Dehydroepiandrosterone (DHEA) supplementation in women with low functional ovarian reserve

Conflict statement

We want to point out to readers of this article that The Center for Human Reproduction (CHR) holds a number of patents, which claim therapeutic benefits from androgen/DHEA supplementation of selected infertile women. CHR and some CHR employees, including this author, receive royalty payments from two companies, which have licensed these patents and, therefore, are entitled to use these claims in their marketing efforts. Moreover, Dr. Gleicher is a shareholder in one of these companies, Fertility Nutraceuticals, LLC.


DHEA is the most abundant steroid hormone in our bodies, and is the precursor hormone for all sex hormones. It, in itself, has very low affinity to the androgen receptor (AR) (and mild affinity to the estrogen receptor) and, therefore, is functionally almost inert as an androgen. Moreover, DHEA within minutes from ingestion is metabolized to DHEAS. Both, DHEA and DHEAS, as storage forms, usually are to a degree in balance in the circulation, constantly changing back and forth. Increasing evidence suggests that this allows different organs to take up sufficient DHEA to produce organ-specific testosterone (T)-levels. It appears that, aside from ovaries and adrenals, other organs, for example, liver, endometrium, etc. also have the apparatus to produce T from DHEA. The overwhelming amount of T in the woman comes, however, almost evenly, from ovaries and adrenals.

The apparatus for adding a sulfate group to DHEA to produce DHEAS, however, exists almost exclusively only in the zona reticularis of the adrenals. This is diagnostically important, and is also important when patient under DHEA supplementation are clinically monitored: Since DHEA and DHEAS should be in reasonable balance, DHEA supplementation should result in quick increases in DHEA as well as DHEAS.

If only DHEA rises, the patient likely suffers from deficiency of sulfotransferase enzymes (SULT1A1 and SULT1E1) in the zona reticularis, and DHEA supplementation will, likely, not lead to adequate increases in T. Such patients then need to be directly supplemented with T.

It is also important to note that the zona reticularis shrinks with advancing age and, therefore, DHEAS levels decline as women (and men) are getting older.

DHEAS levels, thus, in a way reflect on adrenal function. They, therefore, also allow conclusions whether observed androgen abnormalities in a woman are more likely ovarian or adrenal in origin (they, of course can also be caused by a combination of both). So, for example, whether hyper-androgenism (high-T) in a woman with polycystic ovary syndrome (PCOS) is primarily ovarian or adrenal in origin, can be deducted from her DHEAS levels. If they are abnormally high, hyper-androgenism (high-T), is likely, adrenal. If T is high but her DHEAS is normal or even low, then high-T with great probability is of ovarian origin.

In PCOS, these associations have been well recognized for decades. In low functional ovarian reserve (LFOR) patients, whether due to advanced age or premature ovarian aging (POA), also called occult primary ovarian insufficiency (oPOI), which almost uniformly is characterized by low-T, the same principles apply. Based on DHEAS levels, LFOR patients’ likely source of low-T can be equally well determined: If their low-T is associated with low-DHEAS, the source is, likely, adrenal insufficiency of the zona reticularis. If DHEAS in such patients is, however, normal or even high, then low-T is very likely of ovarian origin.

To be able to make this differentiation has significant prognostic relevance because, if low-T is of ovarian origin, it suggests that a patient’s ovaries are, likely, already to a significant degree insufficient. DHEA supplementation in such cases will, therefore, be only marginally effective in improving ovarian function. If low-T, however, is caused by adrenal androgen insufficiency, the likelihood that a patient’s ovaries will still recover function after androgen supplementation has raised intra-ovarian T-levels appropriately, is far better. In these cases, the observed LFOR is not a primary ovarian insufficiency (POI) but a secondary ovarian insufficiency (SOI), actually caused by adrenal insufficiency of androgen production.

Also important for the evaluation of androgen levels in women with LFOR is the inverse relationship between sex hormone binding globulin (SHBG) and T. Though SHBG can also be affected by other hormones, low-T usually is associated with high-SHBG. Evaluation of SHBG, therefore, helps in determining whether a patient is within her normal T range or not. Since normal T ranges vary (for example, the AR of women, who in their younger years were PCOS patients, may still have memory for higher-T levels than in non-PCOS patients), it is important to observe declining SHBG levels in hypo-androgenic women who are being supplemented with DHEA (or T).

How do we know that DHEA works in women with LFOR?

Our group introduced DHEA supplementation into reproductive medicine approximately 12 years ago after an almost 43- year old patient in consecutive cycles demonstrated surprising gains in oocyte and embryo numbers. After six such cycles, she acknowledged she had started supplementing with DHEA after reviewing the medical literature, and finding an article that suggested that DHEA may improve egg yields (she had produced only 1 egg in her first cycle at our center). She continued 12 consecutive IVF cycles, and in her last two developed PCOS-like ovaries (at age 43), requiring a reduction in her gonadotropin stimulation dosage.

This patient’s experience led our center into the investigation of DHEA supplementation in women with LFOR. We at that point had practically no understanding why DHEA may work, and others had suggested that its effect may be mediated by IGF1. While we attempted two prospectively randomized studies (RCTs), one at our center and a second in collaboration with European colleagues, both failed in recruiting enough LFOR patients who were willing to be randomized to placebo. For the same reason, till today not a single well-powered RCT has been published. It, indeed, appears unlikely that such a study will ever be published since most LFOR patients, rightly, are hesitant in wasting six or more months on potential placebo treatment, fully recognizing the importance of time in association with LFOR.

Fortunately, shortly after we had to abandon those RCTs, basic science colleagues started elucidating the importance of androgens on follicle maturation in a number of animal models. From these studies it became abundantly clear why DHEA supplementation, indeed, improved ovarian function clinically, as we had reported in a number of prospective clinical studies with lower evidence levels than offered by RCTs. These animal models also allowed the direct investigation of DHEA supplementation on ovaries, and uniformly confirmed that androgen supplementation, indeed, improved functional ovarian reserve (FOR), as evaluated by anti-Müllerian hormone (AMH).

FOR represents the small growing follicle pool immediately after recruitment of primordial follicles between primary and small pre-antral follicle stages. Uniformly, these studies demonstrated that during these early growth stages of follicles T, via the androgen receptor (AR) on granulosa cells (GCs), acts synergistically with FSH on the follicles. When the AR on GCs is blocked, fewer follicles mature, and resulting oocytes (and, therefore, embryos are of poor quality). Models also demonstrated that, if low-T is supplemented with DHEA or T, folliculogenesis improves, exactly demonstrating the benefits we had reported in our prospective clinical studies of DHEA supplementation in women with LFOR (Table 1).

Table 1. Clinical effects of androgen supplementation

  • Improved AMH in ca. 60% of women
  • No effects on FSH
  • Improved oocyte yields/quality
  • Improved embryo yields/quality
  • Improved spontaneous pregnancy rates
  • Improved IVF pregnancy rates
  • Improved cumulative pregnancy rates
  • Shortened time to pregnancy
  • Decreased embryo aneuploidy
  • Decreased miscarriage rates after age 35


It is important to reemphasize that these beneficial effects of androgen supplementation are on the small growing follicle pool. This means that the follicles that benefit from DHEA supplementation still undergo further maturation for at least 6-8 weeks before they enter the gonadotropin-sensitive last two weeks of folliculogenesis and become available to stimulation by gonadotropins in IVF. Women with LFOR, therefore, need to be pretreated with DHEA for at least six to eight weeks before IVF cycle start.

We reached this conclusion by prospectively observing IVF cycle outcomes even before we understood how DHEA achieved its effects on women with LFOR. In those early studies we, therefore, concluded that a first statistically significant improvement in clinical pregnancy rates was observed after six weeks of stimulation but that chances cumulatively improved up to 4-5 months, at which point a plateau was reached. Once our understanding about the effects of androgen supplementation evolved, these observed outcomes not only made sense, considering the time frames of follicle maturation, but also corresponded well with above described observations of T levels, DHEA, DHEAS and SHBG. In recent years we, therefore, still maintain a minimum period of six weeks of DHEA supplementation before IVF cycle start but, in addition, monitor above noted hormones, and will not start the cycle unless T and SHBG levels are in balance.

How do we supplement with DHEA?

Neither we nor other investigators have performed dose-dependency studies on DHEA supplementation. We started supplementing patients with 25mg of micronized DHEA TID (total 75mg daily) because our above noted index patient used that amount of supplementation without any apparent side effects. She, in turn, came to this dosage from the one manuscript in the literature that gave her the impetus to initiate DHEA supplementation. The study reported in that manuscript utilized this dosage because it slightly exceeded the body’s average daily DHEA production. Overall, this is a very low daily dosage of DHEA in comparison to other reported uses of tis hormone, especially when it is used and abused in body building and other sports.

In the U.S., for difficult to explain reasons, DHEA is considered a “food supplement,” and, therefore, available over-the-counter without prescription. It, therefore, is produced by a large number of companies and freely available at relatively low cost. As studies have demonstrated, the quality of individual products varies greatly, with some even containing no active DHEA. Our center, therefore, a number of years ago started collaborating with a company called Fertility Nutraceuticals LLC in setting strict quality criteria, which were equal to the DHEA our patients were prescribed during our studies, which at that time was compounded by a local pharmacy.

In repeat full disclosure, it is important for readers to understand that our center holds a number of patents for the clinical use of DHEA (and other androgens) in infertile women. Indeed, all DHEA products that wish to claim fertility benefits from DHEA/androgen supplementation, therefore, have to obtain licenses from our center. At the current time only two companies have been licensed since they fulfill quality assurance requirements, and, therefore, can make infertility-related claims for their products. Our center almost exclusively uses the DHEA products produced by Fertility Nutraceuticals LLC (

Again in full disclosure, or center and a number of our investigators, as well as our index patient, are co-inventors on above cited patents and, therefore, receive royalties from license agreements. Moreover, some of the center’s investigators are shareholders in Fertility Nutraceuticals LLC.

Once patients start supplementation, they continue uninterrupted through IVF cycles until second normally rising pregnancy test. DHEA is only discontinued if a patient conceives or decides to end treatments with use of her own eggs. As noted earlier, if we do not see satisfactory increases in DHEAS and T within six weeks, we either add or switch to direct T-supplementation. Desired T-levels should reach at least approximately mid-point for the laboratory’s normal range, and preferably be even a little above mid-point. SHBG should be under 100.0nmol/L, and preferably under 80.0nmol/L.

Side effects of supplementation

DHEA at given dosages is extremely well tolerated. Rare side effects reported are oily skin, acne and hair loss. In general, patients feel more vigorous and report better sex drives. Side effects are reversible with either lower dose supplementation or cessation of supplementation. In general, side effects with DHEA supplementation are much less profound than with T-supplementation. The main reason is that with T-supplementation all organs are exposed to similar T-levels, while with DHEA supplementation, as already previously noted, each organ takes only as much DHEA substrate up as it needs.

Further recommendations

The evaluation of androgens is now routine at our center in all patients because we recently described in the literature a new, previously unknown PCOS phenotype, which in contrast to most PCOS phenotypes, is not characterized by high-T but by low-T. In detailed investigations we determined that these patients’ PCOS phenotype has a very specific age-dependent ontogeny and clinical presentation. Moreover, it appears more treatment resistant than classical PCOS since it is more common in tertiary fertility centers like ours than the classically obese phenotype

This phenotype is characterized by being lean (in contrast to classical PCOS, characterized by truncal obesity), yet is still hyper-androgenic up to the patients’ mid-20s to early 30s, and is characterized by typically high AMH values for all PCOS patients. Between their mid-20 to early 30s, so affected women, quite suddenly, drop androgen levels much quicker than AMH as they advance in age (classical PCOS drops both more less in parallel). The presenting clinical finding in these patients is, therefore, often disproportionally high AMH levels for patient age and/or FSH levels. On further investigation they demonstrate low-T and especially low-DHEAS. Therefore, it is obvious that their sudden hypo-androgenism is adrenal in nature. Indeed, there is convincing evidence that the adrenal insult leading to low-T is, likely, autoimmune in nature since many of these women demonstrate other autoimmune abnormalities, especially anti-thyroid autoimmunity.

Once this PCOS phenotype is recognized by above noted characteristic findings, we were surprised by how frequently we made the diagnosis of hypo-androgenic PCOS in our center’s patient population. Making this diagnosis is very rewarding because DHEA supplementation in these patients is very successful in “rejuvenating” their ovaries. They, typically, are patients, who in prior IVF cycles produced relatively large egg numbers but of poor quality and, therefore, generated poor quality embryos. With proper androgen supplementation, their egg and embryo quality improves and, since they still even at older ages produce good egg and embryo numbers, they achieve surprisingly good pregnancy and live birth rates.

Summary and conclusions

We hope that this brief summary offers a guide to correct androgen supplementation of infertile women, and explains the rational for such treatment as well as its methodology. Androgen supplementation in our opinion is essential in all women with low-T (and/or high SHBG), and in all women above age 40. Androgen supplementation, however, also has to be followed by appropriate ovarian stimulation, which has to avoid anything that is potentially suppressive on ovaries (i.e. oral contraceptives, antagonist, etc.). We also recommend high dosage gonadotropin stimulation with FSH preponderance but at least 150IU of hMG daily since these patient need LH.

After female age, number of available oocytes and especially transferrable embryos is the most important statistical predictor of pregnancy and live birth chances. We, therefore, strongly recommend against mild stimulation protocols in women with LFOR, against culture of embryos to blastocyst stage (cleavage stage embryos give greater cumulative pregnancy chances) and, of course, against PGS, which we consider a worthless procedure, and especially harmful in women with LFOR, since it results in the discarding of perfectly normal embryos.