Anti-Mullerian Hormone (AMH) is a protein produced by preantral and small antral follicles. AMH is a proxy for the size of the primordial follicle pool and therefore a marker for ovarian reserve. AMH values are relatively stable through the menstrual cycle because it is not secreted by a dominant follicle or corpus luteum. However, long-term ovarian suppression with hormonal contraception or pregnancy is associated with lower levels (Dewailly et al 2014). AMH has been well studied in infertile women but data are scarce in non-infertile women. We do not yet have an established “normal” range of AMH values. Seifer et al (2010) published age specific medians, means and standard deviations for AMH for 17,120 women presenting to fertility centers. The figure below shows the yearly decline in mean and median values for AMH with increasing age.
Graph of AMH age-specific median values with mean +/- SD AMH values for women ages 24–50 at 1-year intervals. The median was substantially lower than the average, suggesting non-normally distributed AMH values by age. Seifer. Age-specific AMH values for U.S. clinics. Fertil Steril 2011.
AMH predicts ovarian response to hyperstimulation with gonadotropins, which is helpful when counseling patients about their prognosis for success with fertility treatment. However, AMH is not a good predictor of spontaneous fertility. In a study by Hagen et al., fecundability in healthy young women appear not to be compromised if low AMH levels are present. Conversely, the probability of conceiving was reduced in women with high AMH levels, which is not surprising as these subjects represents those with conditions of anovulation such as PCOS. Women with very low AMH levels will have a lower chance of success with fertility treatment compared to other women of the same age with higher AMH levels. However, the chance of success even with very low AMH values is still decent, so treatment is not withheld purely based on AMH levels. Discussion of the ovarian reserve with the patient is important to set their expectations appropriately, as those with low AMH levels may have only a few oocytes develop and treatment has a high chance of cancellation due to a poor response. Once this happens we usually introduce the topic of third party reproduction such as egg or embryo donation.
A recent observational study by Streuli et al. shows that estimated time to pregnancy was correlated with age but not AMH. This speaks against using AMH in women who are not infertile for the purpose of predicting their chance of pregnancy. AMH is a quantitative but not a qualitative marker of ovarian reserve, and therefore does not reflect a woman’s ability to become pregnant spontaneously. For the time being, age is still the best predictor of egg quality.