Abstract STUDY QUESTION Does FMR1 repeat length confer clinically meaningful predictive value for premature ovarian insufficiency (POI)? SUMMARY ANSWER FMR1 repeat length increases POI risk from ∼36 repeats onward but has limited diagnostic utility compared with a polygenic score for menopause timing. WHAT IS KNOWN ALREADY FMR1 premutation carriers (≥55 repeats) are reported to have high risk of Fragile-X Associated Primary Ovarian Insufficiency (FXPOI), but prior studies were small and highly ascertained. STUDY DESIGN, SIZE, DURATION Cross-sectional analysis of ∼92 000 women from the UK Biobank with genetic and health data. PARTICIPANTS/MATERIALS, SETTING, METHODS Female UK Biobank participants were genotyped for FMR1 repeat length. Associations with self-reported POI, age at menopause, and other reproductive phenotypes were analysed in women. FMR1 protein levels were measured, and genome-wide analyses were conducted to identify potential genetic modifiers. MAIN RESULTS AND THE ROLE OF CHANCE Of 518 female premutation carriers with available age at natural menopause, only 6. 9% reported POI. Elevated POI risk was observed starting at 36 repeats, increasing continuously with repeat length, but no threshold showed strong predictive power (maximum AUC 0. 60 vs AUC 0. 64 for polygenic score). No association was found between repeat length and FMR1 protein levels, consistent with an RNA gain-of-function toxicity mechanism. RAD52 was identified as a potential genetic modifier. LARGE SCALE DATA UK Biobank resource (https: //www. ukbiobank. ac. uk). LIMITATIONS, REASONS FOR CAUTION POI was self-reported rather than clinically confirmed. Analyses could not assess AGG interruptions, mosaicism, or X-inactivation. Genetic modifiers require replication. Findings are limited to a single population dataset. WIDER IMPLICATIONS OF THE FINDINGS These results challenge the utility of the FXPOI disease category, suggest limited diagnostic value of clinical FMR1 premutation testing for POI, and highlight alternative mechanisms and potential modifiers such as RAD52. STUDY FUNDING/COMPETING INTEREST (S) This work was conducted using the UK Biobank resource (application 9905). This work was funded by the Medical Research Council (unit programs: MCUU₁2015/2, MCUU₀0006/2) and Wellcome (Discovery award 302536/Z/23/Z). The sponsors had no role in the study design, collection, analysis, or interpretation of the data, the writing of the manuscript or the decision to submit it for publication. J. R. B. P. and A. M. have engaged in paid consultancy for Ovartix Ltd. TRIAL REGISTRATION NUMBER N/A.
Morbey et al. (Thu,) studied this question.