Pregnenolone

Best-in-class: Pregnenolone

The 'grandmother hormone' precursor to all other steroid hormones (DHEA, cortisol, sex hormones), used for memory, mood and hormonal support with age.

How Pregnenolone actually works

The first steroid made from cholesterol and the precursor to everything downstream — progesterone, DHEA, cortisol, aldosterone and the sex hormones. It is also a neurosteroid in its own right, modulating GABA-A and NMDA receptors and enhancing memory consolidation in animal models.

Where to get Pregnenolone

The evidence for Pregnenolone

Graded by what exists behind each claim.

✅ Clinically validated

• Studies suggest memory/cognition and mood benefits; it's a potent neurosteroid.
• Declines significantly with age.

📊 Correlative data

• Levels fall substantially with age across populations, which is the entire observational basis for supplementing it. Age-related decline is not by itself evidence that restoration helps — the same reasoning has failed for several other hormones.

🧪 Theoretical / extrapolated benefits

• Because it sits upstream of many hormones, effects are individual and hard to predict.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What Pregnenolone actually does

Pregnenolone is the first steroid. Cholesterol is delivered to the inner mitochondrial membrane by StAR, and CYP11A1 — cholesterol side-chain cleavage enzyme — performs two hydroxylations and a carbon–carbon scission that removes six carbons from the side chain. What is left is pregnenolone. Every steroid a human makes is downstream of that single reaction, which is why the marketing calls it the grandmother hormone and why that is the least interesting true thing about it.

The interesting pharmacology is not that pregnenolone becomes other hormones. It is that pregnenolone and its immediate metabolites are themselves ligands at ion channels. Sulfated by SULT2B1, it becomes pregnenolone sulfate, a negative allosteric modulator at GABA-A receptors and a positive modulator at NMDA receptors. Taken the other way — 3-beta-HSD to progesterone, 5-alpha-reductase to 5-alpha-dihydroprogesterone, 3-alpha-HSD to allopregnanolone — it becomes one of the most potent positive allosteric modulators of GABA-A known, acting at a site distinct from the benzodiazepine site. Those two metabolites push the same receptor system in opposite directions, and which one a tissue makes depends on which enzymes that tissue expresses.

Which is why the human study that measured both the steroid and the brain is the one that matters. Thirty-one volunteers received 400 mg of pregnenolone (n = 16) or placebo (n = 15) and were scanned at 3 tesla. Allopregnanolone elevations following pregnenolone administration were associated with enhanced activation of emotion regulation neurocircuits, and reduced self-reported anxiety tracked with changes in amygdala–prefrontal connectivity Sripada 2013. That is a dose, a metabolite measurement, and a brain measurement in the same people — the chain most supplement pages assert and never demonstrate.

Now the claim this page will not make. The commonest story told about pregnenolone is that chronic stress diverts it toward cortisol and starves the sex hormones — the ‘steal’. This site's own bloodwork reference standard says on the rendered marker page that the steal is not supported by the physiology, and the reason is structural: steroidogenesis is not a single shared pool being rationed. The adrenal zona fasciculata and the gonad are different cells expressing different enzymes, each making pregnenolone locally from cholesterol, and cholesterol is not scarce. A cell cannot run short of a substrate it synthesizes on site from something it has in excess.

There is exactly one condition in which pregnenolone genuinely accumulates, and it is a gene, not a mood. In 17-alpha-hydroxylase deficiency, loss-of-function mutations in CYP17A1 block 17-alpha-hydroxylation, so the pathway cannot turn toward cortisol or the sex steroids. Pregnenolone and progesterone back up and are pushed instead down the mineralocorticoid limb toward deoxycorticosterone and corticosterone; the clinical result is hypertension, hypokalemia and absent sex steroid production Yanase 1991, and the same enzyme block can be produced pharmacologically by CYP17A1 inhibitors Auchus 2017. That is what a real pregnenolone traffic jam looks like. It presents to an endocrinologist with high blood pressure and a low potassium, not to a supplement shelf with fatigue.

Cell, rodent, human — and where it stops

There is no cell-to-rodent-to-human ladder here, and that is unusual enough to say first. Pregnenolone went almost directly into randomized human trials, because it is an endogenous human steroid with a known structure and no species question. What follows is therefore four human trials, and then the arithmetic that undoes them for the reader.

Trial one, the proof of concept. Twenty-one patients with schizophrenia were randomized and 18 completed at least four weeks (n = 9 per group), taking fixed escalating doses to 500 mg/day for eight weeks. Negative symptoms on the SANS improved by a mean change of 10.38 on pregnenolone against 2.33 on placebo (p = 0.048), and — the detail that makes this a mechanistic result rather than a clinical one — serum pregnenolone increases predicted cognitive composite scores, r_s = 0.81, p = 0.022 Marx 2009. Oral pregnenolone raised serum pregnenolone, and how much it rose predicted how well people did.

Trial two, the bigger version, and it is mostly negative. One hundred and twenty participants, eight weeks. No significant improvement on the MATRICS cognitive battery — the primary cognitive outcome. Functional capacity on the UPSA-B did improve (n = 56 versus 55, p = 0.03), with the communication subscale at p < 0.001 Marx 2014. Six times the sample size, and the cognitive endpoint the first trial pointed at did not replicate.

Trial three, a different illness and a real result. Eighty adults with bipolar depression were randomized to pregnenolone titrated to 500 mg/day or placebo for 12 weeks. Depression remission rates were 61% on pregnenolone against 37% on placebo (p = 0.046) Brown 2014. A 24-point absolute difference in remission is not a subtle finding.

Trial four is the mechanism study already described, at 400 mg Sripada 2013.

Now the arithmetic, and it is the whole page. Every trial above used 400 to 500 mg a day. The dose on this card, and on essentially every retail bottle, is 10 to 50 mg. That is one-tenth to one-fiftieth of the only doses with randomized evidence behind them. Nobody has run 10 mg against placebo for anything. Nobody has published a serum pregnenolone or allopregnanolone curve after 10 mg. The evidence base and the product are separated by more than an order of magnitude, and the card's own instruction — test and go low — is the sensible clinical advice and simultaneously a decision to leave the evidence behind.

So the specific obstacle is dose, and it is unusually clean. It is not species: these are human trials. It is not the form: it is the identical molecule. It is not absorption: serum pregnenolone measurably rose Marx 2009. The obstacle is that the shelf sells a twentieth of the studied dose, and the studied dose was chosen because smaller ones were not expected to work. The second obstacle is population: every trial above was in psychiatric illness — schizophrenia and bipolar depression — and the card sells it for memory and mood in healthy older adults, where no randomized trial of any dose exists.

Pregnenolone — which form, and does it matter

Pregnenolone is not a nutrient and there is no botanical question here. Retail pregnenolone is synthetic, made from diosgenin from wild yam, and it is chemically identical to the human steroid whatever the starting material was. A label saying ‘from wild yam’ describes a factory input, not a difference in what you swallow. A label saying ‘wild yam extract’ and nothing else is a different product entirely: humans do not convert diosgenin to pregnenolone, so that capsule contains a sapogenin and no hormone.

Micronization is the only oral formulation variable with a rationale. Pregnenolone is a lipophilic, poorly water-soluble steroid, and dissolution limits absorption for compounds of that class; micronized powder and oil-suspended softgels exist for that reason. What can be said with evidence is narrower and more useful: the trial formulations, given orally at 400–500 mg, raised serum pregnenolone measurably, and the size of the rise tracked the clinical result Marx 2009. Oral dosing works. Whether one micronized brand beats another has never been measured against a plasma curve.

Sublingual and transdermal are sold on an argument that is half-right. The argument is first-pass avoidance, and for a lipophilic steroid it is not silly. But the human trials that demonstrated anything used oral dosing Marx 2009 Marx 2014 Brown 2014 Sripada 2013, and no sublingual or transdermal pregnenolone has a published serum curve, let alone an outcome. Choosing a route to avoid the liver also chooses a route with no evidence.

The form distinction that actually changes the pharmacology is one nobody sells. Pregnenolone and pregnenolone sulfate are not the same molecule at a receptor: the sulfate is the NMDA-positive, GABA-negative species, while the allopregnanolone route is GABA-positive. A product delivering the free steroid hands the body a substrate and lets tissue enzymes decide which of two opposing neuroactive metabolites gets made. That is the honest reason the card says effects are individual and hard to predict — it is not vagueness, it is a branch point with real chemistry at it.

What a label would have to carry to make this a considered purchase: a milligram number that matches a trial, which means 400 to 500 mg and not 10 to 50 Marx 2009 Brown 2014; or, failing that, an honest statement that the dose in the bottle has never been tested against placebo for anything. No product on the market carries either.

What would have to be true, and how you would know it was not

1. Serum pregnenolone, because it is the only prediction the trials already validated. Draw it before starting and at four weeks. Predict a rise, and predict that the size of the rise is what matters — serum pregnenolone increases predicted cognitive outcome at r_s = 0.81 Marx 2009. This is the rare supplement where the exposure biomarker is available on a standard requisition, and taking a hormone precursor without measuring the hormone is a choice, not a limitation.

2. dhea-s, total testosterone and estradiol — the downstream panel, and the prediction that cuts against the product. At 10 to 50 mg predict all three unchanged. The dose is a tenth to a fiftieth of anything studied, and no publication reports a downstream hormone panel at retail doses. If the product is doing what its own card says — supplying the upstream precursor so downstream hormones can rise — then a flat DHEA-S at twelve weeks falsifies it outright. Retest at twelve weeks, same lab, morning draw.

3. The dose-escalation prediction, which is the one that could prove this page too pessimistic. Predict that at 400–500 mg the downstream panel does move, and that anxiety and sleep change with it, because that is the dose at which allopregnanolone rose enough to alter measurable brain activity Sripada 2013 and at which two randomized trials found clinical effects Marx 2009 Brown 2014. This prediction is not a dosing recommendation — a 500 mg steroid dose belongs in a supervised trial — but it is the correct reading of the evidence, and it says the product is probably not too dangerous at 10 mg so much as too small.

4. Blood pressure and serum potassium, derived from the enzyme block and labeled as extrapolation. When CYP17A1 is blocked and the 17-deoxy limb runs unopposed, the accumulating steroids are deoxycorticosterone and corticosterone, and the presentation is hypertension with hypokalemia Yanase 1991. Nobody has shown that oral pregnenolone reproduces any part of that. But it identifies the two cheapest markers that would reveal unwanted mineralocorticoid traffic, they are on any basic panel, and nobody monitoring pregnenolone has been told to look at them. Extrapolated from a genetic disease to a supplement, and flagged as such.

5. Sleep and dream recall as the fastest read-out. Allopregnanolone is a GABA-A positive modulator, and the human data attach the metabolite to reduced anxiety within a single dosing session Sripada 2013. Predict that if anything is happening at all it shows up in sleep architecture within one to two weeks — vivid dreams, altered sleep onset — long before any hormone panel changes. A completely unchanged subjective picture at four weeks on a retail dose is consistent with the dose being too small to be detected.

What nobody has tested yet

Nobody has published a serum curve for the dose people actually take. The randomized evidence is at 400–500 mg Marx 2009 Marx 2014 Brown 2014; the shelf sells 10 to 50 mg. Twelve volunteers, one 25 mg dose, serial plasma for pregnenolone, pregnenolone sulfate, allopregnanolone, DHEA-S and progesterone by mass spectrometry, is a single day of clinic time and it would tell every buyer of this product whether their capsule does anything measurable. It has not been done.

Nobody has run it in the population it is sold to. Every trial in this file is in psychiatric illness. The card sells pregnenolone for cognition and mood in older adults, and there is no randomized trial of pregnenolone for memory in healthy aging at any dose. The observation that levels fall with age is the entire rationale, and age-related decline has failed as a rationale for several other hormones already.

Nobody has measured the downstream hormones in the trials that worked. Four randomized trials gave 400–500 mg of a steroid precursor for eight to twelve weeks, and the published outcomes are psychiatric scales and neurosteroid levels. Testosterone, estradiol, DHEA-S and cortisol at those doses are not in the record. The single most-asked question about this compound — what does it do to your other hormones — has been asked in the exact experiment that could answer it, and not measured.

And nobody has tested the branch point. Whether an individual sends pregnenolone preferentially toward the sulfate or toward allopregnanolone would predict whether they find it activating or sedating, and both metabolites are measurable by mass spectrometry in the same sample Sripada 2013. Nobody has phenotyped a cohort that way. It is the obvious explanation for why this compound's user reports are so contradictory, and it is untested.

Pregnenolone — its own safety story, not its category's

The honest starting point is that this is a hormone, not a nutrient, and the dose people take is unstudied in both directions. Randomized safety data exist only at 400–500 mg over 8 to 12 weeks in psychiatric populations Marx 2009 Marx 2014 Brown 2014. At 10–50 mg there is no controlled safety data set at all — not because the dose is dangerous, but because nobody has studied it. ‘Low dose’ and ‘characterized’ are not the same word.

The hormone-sensitive-cancer question has a stronger chemical basis here than for any phytoestrogen on this site. Pregnenolone is an obligate precursor of androstenedione and testosterone, and aromatase in adipose and breast tissue converts those to estrone and estradiol. This is not receptor mimicry at some fraction of estradiol's affinity; it is the actual substrate, entering the actual pathway. Anyone with a history of breast, endometrial, ovarian or prostate cancer, and anyone taking an aromatase inhibitor or tamoxifen, is adding raw material to the pathway their treatment exists to restrict.

The interaction that belongs on this page and appears on no label. Neurosteroid modulation of GABA-A is the mechanism by which allopregnanolone acts Sripada 2013, and it is the same receptor complex that benzodiazepines, zolpidem, barbiturates and alcohol act on through different sites. Additive sedation is the predictable consequence, it has never been formally studied with pregnenolone, and the practical reading is that anyone on a benzodiazepine or a z-drug should assume the combination is stronger than either alone until somebody measures it.

What 17-alpha-hydroxylase deficiency teaches about monitoring. The one physiological state where pregnenolone genuinely accumulates produces hypertension and hypokalemia through the mineralocorticoid limb Yanase 1991 Auchus 2017. Nothing says a supplement reproduces that. It does say that if you are going to monitor anything besides the sex hormones on sustained dosing, blood pressure and potassium are the physiologically motivated choices, and they cost nothing to add.

And the reason to draw blood before rather than after. Pregnenolone is upstream of everything, so a hormone panel taken while on it cannot be interpreted — you cannot tell an endogenous abnormality from the product. The trial that found the strongest signal found it in the change in serum pregnenolone Marx 2009, which requires a baseline. Starting a steroid precursor without one throws away the only measurement that would have made the experiment interpretable, and it also throws away the ability to investigate whatever symptom sent you looking in the first place.

Sources read for this page

• Marx CE, et al. Proof-of-concept trial with the neurosteroid pregnenolone targeting cognitive and negative symptoms in schizophrenia. Neuropsychopharmacology 2009;34(8):1885-1903 · PMID 19339966
• Marx CE, et al. Proof-of-concept randomized controlled trial of pregnenolone in schizophrenia. Psychopharmacology 2014;231(17):3647-3662 · PMID 25030803
• Brown ES, et al. A randomized, double-blind, placebo-controlled trial of pregnenolone for bipolar depression. Neuropsychopharmacology 2014;39(12):2867-2873 · PMID 24917198
• Sripada RK, et al. Allopregnanolone elevations following pregnenolone administration are associated with enhanced activation of emotion regulation neurocircuits. Biological Psychiatry 2013;73(11):1045-1053 · PMID 23348009
• Yanase T, Simpson ER, Waterman MR. 17 alpha-hydroxylase/17,20-lyase deficiency: from clinical investigation to molecular definition. Endocrine Reviews 1991;12(1):91-108 · PMID 2026124
• Auchus RJ. Steroid 17-hydroxylase and 17,20-lyase deficiencies, genetic and pharmacologic. Journal of Steroid Biochemistry and Molecular Biology 2017;165(Pt A):71-78 · PMID 26862015

How you would know if it worked

Pregnenolone — safety & side effects

Bloodwork to run alongside Pregnenolone

Baseline first, then again at 8–12 weeks.

The Basics — Start Here panel covers these in one order — 4 markers, $32.40 with the discount applied.

Check results you already have → · All 103 markers A–Z

Pregnenolone — frequently asked questions

What Pregnenolone is used for

Pregnenolone appears under 2 goals in the goal router.

Related Hormonal & Wellness supplements

← Browse the full Supplement Vault

↑ Back to on this page