DHEA is the most abundant hormone in your blood. That is not a reason to swallow more of it.

What DHEA actually is

DHEA stands for dehydroepiandrosterone. It is a steroid hormone made mainly in the zona reticularis — the innermost layer of the adrenal cortex, the gland that sits on top of each kidney — with smaller contributions from the gonads and the brain. Most of it circulates in a sulfated storage form called DHEAS (dehydroepiandrosterone sulfate), which the body can convert back and forth. Between them, DHEA and DHEAS are the most abundant steroid hormones in human circulation.¹¹

Here is the part that gets glossed over on the label. DHEA is a precursor, not an end product. On its own it does comparatively little. Its influence comes almost entirely from what peripheral tissues convert it into, and the conversion runs in two directions: downstream into androgens such as testosterone, and downstream into estrogens such as estradiol.¹¹ That branching is the single most important fact about this molecule, and almost every honest question about DHEA reduces to it. You are not taking a hormone. You are handing your tissues raw material and letting local enzyme activity — which varies by sex, by tissue, by age, by body fat — decide what signal it pulls.

This is what makes DHEA categorically different from a vitamin, and different again from testosterone therapy. With enclomiphene or hCG you are acting on a specific point in the hormonal axis with a predictable direction of travel. With DHEA you are raising the substrate pool and accepting whatever the downstream machinery does with it — which, in a postmenopausal woman with adipose aromatase activity, is not the same thing it does in a 45-year-old man.

Legal in one country, controlled in the next

Before any discussion of efficacy, there is a jurisdictional fact that most coverage of DHEA skips entirely, and it matters enormously if you are reading this from Canada.

In the United States, DHEA is sold freely over the counter as a dietary supplement. That is not an accident of enforcement — it is written into statute. When Congress passed the Anabolic Steroid Control Act of 2004 and swept dozens of prohormones into the controlled-substances schedule, DHEA was explicitly carved out of the definition of an anabolic steroid.¹³ It remains a supplement, regulated under the far looser dietary-supplement framework rather than as a drug.

In Canada, the same molecule appears by its pharmaceutical name, prasterone, as item 36 on Schedule IV of the Controlled Drugs and Substances Act.¹² It is a controlled substance. There is no legal over-the-counter DHEA supplement in Canada; the legitimate route is a prescription product, and Health Canada has authorized the intravaginal prasterone insert through the ordinary drug-review process. If you are Canadian and you have DHEA capsules in your cupboard, they were almost certainly imported from a jurisdiction where the rules are different.

In sport, the question is settled and has been for two decades. DHEA is listed by the World Anti-Doping Agency under class S1, Anabolic Agents, and is prohibited at all times — in and out of competition.¹⁴ Any tested athlete who takes it, deliberately or through a contaminated product, is looking at an adverse analytical finding.

The age-decline pitch, and what it does not prove

The commercial case for DHEA is one clean graph. Circulating DHEAS peaks in the early twenties and then falls steadily — roughly a couple of percent per year — so that by the seventh or eighth decade a person is carrying a small fraction of their young-adult level. The phenomenon even has a name, adrenopause. The pitch writes itself: something that high in youth and that low in age must matter, so put it back.

The logic has a hole in it, and it is the same hole that has swallowed a long line of hormone-replacement enthusiasms. Observing that a hormone falls with age tells you nothing about whether restoring it does anything useful. The decline could be a driver of aging. It could equally be a downstream marker of it, or an adaptive change, or simply irrelevant. The only way to find out is to put the level back in a randomized trial and measure whether anything the person cares about actually changes.

That trial exists, and it is the most important study in this entire literature. In 2006 a Mayo Clinic group published a two-year, placebo-controlled, double-blind randomized trial in the New England Journal of Medicine: 87 elderly men and 57 elderly women, all selected specifically because they had low DHEAS.² DHEA supplementation worked exactly as advertised on the biochemistry — median DHEAS rose by 3.4 µg/mL in men and 3.8 µg/mL in women versus placebo. The levels were restored.

Then they measured whether it mattered. No significant effect on body composition in either sex. No change in peak oxygen consumption, muscle strength, or insulin sensitivity. No improvement in quality of life. The authors’ conclusion is one sentence and it has aged well: neither DHEA nor low-dose testosterone replacement in elderly people has physiologically relevant beneficial effects on body composition, physical performance, insulin sensitivity, or quality of life.²

The best-designed test of the entire anti-aging premise took the exact population the marketing describes, restored their levels for two years, and found nothing worth having.

A 2013 meta-analysis of 25 placebo-controlled trials in 1,353 elderly men appeared, at first glance, to offer a rescue: DHEA was associated with a small reduction in fat mass, a standardized effect of −0.35.³ But the same paper dismantles its own finding. In multivariate regression, the association with fat mass disappeared after adjusting for the rise in testosterone and estradiol. In other words, whatever small body-composition signal DHEA pulls is not DHEA doing anything — it is the sex hormones it converts into, arriving by an indirect and poorly controlled route. And on every other outcome the analysis examined — lipids, glucose metabolism, bone, sexual function, quality of life — DHEA did nothing at all versus placebo.

The evidence, sorted by what you would take it for

This is where averaging becomes dishonest. DHEA is not one intervention with one grade. It is at least five different propositions wearing the same name, and they do not perform alike.

Intravaginal prasterone for genitourinary syndrome of menopause — the one that holds

Genitourinary syndrome of menopause (GSM) is the modern term for the vaginal dryness, tissue thinning and painful intercourse that follow the estrogen loss of menopause. It is common, it is under-treated, and women rarely bring it up unprompted.

Delivered locally as a vaginal insert — the pharmaceutical form is called prasterone — DHEA has a real evidence base here. A 2026 systematic review and meta-analysis pooled six reports representing five unique randomized controlled trials in 1,611 postmenopausal women. Against placebo, intravaginal DHEA significantly improved both primary outcomes: vaginal dryness, with a mean difference of −0.23 (95% CI −0.35 to −0.11), and dyspareunia — painful sex — at −0.40 (95% CI −0.66 to −0.15). Heterogeneity was low to moderate, adverse effects were mild and infrequent, and no major safety signal emerged.¹

That is a genuinely respectable body of evidence, and it is why this claim grades MODERATE rather than lower. But look at the numbers rather than the p-values before you decide what MODERATE means. Those mean differences are on symptom severity scales that typically run 0 to 3. A shift of 0.23 for dryness is real, consistent, and small. This is a treatment that helps — not a transformation. It also grades MODERATE rather than STRONG because five trials, several sharing sponsorship lineage, is a narrower base than the phrase “regulator-approved” suggests to most readers.

The critical qualifier: this route is not the capsule route. A vaginal insert delivering DHEA into local tissue, where it is converted to estrogens and androgens exactly where they are needed, is a different intervention from a 25 mg tablet passing through your liver. Evidence for one is not evidence for the other, and the supplement industry has been very comfortable letting that distinction blur.

Adrenal insufficiency — the theoretically strongest case that still underperforms

People with adrenal insufficiency — Addison’s disease and its secondary forms — are the one population with an unambiguous physiological rationale. Their adrenal glands genuinely cannot produce DHEA, standard replacement therapy with glucocorticoids and mineralocorticoids does not replace it, and many report persistently poor wellbeing despite otherwise adequate treatment. If DHEA replacement works anywhere systemically, it should work here.

A meta-analysis of ten randomized placebo-controlled trials in women with primary or secondary adrenal insufficiency found a statistically significant improvement in health-related quality of life — with an effect size of 0.21 (95% CI 0.08 to 0.33).⁶ There was a small benefit for depression. Effects on anxiety and sexual wellbeing were small and not statistically significant.

The authors’ own summary is unusually blunt and worth quoting in shape if not in full: DHEA may improve quality of life and depression in a small and perhaps trivial manner, and the evidence appears insufficient to support routine use.⁶ An effect size of 0.21 is at the boundary where a difference becomes hard to notice in a single person’s life. That is why this grades WEAK despite ten randomized trials behind it — the finding is statistically real and clinically close to nothing. It remains a reasonable individualized conversation between a patient and an endocrinologist. It is not a result the supplement aisle has earned the right to borrow.

Fertility and IVF — where the intermediate markers move and the outcome does not

DHEA priming before in-vitro fertilization is standard practice in a number of clinics for women with poor ovarian response or diminished ovarian reserve. A 2023 meta-analysis retrieved 32 studies — but crucially reported the randomized trials separately from the uncontrolled ones, which is exactly the discipline this question needs.⁷

In the randomized-trial subgroup, DHEA moved several intermediate measures in a favourable direction: antral follicle count rose, baseline FSH fell, the required gonadotropin dose dropped by roughly 382 international units, stimulation was shorter by about a day, and miscarriage rate fell (relative risk 0.46, 95% CI 0.29 to 0.73). Encouraging — and then the endpoints that matter. Among randomized trials only, there was no significant difference in clinical pregnancy rate and no significant difference in live birth rate. Higher pregnancy and live-birth rates did appear in the non-randomized studies, and the authors explicitly warn that those should be interpreted with caution because of potential bias.⁷

The single largest randomized trial on the question settles it more bluntly. Across nine reproductive centres in China, 821 women meeting the Bologna criteria for poor ovarian response were randomized double-blind to DHEA or placebo for four to twelve weeks before transfer. Live birth after the first embryo transfer: 8.8% on DHEA, 9.0% on placebo — relative risk 0.98 (95% CI 0.63 to 1.51, p = 0.911). No difference in oocytes retrieved, clinical pregnancy, pregnancy loss, or cumulative live births either.¹⁵

That gap — between what looks promising in uncontrolled clinic data and what survives randomization — is the whole story of DHEA in fertility medicine, and it is why the claim grades WEAK. The miscarriage signal is interesting and deserves a properly powered trial of its own. The live-birth claim, which is the only one a patient actually cares about, is not currently supported.

Cognition — the claim that goes the wrong way

DHEA is a neurosteroid, brain tissue makes it, levels fall with age, and dementia is frightening. The inferential leap builds itself. It has not held up.

A Cochrane review of randomized placebo-controlled trials in people over 50 without dementia found only three studies with adequate parallel-group data, and could not even pool them for lack of reported statistics. One found no significant effect over three months in perimenopausal women. One found no effect over three months in 46 older men. The third — and this is the detail that never makes it into a product page — found that under a psychosocial stressor, DHEA significantly impaired performance on a visual memory recall test compared with placebo. The review’s conclusion is that the available controlled trials do not support a beneficial effect of DHEA on cognitive function.⁸

A 2023 systematic review in Menopause went looking specifically at postmenopausal women — the group most heavily marketed to on this claim — and found only four randomized trials that qualified, all using 50 mg oral DHEA daily against an identical placebo. One positive result surfaced: a four-week crossover study in 24 cognitively normal women where DHEA improved five of six visual-spatial tests. No other study showed any cognitive improvement, and the reviewers rated the trials at high risk of bias across multiple domains. Their conclusion: the evidence does not support a beneficial effect of DHEA on cognitive performance in postmenopausal women.¹⁶

Seventeen years separate those two reviews and the verdict has not moved, which is itself informative about how seriously the field takes the hypothesis. This grades WEAK, and the honest framing is that the single most-cited cognitive finding in the controlled literature is a decrement, not an improvement.

What it does in women specifically

Most writing about DHEA treats women as an afterthought appended to a male default. That is backwards. Women are the population in whom oral DHEA demonstrably does the most — and therefore the population carrying the most risk.

A 2025 meta-analysis of 21 randomized trials in postmenopausal women is the cleanest data on this point. DHEA supplementation raised estradiol by a weighted mean difference of 7.86 pg/mL (95% CI 6.33 to 9.40) and testosterone by 24.31 ng/dL (95% CI 15.22 to 33.40), both highly significant. Effects were larger at doses of 50 mg per day or more, where the testosterone rise reached 29.65 ng/dL, and the estradiol rise was largest in women aged 60 and over.⁵ That is the STRONG grade in the card above — and note carefully what it is a grade for. It is a biomarker claim. DHEA unambiguously moves hormone levels in postmenopausal women. Whether moving them improves anything is the separate question the rest of this article answers, mostly in the negative.

A pooled analysis of four randomized trials in 295 women and 290 men aged 55 and over sharpens the sex difference into something clinically useful. Women on DHEA for twelve months saw significant rises in DHEAS, testosterone, estradiol and IGF-1, plus small gains in lumbar-spine and trochanter bone density and maintenance of total hip bone density. Men saw rises in DHEAS, a smaller estradiol rise, a marginal IGF-1 change — and no bone-density benefit at all.⁴ The authors’ framing is that DHEA may be an approach worth studying for preserving bone and muscle in women. In men, the same intervention largely does not land.

Read that against the female hormonal picture and the practical implications follow directly. Because oral DHEA raises both estradiol and testosterone, it is not a neutral addition for a woman with a hormone-sensitive condition — a history of breast or endometrial cancer, or endometriosis, sits squarely in the path of an unmonitored rise in circulating estrogen. Because the androgen arm is real, the androgenic complaints are real too: acne, oilier skin, unwanted facial or body hair, and scalp hair thinning are the predictable consequences of pushing a woman’s testosterone up by 25 to 30 ng/dL without measuring where she started. In a premenopausal woman still cycling, that androgen push also has the potential to disturb the LH and FSH signalling that governs ovulation, and the luteal phase — already the phase where progesterone-sensitive women feel most fragile — is the worst window in which to add an unpredictable androgen load. Almost none of the trial data above was generated in premenopausal women, which means the group most likely to buy DHEA off a shelf is the group with the least evidence supporting it.

If you want the map of what a woman’s hormonal decline actually looks like and which levers have data behind them, our reads on testosterone in women and black cohosh for menopause symptoms cover adjacent ground with the same grading discipline.

And what it does in men

For men the summary is shorter, and it is not flattering to the marketing.

Oral DHEA in men raises DHEAS reliably and estradiol modestly. It does not reliably raise testosterone to any degree that changes anything, it produces no bone-density benefit, and across 25 placebo-controlled trials it showed no effect on lipids, glucose metabolism, bone, sexual function, or quality of life.³⁴ The small fat-mass reduction that did appear was fully explained by conversion into testosterone and estradiol, which means a man taking DHEA for body composition is running an inefficient, unmonitored and slightly randomized version of hormone therapy.

That is the core problem with DHEA as a male testosterone strategy: the one downstream product men want is the one the conversion machinery delivers least. Men have functioning testes producing testosterone in far greater quantity than peripheral conversion of an oral precursor can add. What the precursor does reliably add is estradiol — the direction most men taking it are hoping to avoid. If the goal is genuinely to move testosterone, the evidence-based conversations are about the hypothalamic-pituitary-gonadal axis and about what is actually driving the decline in the first place, not about a precursor that arrives at the wrong destination.

Not a benign supplement

The safety case for DHEA is usually made by pointing out that trials report few serious adverse events. That is true, and it is a low bar — those trials are mostly short, mostly small, and mostly not designed to detect the things that would matter over a decade.

The clearest quantified harm signal is metabolic. A dose-response meta-analysis of 23 randomized controlled trials found that DHEA did not change total cholesterol, LDL cholesterol or triglycerides — but it significantly reduced HDL cholesterol, by a weighted mean of 3.1 mg/dL overall. The subgroup breakdown is where it gets pointed: in studies of women, HDL fell by 5.1 mg/dL (95% CI −7.2 to −3.0). In men, it did not move at all (0.13 mg/dL).⁹ That is a MODERATE-grade finding across 23 randomized trials, it is consistent with what oral androgens do to HDL generally, and it lands precisely on the population in whom DHEA is most often recommended. Whether a 5 mg/dL HDL drop translates into cardiovascular events over years is genuinely unknown — the authors call for exactly that trial. It is not a reassuring unknown.

Beyond lipids, the risk profile follows straight from the mechanism. Anything that raises circulating estrogen is a live question in hormone-sensitive cancers. Anything that raises androgens carries dermatological and hair consequences in women and, in theory, prostate implications in men that no trial has been long enough to settle. DHEA also interacts with the reason people are often taking it: someone with adrenal insufficiency is on glucocorticoid replacement, and adding an adrenal steroid precursor to that regimen is a clinician’s decision, not a shelf decision. If you want to see how any of this is actually landing in your own body, the answer is a hormone and lipid panel before and after — our lab interpreter will tell you what the numbers mean, and our labs reference covers which ones to ask for.

The label problem

There is one more issue specific to DHEA, and it predates most of the current supplement market. A quality-control analysis published in JAMA tested commercially available DHEA dietary-supplement products against their labelled content and found the actual amounts diverged substantially from what the labels claimed — in both directions, including products containing far less active compound than stated.¹⁰

That study is old, and it would be unfair to present a 1998 letter as a description of today’s market. But two things make it worth flagging anyway. First, nothing in the regulatory structure has changed to fix it: DHEA is still sold in the United States under the dietary-supplement framework, which does not require pre-market content verification. Second, the consequence of label inaccuracy is much larger for a hormone precursor than for a vitamin. Getting three times the labelled dose of vitamin C is a non-event. Getting three times the labelled dose of a compound that raises your estradiol and testosterone is not, and neither is getting a third of it while assuming you are running a studied dose. If you use DHEA under clinical supervision, third-party batch testing is not a nicety.

Open questions

Four gaps are worth naming precisely, because they define what would actually change these grades. First, the miscarriage signal in fertility is unexplained and untested on its own — a relative risk of 0.46 in the randomized subgroup is the single most interesting orphan finding in this literature, and it has never been the primary endpoint of an adequately powered trial.⁷ Second, the female bone signal deserves a proper long-duration trial: twelve months of DHEA maintained or slightly improved bone density in women but not men, and nobody has run it long enough to see whether that translates into fractures avoided.⁴ Third, the HDL question is a genuine safety unknown, not a resolved one — a consistent 5 mg/dL drop in women across 23 trials needs a cardiovascular-outcome study, and the meta-analysis authors say so explicitly.⁹ Fourth, premenopausal women are almost entirely unstudied, which is remarkable given how much DHEA is sold to women still cycling. Every claim in this article about women rests on postmenopausal or adrenal-insufficient cohorts.

The verdict

DHEA is a real hormone with a real pharmacology and one legitimate, regulator-approved application. Delivered as an intravaginal insert for the dryness and painful sex of menopause, it works — modestly, safely, and with five randomized trials behind it.¹ If that is your problem, this is a conversation worth having with a gynecologist, and in Canada it is the only lawful route anyway.

Everything else the bottle implies is either unproven or actively contradicted. The anti-aging premise was tested properly, over two years, in exactly the low-DHEAS population the pitch describes, and it produced nothing of physiological relevance.² The adrenal-insufficiency benefit is statistically real and clinically close to trivial.⁶ The fertility benefit evaporates at the endpoint patients care about once you restrict to randomized data.⁷ The cognitive benefit does not exist, and the best-documented cognitive finding in the controlled literature is an impairment.⁸ The claim that DHEA is a benign supplement that safely restores youthful hormones grades HYPE, and it grades HYPE on three separate counts: the restoration does not deliver benefits, the safety profile includes a consistent HDL reduction in women, and the product itself is a controlled substance in Canada and a doping violation in sport.⁹¹²¹⁴

The honest way to hold DHEA is as a prescription-grade intervention that happens to be sold like a vitamin in one country. It converts into both androgens and estrogens, it moves measurable hormone levels in women more than in men, it lowers HDL in women, and it interacts with cancers, cycles and adrenal replacement regimens in ways that require someone monitoring labs. Nobody should be self-prescribing it off an article — including this one. If DHEA belongs in your protocol, it belongs there with a clinician, a baseline panel, a follow-up panel, and a specific indication that is on the short list above rather than the long list below it.

For adjacent reads with the same grading discipline: enclomiphene and hCG cover the fertility-preserving side of male hormone therapy, DIM and estrogen metabolism takes on the other big estrogen-modulation supplement claim, and boron for testosterone is a study in how a small real signal becomes a large marketing promise. The wider picture lives on our state of the evidence on sex hormones page, and the broader supplements reference covers how we assess this category generally.