To evaluate the association, if any, between serum antimüllerian hormone (AMH) levels and probability of clinical pregnancy and spontaneous abortion (SAB) in the infertility setting.

Retrospective cohort study.

Academic fertility center.

A total of 1,861 gonadotropin stimulation/intrauterine insemination cycles stratified by AMH levels into 3 groups: Low, <25th percentile (<0.7 ng/mL); Middle, ≥25th and <75th percentile (0.7–4.4 ng/mL); and High, ≥75th percentile (≥4.5 ng/mL).

Intrauterine insemination after stimulation with gonadotropins.

Cumulative probability of clinical pregnancy over a maximum of 3 and/or 6 cycles and SAB incidence risk rate (IRR). The Kaplan-Meier failure function (log rank test), Cox proportional hazards models, and multilevel mixed-effects Poisson regression models were performed to compare outcomes among the AMH groups.

Overall, in both unadjusted and adjusted models, the probability of achieving a clinical pregnancy was higher in the Middle and High AMH groups compared with that in the Low AMH group, both over 3 (hazard ratios [95% confidence interval], 1.55 [1.05–2.29] and 1.85 [1.22–2.81], respectively) and 6 (1.71 [1.17–2.48] and 2.12 [1.42–3.16], respectively) cycles.

In the unadjusted models, the SAB IRR was higher among the Low AMH group (IRR [95% confidence interval], 2.17 (1.11–4.24]), with the relationship persisting after adjusting for age (1.83 [0.93–3.60]). When the SAB IRR were calculated separately for the subpopulations with and without polycystic ovary syndrome, a similar relationship was noted among the latter in the unadjusted (1.94 [0.97–3.88]) and adjusted (1.74 [0.86–3.49]) analyses.

In women undergoing gonadotropin stimulation/intrauterine insemination, AMH appears to affect the probability of achieving a clinical pregnancy. A possible negative impact, independent of age, on the risk of SAB was also suggested.