Health ArticleEducational review — not personal medical advice

Understanding Anti-Müllerian Hormone: A Key Blood Test for Female Fertility

Anti-Müllerian hormone (AMH) is a promising new blood test that helps doctors assess a woman's ovarian reserve — the quantity and quality of remaining eggs.

21 min

Table of Contents

Key Points

  • AMH blood test helps assess ovarian reserve, reflecting the quantity and quality of remaining eggs.
  • AMH predicts poor or excessive ovarian response before IVF, aiding individualized treatment planning to prevent complications like OHSS.
  • Low AMH does not mean you cannot get pregnant; many women, especially younger ones, still achieve pregnancy.
  • AMH is elevated in polycystic ovary syndrome and can support diagnosis and severity assessment.
  • AMH is a tumor marker for granulosa cell tumors and helps evaluate premature ovarian failure risk.

Why This Research Matters: The Growing Need for Fertility Biomarkers

In Western societies over the last 50 years, rising educational levels and increased participation in the workforce have pushed the average age at which women deliver their first child steadily upward. This delay has important biological consequences: female fertility begins to decline as early as the early twenties due to decreasing ovarian reserve — a term that refers to both the quality and quantity of the ovarian follicle pool.

Many women are consequently faced with unexpected fertility problems. The rate at which ovarian reserve declines varies greatly from woman to woman, making it a significant challenge for doctors to estimate an individual woman's remaining reproductive lifespan. There is therefore a growing demand for sensitive biomarkers that can accurately measure ovarian reserve and provide proper counseling and treatment for women experiencing infertility.

Anti-Müllerian hormone (AMH), also known as Müllerian-inhibiting substance, is a dimeric glycoprotein belonging to the transforming growth factor-β superfamily that is involved in growth and differentiation. The hormone was originally described as a gonadal factor in males that causes regression of the Müllerian ducts — the precursors to the female reproductive tract in early mammalian embryos. In males, AMH is expressed by Sertoli cells from the eighth week of gestation until puberty begins.

In females, however, AMH is expressed in granulosa cells and secreted into the circulation from birth until menopause, making it measurable in plasma. AMH levels are barely detectable at birth, increase significantly at puberty, then decline slowly throughout the reproductive period until they become undetectable at menopause.

How This Review Was Conducted

Researchers systematically searched the PubMed database for studies published up to November 2011, using the medical subject heading term "anti-müllerian hormone" alone and in combination with "analysis," "blood," "diagnostic use," "physiology," "secretion," and "therapeutic use." This search strategy yielded 235 publications.

Of these 235 publications, 96 were excluded because they were not in English, were not related to humans, or were not related to females. The remaining 139 publications were examined and assessed for methodological quality and relevance to predefined themes: AMH in female infertility, ovarian physiology, ovarian reserve, IVF, and PCOS. Preference was given to original clinical studies over review articles. References of retrieved publications were also reviewed to identify potentially relevant studies missed in the original search.

To specifically evaluate AMH as a predictor of IVF outcome, the researchers preselected articles examining the association between AMH and poor ovarian response, pregnancy, or hyper-response. Studies were included only if adequate data could be obtained — including specificity, sensitivity, cut-off values, and the assay used for AMH measurement. Publications reporting single cases or small case series were excluded as insufficiently strong evidence. Only studies that defined poor ovarian response as four or fewer oocytes were included to make results reliably comparable.

Ultimately, 12 prospective and 7 retrospective cohort studies, along with 1 case-control study, were included to evaluate AMH as a predictor of IVF outcome. The final review was based on predominantly original publications and a few reviews — 80 publications in total — considered to be of sufficiently high scientific standard and high relevance.

AMH and the Basics of Ovarian Function

Follicles represent the basic functional unit of the ovaries. During fetal life, germ cells populate the ovary and become surrounded by somatic cells, forming primordial follicles. The primordial follicle formation begins at mid-gestation and is complete shortly after birth, by which time about one million primordial follicles are present.

Throughout life, until their numbers are exhausted, primordial follicles leave the resting pool to enter the growing pool — a process called initial recruitment. This is a continuous process not controlled by the pituitary–gonadal endocrine axis. The majority of these growing follicles will undergo atresia (degeneration) unless they are rescued by follicle-stimulating hormone (FSH).

The rescue of a cohort of growing follicles by FSH begins after puberty when the hypothalamic–pituitary–gonadal endocrine axis is activated. This process, called cyclic recruitment, results in only one follicle being selected to become the dominant follicle, which will ovulate under the influence of luteinizing hormone.

AMH cannot be detected in resting primordial follicles, but its expression rapidly increases once a follicle reaches the preantral and antral stages. The highest level of AMH expression is found in granulosa cells of preantral and small antral follicles of no more than 4 mm in diameter. Expression disappears as follicles develop into larger antral and preovulatory stages.

Mouse studies on AMH-knockout models have proven illuminating. These mice exhibit a more rapid rate of primordial follicle recruitment, and their primordial follicle pool becomes prematurely exhausted. This indicates that AMH exerts a negative, restraining influence on initial recruitment. In vitro culture of neonatal ovaries in the presence of AMH confirmed its inhibitory effect. AMH also appears to inhibit cyclic recruitment of follicles by lowering their FSH sensitivity, thereby controlling the number of preantral and small antral follicles that escape atresia and continue growing to reach the preovulatory stage. In short, AMH is a strong suppressor of both initial and cyclic recruitment, preventing premature depletion of the follicle pool.

How AMH Is Measured in the Blood

AMH is secreted by granulosa cells into the circulation and is therefore measurable in plasma. Since 2004, two commercial enzyme-linked immunosorbent assays (ELISAs) have been available: one from Immunotech (Marseille, France) and another from Diagnostic Systems Laboratories (DSL; Webster, Texas, USA). The Beckman Coulter Company now owns both.

A high correlation exists between values obtained by the two kits, although the absolute values are around four times lower with DSL. No international reference standard has yet been established for AMH. Results may be reported in nanograms per milliliter or picomoles per liter — 1 ng/mL corresponds to 7.14 pmol/L. Caution is required when comparing results from different publications, since different ELISAs may have been used.

The Beckman Coulter Company has developed a second-generation ELISA called the AMH Gen II assay, which will most probably replace the earlier two assays. No automated assays have yet been developed, although an automated platform may become available in the near future.

AMH as a Marker of Ovarian Reserve

The term "ovarian reserve" describes the number and quality of remaining oocytes in the ovaries, while ovarian aging describes the age-related decline in that reserve. The number of resting primordial follicles left in the ovary is an important parameter, but it is difficult to measure directly. However, the size of the resting primordial follicle pool correlates with the number of follicles entering the growing pool. Since only growing follicles produce AMH, plasma AMH levels reflect the size of the resting primordial follicle pool.

Studies in mice and monkeys have shown a strong correlation between AMH levels and the number of primordial follicles. The role of AMH in predicting oocyte quality is still to be clarified, but the age-related decrease of the follicle pool is associated with a decrease in oocyte quality.

Other current ovarian reserve tests include hormonal markers (FSH, estradiol, and inhibin B) and ultrasonographic markers (antral follicle count and ovarian volume). These tests reflect, directly or indirectly, the size of the antral follicle pool. Antral follicle count is a direct ultrasound measure, while inhibin B and estradiol levels are considered dependent on the number of antral follicles. FSH levels are regulated by negative feedback from these granulosa cell products, making FSH a more indirect reflection.

Compared with these hormonal markers, plasma AMH levels better reflect the longitudinal decline in the oocyte/follicle pool over time, even before irregular cycles occur. In contrast to the cyclic fluctuations characteristic of FSH, estradiol, and inhibin B, AMH shows little or no fluctuation during the menstrual cycle. AMH levels are also relatively unaffected by conditions that suppress the later FSH-dependent stages of follicle development, such as pregnancy, use of hormonal contraceptives, and treatment with gonadotropin-releasing hormone agonists. Furthermore, AMH does not appear to be affected by clinical and behavioral variables such as body mass index and smoking status.

The predictive value of AMH for ovarian reserve is still uncertain, but several prospective longitudinal studies involving up to 11 years of follow-up of normo-ovulatory women have found that AMH seems to be the best endocrine marker for assessing ovarian aging and that plasma AMH levels, with reasonable accuracy, can predict the onset of menopause. Together, this data suggests plasma AMH can be used as a reliable marker of ovarian reserve.

AMH and In Vitro Fertilization (IVF)

Assessment of ovarian reserve is important before IVF treatment is initiated. The response to IVF treatment is highly variable, even among women of similar ages — a variability that undoubtedly reflects differences in ovarian reserve. Identifying both low and high responders prior to ovarian hyperstimulation allows doctors to optimize stimulation protocols, reducing cycle cancellation rates and severe complications such as ovarian hyperstimulation syndrome (OHSS).

An ideal ovarian reserve test would identify women with a practically zero chance of becoming pregnant due to diminished ovarian reserve. These women could be advised against undergoing treatment, thereby avoiding a disappointing outcome at high cost plus adverse effects. Conversely, false-positive tests (low AMH levels falsely predicting poor response) would incorrectly discourage women from treatment.

Predicting Poor Response to IVF

In selected literature defining poor response as an oocyte yield of four oocytes or fewer, AMH was found to be a predictor of poor response, with reported sensitivity of 72–97% and specificity of 41–93% at varying cut-off values. The varying cut-off values appear to be caused by variations in study methods and study populations, especially age, inclusion criteria, and the etiology of infertility.

Across 11 studies involving 2,315 patients total, the cut-off values for predicting poor response ranged from 1.43 to 14.0 pmol/L:

  • Muttukrishna et al. (2005): 108 patients, retrospective study, cut-off 1.43 pmol/L, sensitivity 87%, specificity 64%
  • Tremellen et al. (2005): 75 patients, prospective study, cut-off 8.10 pmol/L, sensitivity 80%, specificity 85%, PPV 67%, NPV 92%
  • LaMarca et al. (2007): 48 patients, prospective study, cut-off 5.36 pmol/L, sensitivity 80%, specificity 93%
  • McIlveen et al. (2007): 84 patients, prospective study, cut-off 8.93 pmol/L, sensitivity 85%, specificity 63%
  • Lekamge et al. (2007): 126 patients, retrospective study, cut-off 14.0 pmol/L, sensitivity 73%, specificity 73%
  • Gnoth et al. (2008): 132 patients, prospective study, cut-off 9.00 pmol/L, sensitivity 97%, specificity 41%, PPV 36%, NPV 98%
  • Riggs et al. (2008): 123 patients, retrospective study, cut-off 5.93 pmol/L, sensitivity 83%, specificity 79%, PPV 30%, NPV 97%
  • Barad et al. (2009): 76 patients, retrospective study, cut-off 3.57 pmol/L, sensitivity 87%, specificity 84%, PPV 79%, NPV 90%
  • Nardo et al. (2009): 165 patients, prospective study, cut-off 7.14 pmol/L, sensitivity 87%, specificity 67%
  • Al-Azemi et al. (2011): 352 patients, prospective study, cut-off 9.71 pmol/L, sensitivity 76%, specificity 75%
  • Honnma et al. (2011): 1,026 patients, retrospective study, cut-off 10.0 pmol/L, sensitivity 72%, specificity 76%

The positive predictive values (PPV) from these studies varied between 30% and 79%, meaning many false positives would be expected if low AMH alone were used to predict poor response. To prevent a high rate of false positives, extreme cut-off values must be used — but this would mean only a small percentage of abnormal tests would be found, and many poor responders would go unrecognized.

However, the negative predictive values (NPV) were notably higher, ranging from 90% to 98%. This implies that a negative test (i.e., a high AMH level) is less likely to be a false negative — meaning a woman with a high AMH level is very unlikely to be a poor responder.

Predicting Pregnancy Success with IVF

Prediction of live birth is the ultimate goal of any IVF screening test, but the application of AMH as a predictor of ongoing pregnancy following IVF appears to be limited in unselected infertile patients. Studies have found that a number of poor responders do achieve pregnancy. In particular, young poor responders have a better prognosis than older ones, and successful pregnancies have been reported even in cases of very low and even undetectable AMH levels.

A recent study found that patients with extremely low serum AMH levels have moderate but reasonable chances of achieving pregnancy and live births. There is reason to believe the ability of AMH to predict IVF pregnancy outcome varies greatly with age.

A retrospective study of 1,558 women who had undergone IVF treatment demonstrated that the clinical pregnancy rate for women younger than 34 years did not differ with AMH levels. At the same time, women aged 42 years or older had poor prognosis regardless of AMH levels. The authors concluded that serum AMH concentrations are associated with pregnancy only for women between the ages of 34 and 41 years.

This age-dependent finding may reflect different mechanisms behind depletion of ovarian reserve. In older women, the physiological aging of the oocytes (declining oocyte quality) may play an important role. In younger women, non-physiological mechanisms may dominate, such as congenital reduced ovarian reserve and faster consumption of primordial follicles without deterioration of oocyte quality.

Seven studies evaluated AMH's ability to predict pregnancy or live birth, with the following results:

  • Lekamge et al. (2007): 126 patients, retrospective, cut-off 14.0 pmol/L, sensitivity 50%, specificity 62%, cumulative clinical pregnancy
  • Elgindy et al. (2008): 29 patients, prospective, cut-off 19.28 pmol/L, sensitivity 83%, specificity 82%, PPV 77%, NPV 87%, clinical pregnancy
  • Lee et al. (2009): 123 patients, prospective, cut-off 12.0 pmol/L, sensitivity 60%, specificity 66%, PPV 49%, NPV 75%, live birth
  • Barad et al. (2009): 76 patients, retrospective, cut-off 7.14 pmol/L, sensitivity 62%, specificity 75%, PPV 48%, NPV 84%, clinical pregnancy
  • Li et al. (2010): 189 patients, retrospective, cut-off 13.0 pmol/L, sensitivity 77%, specificity 44%, cumulative live birth
  • Majumder et al. (2010): 97 patients, prospective, cut-off 19.3 pmol/L, sensitivity 66%, specificity 55%, PPV 31%, NPV 84%, live birth
  • Kini et al. (2010): 170 patients, retrospective, cut-off 5.0 pmol/L, sensitivity 86%, specificity 28%, cumulative clinical pregnancy

AMH, along with other endocrine markers, performs relatively poorly as a marker for withholding IVF treatment. It is not feasible to suggest that a woman should not undergo IVF treatment based solely on a low AMH value. However, in the absence of a perfect predictive marker, plasma AMH concentration could serve as a reference in pre-treatment counseling, along with other factors such as age.

Predicting Excessive Response and OHSS

AMH also appears useful in predicting ovarian hyperstimulation syndrome (OHSS) following gonadotropin stimulation. An excessive ovarian response to IVF treatment may lead to OHSS, a potentially life-threatening condition. Elevated mean basal AMH values are associated with the occurrence of an excessive response, and AMH measurements prior to gonadotropin stimulation could provide valuable information to help prevent OHSS.

Across seven studies, the reported sensitivity for predicting hyper-response varied between 69% and 93%, with specificity between 67% and 81% at varying cut-off values:

  • Nelson et al. (2007): 340 patients, prospective, cut-off 15.0 pmol/L, sensitivity 88%, specificity 77%, >21 oocytes
  • Lee et al. (2008): 262 patients, prospective, cut-off 24.0 pmol/L, sensitivity 91%, specificity 81%, PPV 30%, NPV 99%, moderate to severe OHSS
  • Riggs et al. (2008): 123 patients, retrospective, cut-off 11.4 pmol/L, sensitivity 84%, specificity 67%, PPV 22%, NPV 98%, >15 oocytes
  • Nardo et al. (2009): 165 patients, prospective, cut-off 25.0 pmol/L, sensitivity 88%, specificity 70%, >20 oocytes and/or estradiol >21,000 pmol/L
  • Alfatoonian et al. (2009): 159 patients, prospective, cut-off 34.5 pmol/L, sensitivity 93%, specificity 78%, PPV 65%, NPV 96%, >15 oocytes
  • Honnma et al. (2011): 1,026 patients, retrospective, cut-off 17.6 pmol/L, sensitivity 69%, specificity 75%, ≥20 oocytes
  • Ocal et al. (2011): 82 patients, case-control, cut-off 23.6 pmol/L, sensitivity 90%, specificity 71%, PPV 61%, NPV 94%, OHSS

The corresponding PPV and NPV range between 22–65% and 94–99%, respectively. AMH thus seems to be a fair predictor of excessive response and may improve patient counseling and permit individualization of treatment with the aim of reducing the incidence of OHSS.

AMH and Polycystic Ovary Syndrome (PCOS)

Polycystic ovary syndrome is a heterogeneous syndrome affecting 5–10% of the female population. The syndrome is characterized by at least two of the following three key components:

  • Oligo-/amenorrhea (infrequent or absent menstrual periods)
  • Clinical or biochemical hyperandrogenism (excess male hormones)
  • Polycystic ovary morphology on ultrasound

Women with PCOS also frequently present metabolic disturbances, including obesity, insulin resistance, and the metabolic syndrome. Women with PCOS are known to have an excessive amount of preantral and small antral follicles in the ovaries and to have increased plasma AMH concentrations.

The elevated AMH levels in PCOS are not only due to the excessive accumulation of preantral follicles, but also to increased granulosa cell AMH production. Since AMH plays a major inhibitory role during folliculogenesis, this elevated AMH production may contribute to the arrest of follicle development characteristic of PCOS. AMH measurement therefore has diagnostic potential in PCOS, and the degree of AMH elevation may correlate with syndrome severity.

AMH in Other Ovarian Conditions

Beyond ovarian reserve assessment and IVF, AMH has several additional clinical applications. In granulosa cell tumors, AMH serves as a valuable tumor marker. Because these rare ovarian tumors arise from granulosa cells — the very cells that produce AMH — elevated AMH levels can help detect the presence of these tumors and monitor for recurrence after surgical removal.

AMH also plays a role in the evaluation of premature ovarian failure (also called premature ovarian insufficiency). In this condition, ovarian function ceases before age 40. AMH levels can help identify women at risk for accelerated ovarian aging, potentially allowing earlier counseling about family planning options, including egg freezing or other fertility preservation methods.

Recent research has highlighted the use of AMH in a variety of ovarian pathological conditions, making it a versatile biomarker beyond its original role in assessing ovarian reserve.

Future Directions for AMH Testing

One of the major challenges in the clinical application of AMH measurements has been the lack of standardization between different assays. The development of the new commercial ELISA for measuring AMH levels — the AMH Gen II assay — makes results from different studies more comparable than before. However, widespread clinical application awaits an international standard for AMH, so that results using future assays can be reliably compared.

Whether individualized treatment strategies based on AMH values may increase oocyte yield remains unclear. Randomized clinical trials are needed to clarify whether different types and dosages of hormone stimulation influence oocyte yield in poor responders. So far, published results on this topic have been ambiguous.

Study Limitations

Several limitations must be considered when interpreting this review. First, hardly any randomized controlled trials were identified in the literature search — the evidence base consists largely of prospective and retrospective cohort studies, which carry inherent risks of bias. Second, the varying cut-off values for AMH across studies reflect differences in study methods and populations, making direct comparisons challenging.

Third, only sparse longitudinal prospective data exist on AMH's ability to predict reproductive lifespan. While AMH appears to be the best endocrine marker for assessing age-related decline of the ovarian pool, long-term studies with extended follow-up are needed to firmly establish its predictive power for menopause timing and reproductive longevity.

Fourth, the role of AMH in predicting oocyte quality — as distinct from oocyte quantity — is still to be clarified. AMH reflects the follicular pool size, but whether it can predict the chromosomal or developmental competence of oocytes remains an open question.

Finally, the finding that AMH's association with pregnancy is limited to women aged 34–41 years suggests that AMH cannot be used as a universal predictor across all age groups. Younger women with low AMH may still have reasonable pregnancy prospects, while older women with normal AMH may face age-related oocyte quality issues that AMH does not capture.

What This Means for Patients

For women considering fertility treatment or concerned about their reproductive timeline, AMH testing offers several practical benefits:

  1. Before IVF: An AMH blood test can help your doctor predict how your ovaries will respond to stimulation medications. A high AMH level suggests a lower risk of poor response, while a low level may prompt your doctor to adjust your treatment protocol.
  2. Preventing complications: Women with elevated AMH levels are at higher risk for ovarian hyperstimulation syndrome (OHSS). Knowing this in advance allows your doctor to use gentler stimulation protocols or lower medication doses to reduce that risk.
  3. Understanding PCOS: If you have polycystic ovary syndrome, AMH levels are typically elevated, reflecting the excessive accumulation of small follicles. AMH can help confirm the diagnosis and assess severity.
  4. Monitoring certain tumors: For women treated for granulosa cell tumors, AMH can be used as a tumor marker to monitor for recurrence.
  5. Planning for the future: AMH may help estimate your reproductive lifespan and predict when menopause might occur, which can inform family planning decisions — though it cannot predict whether you will be able to conceive naturally.

However, patients should understand that AMH is not a crystal ball. A low AMH level does not mean you cannot get pregnant — many women with low or even undetectable AMH levels achieve successful pregnancies, especially younger women. Similarly, a normal AMH level does not guarantee IVF success. AMH should be interpreted in the context of your age, medical history, and other fertility tests as part of comprehensive counseling with your fertility specialist.

If you are under 34 years old, pregnancy rates appear similar regardless of AMH levels. If you are 42 or older, your prognosis may be limited regardless of AMH. The most useful application of AMH is in the intermediate age range of 34–41 years, where it can provide meaningful prognostic information.

Frequently Asked Questions

What is an AMH blood test and what does it measure?

AMH stands for anti-Müllerian hormone, a blood test that helps doctors assess ovarian reserve, meaning the quantity and quality of remaining eggs. It is produced by granulosa cells from small ovarian follicles. AMH levels reflect the size of the resting primordial follicle pool and decline as women age, becoming undetectable at menopause.

How is AMH testing used before IVF treatment?

Before IVF, an AMH test can help predict whether your ovaries will respond poorly or excessively to stimulation medications. A high AMH level suggests a lower risk of poor response, while a low level may lead your doctor to adjust your treatment protocol. It can also help identify women at higher risk of ovarian hyperstimulation syndrome (OHSS).

Can a low AMH level mean I cannot get pregnant?

No. A low AMH level does not mean you cannot get pregnant. Many women with low or even undetectable AMH levels achieve successful pregnancies, especially younger women. AMH is not a crystal ball and should be interpreted with your age, medical history, and other fertility tests. In women under 34, pregnancy rates appear similar regardless of AMH levels.

What is ovarian hyperstimulation syndrome (OHSS) and how does AMH help prevent it?

OHSS is a potentially life-threatening complication of IVF that can occur when the ovaries respond excessively to gonadotropin stimulation. Elevated AMH levels before stimulation are associated with a higher risk of excessive response. Measuring AMH beforehand allows your doctor to use gentler stimulation protocols or lower medication doses to reduce the chance of OHSS.

Is AMH testing useful for diagnosing polycystic ovary syndrome (PCOS)?

Yes. Women with PCOS typically have elevated AMH levels, reflecting an excessive number of preantral and small antral follicles, as well as increased AMH production by granulosa cells. AMH measurement has diagnostic potential for PCOS, and the degree of elevation may correlate with syndrome severity.

Can AMH predict when I will reach menopause?

Some prospective studies with up to 11 years of follow-up suggest that plasma AMH levels can, with reasonable accuracy, predict the onset of menopause. AMH appears to be the most effective endocrine marker for assessing ovarian aging. However, long-term studies with extended follow-up are still needed to firmly establish its predictive power for menopause timing.

There are different AMH test kits with different results. How can I compare my value?

Two commercial ELISAs have been used, and values from the DSL kit are about four times lower than from the Immunotech kit. No international reference standard exists yet, and results may be in ng/mL or pmol/L (1 ng/mL = 7.14 pmol/L). A newer Gen II assay is replacing older tests, which should improve comparability across studies.

Source Information

This patient-friendly article is based on the following peer-reviewed research:

Original Title: The role of anti-Müllerian hormone in female fertility and infertility – an overview

Authors: Anna Garcia-Alix Grynnerup, Anette Lindhard, and Steen Sørensen

Journal: Acta Obstetricia et Gynecologica Scandinavica, 2012; 91:1252–1260

Affiliations: Fertility and Endocrinology Unit, Department of Gynecology and Obstetrics, Roskilde Hospital, University of Copenhagen, Roskilde, Denmark; and Department of Clinical Biochemistry, Hvidovre Hospital, University of Copenhagen, Hvidovre, Denmark

DOI: 10.1111/j.1600-0412.2012.01471.x

Received: 28 October 2011; Accepted: 10 May 2012

The authors stated explicitly that there are no conflicts of interest in connection with this article. This patient-friendly translation is intended for educational purposes and does not constitute medical advice. Always consult a qualified healthcare provider regarding your individual fertility situation.