DHT: the most potent androgen, and the one you never need to measure
It triples on a transdermal gel and the report flags it. What the dutasteride trials proved about what DHT is actually for, and why three major guidelines never ask for the test.
The most potent androgen in the body, and the one you almost never need to measure.
What it measures
Dihydrotestosterone is testosterone with a single double bond removed from its first ring. That one change makes it bind the androgen receptor about four times as tightly, dissociate roughly three times more slowly, and hold the receptor in its DNA-binding state. It is the more powerful molecule, and it is made from the less powerful one.
The enzyme responsible is 5-alpha reductase, and it comes in two forms with different addresses. Type 2 predominates in the prostate, the male accessory sex glands and fetal genital skin. Type 1 is absent in the fetus, appears in skin at puberty, and thereafter is abundant in sebaceous glands, hair follicles, sweat glands and epidermis. Both are present in the liver after birth. That distribution — type 1 all over the skin — is the entire explanation for what gels do to a DHT result.
Two tracer studies from 1971 established where circulating DHT comes from, and they agree. Ito and Horton found that testosterone conversion accounts for at least 70% of plasma DHT in men, with only about 5.6% of circulating testosterone converted; Mahoudeau and colleagues put the conversion at about 4% and reached the same conclusion, that most plasma DHT derives from circulating precursors rather than from glandular secretion. Back-conversion of DHT to testosterone was undetectable.
This has a consequence that is rarely stated. Circulating DHT is spillover from peripheral tissue, not a hormone the body secretes. And it under-represents the two tissues anyone actually cares about: little of the DHT made inside the prostate or liver reaches the bloodstream, because intracellular enzymes convert it rapidly to androstanediols with almost no androgen activity.
The assay problem is worse than for testosterone
DHT differs from testosterone by one double bond, and in men it circulates at roughly one seventh to one tenth of the testosterone concentration. Antibodies raised against one bind the other, and because the interfering molecule is present at seven to ten times the concentration of the target, even modest cross-reactivity produces a large absolute error.
The measured consequence: a 2008 Clinical Chemistry paper introducing simultaneous testosterone and DHT measurement by mass spectrometry reported that DHT concentrations by LC-MS/MS were “markedly lower” than those from radioimmunoassay, because the immunoassay is non-selective without a chromatography step. A 2017 Endocrine Reviews survey states that older direct immunoassays read up to 25% high, and adds that others have reported over-reads of as much as 40%. Those two figures appear in the same paragraph and the review does not reconcile them; we report both rather than choosing.
One nuance worth carrying: when DHT is administered in pharmacological amounts, the sheer mass of DHT overwhelms the cross-reactivity, so old immunoassay data from high-DHT states are probably roughly right. Old immunoassay data from normal or low DHT states are not.
And there is no harmonisation programme. The CDC’s hormone standardisation programme certifies assays for testosterone and for estradiol. There is no DHT equivalent — no reference measurement procedure, no certified assay list. The practical result shows up in published reference intervals:
| Source | Method | Adult male interval |
|---|---|---|
| Clinical Chemistry 2008, 113 healthy men | LC-MS/MS | 14–77 ng/dL |
| Pooled community-dwelling men under 65 | LC-MS | 23–102 ng/dL |
| Mayo Clinic Laboratories | LC-MS/MS | 11.2–95.5 ng/dL |
| Labcorp Endocrine Sciences | HPLC–tandem MS | 30–85 ng/dL |
Between two large American reference laboratories, both using mass spectrometry, the lower limit differs almost threefold. A DHT of 25 ng/dL is normal at one and low at the other, with no change in the patient. Serial DHT values from different laboratories should not be compared, and a DHT result means nothing except against the interval of the laboratory that produced it.
To convert: ng/dL × 0.0344 = nmol/L. Some laboratories report pg/mL, where 1 pg/mL = 0.1 ng/dL.
Evidence: Established that the assay problem exists and that no harmonisation programme covers DHT.
What DHT does that testosterone does not
This is the question the whole page turns on, and it has been answered unusually well — first by a natural experiment, then by deliberate ones.
The natural experiment
In 1974 Imperato-McGinley and colleagues described 46,XY individuals with markedly reduced DHT from 5-alpha reductase type 2 deficiency. They are born with ambiguous external genitalia — and they virilise markedly at puberty. In a follow-up of 38 affected people, 18 had been raised as girls; 17 of those 18 adopted a male gender identity during or after puberty. Plasma testosterone was high-normal throughout.
What that proves is precise. DHT is obligatory for masculinisation of the external genitalia and development of the prostate before birth. It is not required for testosterone’s effects on muscle, voice, phallic growth at puberty, or male gender identity — all of which occur in men who have almost none of it. The mechanism for the pubertal catch-up is that type 1 5-alpha reductase switches on in skin at puberty, generating some DHT even in men who lack the type 2 enzyme. Affected men also show less temporal hair regression: they do not go bald.
The deliberate experiments
Three trials gave men testosterone with and without a 5-alpha reductase inhibitor, which separates the testosterone effect from the DHT effect experimentally. They are the strongest evidence on this page.
Muscle. In a 2012 JAMA trial, 139 healthy men aged 18 to 50 had their own testosterone production suppressed and were randomised to testosterone enanthate at 50, 125, 300 or 600 mg weekly, each with placebo or dutasteride 2.5 mg daily, for 20 weeks. Fat-free mass rose steeply with testosterone dose — and dutasteride made no difference at any dose (dose-adjusted p = 0.18). Nor was there any difference in fat mass, muscle strength, sexual function, prostate volume, sebum production, hematocrit or lipids. The authors’ conclusion: conversion of testosterone to DHT is not essential for its anabolic effects on muscle.
Bone. Seventy men aged 65 or over received testosterone with placebo, testosterone with finasteride, or placebo, for 36 months. Lumbar spine bone density rose 10.2% with testosterone alone and 9.3% with testosterone plus finasteride, against 1.3% on placebo. Hip density behaved the same way. Blocking DHT did not blunt the bone benefit.
Prostate. Fifty-three men with symptomatic benign prostatic enlargement received testosterone gel with or without dutasteride for six months. Serum testosterone rose similarly in both; DHT rose in one group and fell in the other. Prostate volume fell 12% and PSA fell 35% with dutasteride added, versus rises of 7.5% and 19% on testosterone alone. Symptoms improved in both groups.
Inside the prostate. In 32 healthy men biopsied after 12 weeks, testosterone gel alone did not change intraprostatic androgens at all despite raising serum testosterone. Adding dutasteride produced a 90% fall in intraprostatic DHT, an eleven-fold rise in intraprostatic testosterone — and no significant difference in prostate epithelial gene expression, globally or in androgen-regulated genes.
And the counter-experiment
The cleanest test of “is a high DHT harmful?” gave men DHT itself. In a two-year randomised trial, 114 healthy men over 50 received transdermal DHT 70 mg daily or placebo. Prostate volume and PSA rose over the two years in both arms — and DHT had no effect on either (p > 0.2). It did lower spinal bone density by 1.4%, raise hemoglobin by 7% and lean mass by 2.4%, lower fat mass by 5.2%, and suppress testosterone, estradiol, LH and FSH. Eight men stopped for a raised hematocrit, all resolving afterwards.
Read that carefully, because it cuts both ways. No prostate signal at all, over two years, at supraphysiological DHT. But a real bone signal and a real erythrocytosis signal — both plausibly driven by DHT suppressing the man’s own testosterone and estradiol rather than by DHT itself, since estradiol is what male bone actually depends on.
Evidence: Established that DHT is not required for testosterone’s effects on muscle, bone, erythropoiesis, sexual function or lipids. Established that DHT is obligatory for fetal genital masculinisation.
What testosterone therapy does to it
Transdermal testosterone crosses skin, and skin is where type 1 5-alpha reductase lives. A larger fraction is converted on the way in. Injected esters bypass that compartment entirely. This is not a subtle effect.
| Formulation | What happens to serum DHT | DHT/T ratio at steady state |
|---|---|---|
| 1% transdermal gel | Rises 3.6-fold at 50 mg/day and 4.6-fold at 100 mg/day; absolute levels of 200–266 ng/dL, at or above the top of the reference range | 0.25–0.30 |
| 1.62% transdermal gel | Lower than 1% gel, still about twice baseline | 0.15–0.17 |
| 2% axillary solution | 18 → 98 ng/dL over 120 days | 0.26 |
| Testosterone patch | Rises 1.3-fold | — |
| Injectable enanthate, 125 mg weekly | About 20 → 70 ng/dL. Rises in absolute terms, but the ratio falls as dose rises | Falls with dose |
| Injectable undecanoate | About 17 → 25–30 ng/dL; stays in range | 0.075 |
| Nasal gel | 33–40 ng/dL at 90 days — the lowest of the products | <0.1 |
| Subcutaneous pellets | Peak around 290 ng/dL near day 60 | — |
| Oral undecanoate | Raises the ratio more than parenteral routes, because gut wall and lymphatics have their own 5-alpha reductase | High |
So: on a 1% gel, expect DHT to triple or quadruple and expect the report to flag it. On weekly injections, expect it to rise in absolute terms while the ratio to testosterone falls. On long-acting injectable undecanoate or nasal gel, expect it barely to move.
Evidence: Established.
Should anyone measure it?
No major guideline asks for it, and the way each one declines to is informative.
The Endocrine Society’s 2018 guideline mentions DHT five times and never as a test. It states that conversion of testosterone to DHT “is not obligatory for mediating its effects on erythropoiesis or muscle mass and strength.” Its formulation table lists, as a disadvantage of transdermal gels and the axillary solution, “moderately high DHT concentrations (of unknown significance)” — a phrase that neither reassures nor alarms. Its monitoring recommendation specifies symptoms, adverse effects, adherence, serum testosterone, hematocrit and prostate cancer risk. DHT is not on the list.
The AUA’s 2018 testosterone deficiency guideline does not mention DHT anywhere, including in its evaluation and monitoring tables.
The EAU guideline mentions DHT only as a drug, not as a measurement.
Three guidelines, zero recommendations to measure it. That silence is the answer to the question.
Where a DHT level does earn its place
Suspected 5-alpha reductase deficiency, in the workup of a 46,XY difference of sex development. This is the one supported use — and even here it is imperfect. A testosterone-to-DHT ratio above 10 misses roughly 28% of genetically confirmed cases in one series, and a 2013 review concluded DHT measurement is “not essential” when genetic testing is available.
Confirming adherence to a 5-alpha reductase inhibitor. Pharmacologically coherent: dutasteride 0.5 mg suppresses serum DHT by 94.7% and finasteride 5 mg by 70.8%, so a normal DHT on either drug is informative. But this use has never been formally validated as an adherence test, and we are not going to describe it as though it had been.
Investigating a symptom-level mismatch. Speculative. There is no evidence that a DHT measurement resolves the discordance between how a man feels and what his testosterone reads, and no study has tested whether acting on one improves anything.
Evidence: Limited for the DSD indication as a stand-alone test; Unsupported as a routine measurement in men on or considering testosterone.
One practical point about serum versus tissue
Finasteride suppresses DHT by about 70% in serum and by as much as 85 to 90% in the prostate. Same drug, same patient, two different numbers. Serum DHT and prostate DHT are not interchangeable, and neither is a proxy for the other — which is another reason a blood level is a weak guide to anything happening in the tissues people worry about.
Related
Total testosterone · Estradiol · PSA · Testosterone replacement therapy · Testosterone, the prostate and the heart · The Lab Library
Questions patients ask
What the number means
My DHT came back above the reference range on gel, so my dose is too high.
It is the formulation, not the dose, and it is the expected pharmacology.
What the evidence showsA 1% transdermal gel raises serum DHT 3.6-fold at 50 mg/day and 4.6-fold at 100 mg/day, to absolute concentrations of 200 to 266 ng/dL — which the authors of that study noted were "at or above the normal adult male range." Skin is rich in type 1 5-alpha reductase, so transdermal testosterone is partly converted on its way into the circulation. A patch, which delivers through a smaller area, raises DHT only 1.3-fold.
What remains uncertainWhether the elevation has any consequence. The Endocrine Society's own guideline calls it "of unknown significance," which is a fair summary of the literature.
Bottom lineAn above-range DHT on gel is a property of the route of administration. It is not by itself evidence that anything is wrong.
Strong on the magnitude; Limited on whether it matters
A DHT of 25 means the same thing wherever it was measured.
It does not, and the difference is larger than most people would guess.
What the evidence showsTwo major American reference laboratories, both using mass spectrometry, publish adult male lower limits of 11.2 and 30 ng/dL. Published research intervals run from 14 to 23 at the bottom and 77 to 102 at the top. There is no CDC standardisation programme for DHT as there is for testosterone and estradiol — no reference measurement procedure and no certified assay list.
What remains uncertainWhether the between-laboratory gap is methodological, population-based or statistical. Neither laboratory's public documentation says.
Bottom lineA DHT result is interpretable only against the reference interval of the laboratory that produced it, and serial values from different laboratories should not be compared.
Strong
If the DHT was measured by immunoassay it is close enough.
More complicated than it sounds — the direction of the error is known and it is not small.
What the evidence showsDHT differs from testosterone by one double bond and circulates at a seventh to a tenth of the testosterone concentration, so antibody cross-reactivity produces a large absolute error. Mass spectrometry values are "markedly lower" than radioimmunoassay values without a chromatography step. Published over-read estimates are up to 25% in one place and as much as 40% in another, in the same review.
What remains uncertainThe cross-reactivity of specific commercial DHT immunoassays. We could not find a published figure for any of them.
Bottom lineAsk for mass spectrometry, or treat an immunoassay DHT as an upper bound rather than a measurement.
Strong
What DHT actually does
DHT is the real androgen and testosterone is just a prohormone, so DHT is what should be optimised.
This claim is more complicated than it sounds. The receptor pharmacology is right; the physiological conclusion does not follow.
What the evidence showsDHT does bind the androgen receptor about four times as tightly as testosterone and dissociates about three times more slowly. But when 139 men were given testosterone across a twelve-fold dose range with dutasteride or placebo, blocking DHT changed nothing — not fat-free mass, strength, fat mass, sexual function, prostate volume, sebum, hematocrit or lipids. Over 36 months, testosterone plus finasteride produced the same lumbar spine bone gain as testosterone alone. And men born with 5-alpha reductase type 2 deficiency virilise at puberty.
What remains uncertainWhether DHT matters for outcomes nobody measured in those trials, over horizons longer than three years.
Bottom lineDHT is the more potent molecule and the less necessary one. Outside fetal development and the prostate, testosterone does the work.
Strong
A high DHT is what causes prostate problems, so a high DHT on testosterone is dangerous.
Lowering DHT changes the prostate. Raising it, in the one trial that tested this directly, did not.
What the evidence showsIn 114 men given transdermal DHT or placebo for two years, prostate volume rose 29% and PSA rose 15% over time in the study as a whole — with no effect of DHT on either (p>0.2). A 2017 review reported it could not identify a single epidemiological study showing an association between circulating DHT and prostate cancer risk, and the REDUCE trial's placebo arm found no relationship between baseline androgens and cancer risk. In the opposite direction, dutasteride and finasteride do reduce prostate cancer prevalence, by 22.8% and 24.8% respectively — though both trials found more high-grade tumours in the treated arm, which has never been fully resolved.
What remains uncertainWhether the high-grade signal in those chemoprevention trials was a detection artefact or a real effect. The debate is unsettled.
Bottom lineBlocking DHT shrinks the prostate. It does not follow that a naturally higher DHT is a hazard, and the direct test found no prostate effect.
Mixed — Strong that 5-alpha reductase inhibitors shrink the prostate; Limited that higher circulating DHT confers risk
My DHT is high, so that explains my hair loss and I should get it down.
More complicated than it sounds. DHT is genuinely involved in male pattern hair loss; the blood level is not the variable.
What the evidence showsMen with 5-alpha reductase deficiency and castrated men do not develop male-pattern baldness, so androgen conversion is necessary for the process. But type 1 5-alpha reductase expression does not differ qualitatively between balding and non-balding scalp in the same men, which points to receptor and follicle-level factors rather than to circulating supply. No study has shown that a serum DHT level predicts the rate of hair loss or the response to treatment.
What remains uncertainEssentially the whole question of what distinguishes a susceptible follicle from a resistant one on the same head.
Bottom lineDHT matters for hair. Your serum DHT number does not tell you about your hair.
Mixed — Strong that androgens drive the process; Unsupported that the blood level is informative
Whether to order it
DHT should be part of routine monitoring on testosterone therapy.
No guideline asks for it, and two of the three never mention it as a test at all.
What the evidence showsThe Endocrine Society's 2018 monitoring recommendation lists symptoms, adverse effects, adherence, serum testosterone, hematocrit and prostate cancer risk — not DHT. The same guideline describes the gel-associated DHT rise as "of unknown significance." The AUA's 2018 guideline does not mention DHT anywhere. The EAU mentions it only as a drug.
What remains uncertainNothing about what the guidelines say. What remains genuinely unknown is whether the gel-associated elevation has any long-term consequence — nobody has looked.
Bottom lineOrdering it will produce a number that no society knows what to do with, and that a laboratory switch could change by threefold.
Unsupported as routine monitoring
Adding a 5-alpha reductase inhibitor would make gel safer.
It would lower the number. Whether it makes anything safer has not been shown, and there is a cost.
What the evidence showsDutasteride added to testosterone gel does shrink the prostate and lower PSA. But across the dutasteride-plus-testosterone trials, blocking DHT removed no adverse effect that was being measured — hematocrit, lipids, sexual function and sebum were unchanged. Meanwhile finasteride and dutasteride carry their own well-documented sexual and mood adverse effects, and both chemoprevention trials showed more high-grade prostate tumours in the treated arm.
What remains uncertainWhether there is any man on testosterone for whom this combination has a net benefit outside of treating co-existing prostatic enlargement.
Bottom lineIf a man on testosterone also needs treatment for prostatic obstruction, that is a reason. Making a DHT number look better is not.
Limited
Ready for testosterone care built on a diagnosis?
The first step is a full endocrine evaluation, not a prescription. We see patients across San Diego County and welcome referrals from other physicians.
