Can FSH or AMH accurately predict pregnancy chances with IVF?

A question we hear quite frequently from patients is whether follicle stimulating hormone (FSH)anti-Müllerian hormone (AMH)or both are predictive of pregnancy chances with IVF. Though on first impulse the answer appears to be a simple YES, reality is much more complex. Let us explain why: 

Both FSH and AMH are hormones that to a significant degree reflect what is called the ovarian reserve (OR). At CHR, we prefer the term functional ovarian reserve (FOR),because OR is really made up of two distinct components, the so-called resting follicles (primordial follicles)and growing follicles, which are resting follicles that have been recruited into a journey of 3-4 months of maturation. Resting follicles represent the vast majority of total OR but cannot be measured. What FSH and AMH can tell us about is the growing follicle pool only. But because the growing follicle pool, of course, is dependent on the resting follicle pool, there is always a correlation.

Age-specific normal AMH levels and FSH levels change as women age.

As women age, with declining growing follicle pool, FSH increases and AMH decreases. Both hormones make these age-adjustments in concert, but only to a degree. At young ages, AMH is somewhat more predictive of IVF outcomes than FSH, while at more advanced ages, the opposite is the case. They correlate best at median ages, but, in principle, the more normal they are, the better will be the prognosis in IVF cycles.

What represents “normal” is, however, not always as clear as it may sound. The principal reason is that what laboratories use as “normal range” includes women of all ages. This, of course, means that for younger women, upper margins of normal FSH levels and lower margins of normal AMH levels are highly exaggerated. In other words, FSH and AMH must be assessed in age-specific ways. Since most laboratories list FSH levels up to 12.0 mIU/mL as normal (some up to 10.0 mIU/mL), most IVF centers still consider a woman to have abnormally low FOR only if FSH is over 10.0-12.0 mIU/mL. These levels would, however, be clearly abnormal in a 35-year-old patient. Similarly, an AMH level of 1.0 ng/mL can be considered excellent in a 43-year-old but would reflect low FOR in a 37-year-old.

CHR has been assessing FSH and AMH levels in age-specific ways for over 10 years and has, indeed, published the center’s age-specific cut-offs (see figure). Additional publications have since appeared, mostly defining age-specific AMH levels. Utilization of age-specific FSH and AMH levels has, however, been disappointing and, in CHR’s opinion, is the principal reason why the diagnosis of premature ovarian aging (POA)/occult primary ovarian insufficiency (oPOI) is still so frequently overlooked in daily infertility practice at so many IVF centers.

In summary, FSH and/or AMH are predictive of IVF outcomes to a degree because both represent FOR and FOR, in turn, is reflective of egg and embryo numbers obtained in an IVF cycle, which is the second most important predictor of IVF success. (The most important predictor of IVF success is the female age.) Quite a number of additional factors, including how ovaries are stimulated, also influence IVF outcomes and, in isolation, FSH/AMH, therefore, have only limited predictability.

This is a part of the March 2019 CHR VOICE.

Norbert Gleicher, MD, leads CHR’s clinical and research efforts as Medical Director and Chief Scientist. A world-renowned reproductive endocrinologist, Dr. Gleicher has published hundreds of peer-reviewed papers and lectured globally while keeping an active clinical career focused on ovarian aging, immunological issues and other difficult cases of infertility.