We use several blood tests to assess a woman's potential to respond to fertility treatment. Initial research on measuring anti-Mullerian hormone (AMH) levels, used for gauging assessing ovarian reserve, showed that they do not vary during and between menstrual cycles. A single blood test done at any time during the cycle was supposed to be diagnostic.
This was a huge improvement over FSH levels, which must be measured on cycle days 2, 3 or 4 and are known to vary widely from cycle to cycle.
At InVia, we have been measuring AMH levels for many years now. Our clinical experience was that repeat AMH levels did seem to vary in the same patient. We have therefore been careful not to base treatment protocols based on AMH alone. We do keep other relevant factors (age, day-3 FSH, AFC, previous ovarian response, BMI) in mind when planning an IVF treatment protocol. The basic question, however, remains “Do AMH levels vary”?
This very question was addressed in a recent publication by Rustamov et al. from the University of Manchester, UK (Fertil Steril 2011;95:1185-7). They investigated the variation between repeated AMH measurements in the same patient and between patients and compared AMH variability with that of serum FSH in the same patient group.
186 women who had repeated AMH levels over a year were investigated. Blood samples were randomly taken during the menstrual cycle when patients attended the clinic for the routine work-up. The reasons for repeating AMH measurements were:  clinic’s protocol for updating results every 6 months,  patient’s request, and  clinician not aware of the initial AMH measurement.
They used a Diagnostic Systems Laboratories (DSL) assay to measure AMH. Basal FSH levels were measured in the same study population. AMH levels are allocated to five bands with differing ovarian reserve: ≤ 2.2 pmol/L, very poor ovarian reserve; > 2.2 to ≤ 15.7 pmol/L, low ovarian reserve; >15.7 to ≤ 28.6 pmol/L, satisfactory (i.e., normal) ovarian reserve; >28.6 to ≤ 48.5 pmol/L, high ovarian reserve; > 48.5 pmol/L, suggestive of polycystic ovary syndrome (PCOS). (Please divide these numbers by 7.14 to convert the values into ng/mL). Here is what they found.
They observed a variation in AMH levels of 28%!!! The variability terms between samples and patients were statistically significant (P<.001). Repeated AMH values in 42 of 172 women (24%) decreased into different (but adjacent) clinical categories when compared with the first AMH sample.
FSH levels also varied. In fact the variation in FSH levels (27%) was comparable to the variation in AMH levels!
They could not find any specific clinical factors that accounted for the variability in AMH levels. The looked at stage of the menstrual cycle, presence of ovarian cysts, recent ovarian surgery, treatment with birth control pills and IVF treatment. The CV (co-efficient of variation) of the assay was only 4.8%. So, this could not be an explanation for a 28% decrease in AMH values.
They conclude that contrary to the original research publications, AMH levels do vary and this should be kept in mind when IVF treatment protocols are planned. In a subsequent publication, the authors investigated a much larger group of patients and suggest there may be a problem with the assay used for measuring AMH. This will be discussed in a follow up blog.
So what does all this mean?
Using currently available assay kits, AMH levels do vary. In fact, they vary just as much as FSH levels! One must therefore be careful when making clinical recommendations based on a single AMH level by itself.
Dr. Karande is Board Certified in the specialty of Obstetrics and Gynecology as well as the subspecialty of Reproductive Endocrinology and Infertility. He is a Fellow of the American College of Obstetricians and Gynecologists and Member of the American Society for Reproductive Medicine.