic regression was performed to assess association between first and last authorship and adjusted for publication year, journal, geographic differences and article type. RESULTS: In the 7,477 publications that met inclusion criteria, women accounted for 24.8% of first authors (1,858/7,477) and 17.3% of last authors (1,290/7,477). Female first authorship increased from 21.5% in 2016 to 31.6% in 2025 (P<0.001), and last authorship from 15.2% to 21.8% (P<0.001). Female authorship varied across journals, with first authorship varying from 33.7% (Journal of Neuroimaging) to 19.1% (AJNR). Publications with female last authors had 2.12-fold higher odds of having a female first author compared with publications with male last authors (95% CI, 1.97-2.28; P<0.001). Wide geographic variations were noted with female first authorship ranging from 12.2% (Japan) to 34.2% (Switzerland). Female first authorship from United States was 21.4% while last authorship was 17.6%. CONCLUSION: Female authorship in neuroradiology journals has increased over the past decade, but remains low, particularly in senior authorship positions. Continued efforts to provide mentorship and leadership opportunities will help sustain progress towards an equitable workforce.