Is there a relationship between anti-Mullerian hormone (AMH) levels and the prediction of ovarian reserve in reproductive-age women undergoing fertility evaluation? : A Systematic Review

Introduction: Accurate assessment of ovarian reserve is crucial for counseling women undergoing fertility evaluation. Anti-Müllerian hormone (AMH) has emerged as a promising biomarker, yet its predictive value for ovarian reserve in reproductive-age women requires systematic evaluation. This systematic review aims to determine the relationship between AMH levels and ovarian reserve prediction in reproductive-age women undergoing fertility assessment. Methods: A systematic review of 80 studies involving reproductive-age women (18-45 years) undergoing fertility evaluation was conducted. Studies were included if they measured serum AMH levels, assessed ovarian reserve through validated markers (antral follicle count [AFC], ovarian volume, response to stimulation), and examined their correlation. Data extraction encompassed AMH measurement protocols, ovarian reserve assessment methods, statistical relationships, and confounding factors. Results: AMH demonstrated consistently strong correlations with AFC across studies (r=0.48-0.89), with the strongest correlations observed with histological primordial follicle count (ρ=0.75). For predicting poor ovarian response, AMH showed excellent discriminatory ability (AUC 0.75-0.93), with cutoff values ranging from 0.7-1.37 ng/mL across different assays. AMH outperformed traditional markers including basal FSH and demonstrated superior reproducibility (ICC=0.89) compared to AFC (ICC=0.73). Age-stratified analyses revealed that in women >35 years, AMH (AUC=0.858) significantly outperformed AFC (AUC=0.675) in predicting suboptimal response. For high response prediction, AMH achieved AUC values of 0.81-0.91. However, AMH's predictive ability for pregnancy outcomes was more modest (AUC 0.56-0.63), operating primarily through oocyte quantity rather than quality. Discordance between AMH and AFC occurred in approximately 20% of women, increasing with age. Discussion: AMH consistently demonstrates strong correlation with ovarian reserve markers, particularly AFC, and shows excellent predictive accuracy for extremes of ovarian response in assisted reproduction. Its superior reproducibility and age-dependent performance—especially superior to AFC in women over 35 years—support its role as a primary ovarian reserve marker. However, significant heterogeneity exists due to assay platforms, population characteristics, and outcome definitions. The biological variability and assay-specific differences necessitate platform-specific reference ranges. Conclusion: AMH is a reliable predictor of ovarian reserve in reproductive-age women undergoing fertility evaluation, with strongest performance for quantitative outcomes and extremes of response. Clinical implementation requires standardized assays, age-specific interpretation, and recognition that AMH predicts oocyte quantity rather than quality. Future research should focus on assay standardization, population-specific thresholds, and integration with other clinical parameters.