Why Minoxidil Doesn't Work for Everyone: The SULT1A1 Enzyme Explained

Two men, same bottle of 5% minoxidil, same twice-a-day discipline for eight months. One gets a visibly thicker crown. The other gets nothing but an itchy scalp and a wasted year. For decades that looked like bad luck. It isn’t. It’s an enzyme.

Minoxidil is the most used hair-loss drug on earth and one of the least understood by the people using it. It’s a prodrug. The molecule in the bottle does nothing to a follicle until an enzyme in your scalp bolts a sulfate group onto it. If you have plenty of that enzyme, minoxidil works. If you have little, it mostly doesn’t, and no amount of “give it another six months” changes the biology.^[1]

That enzyme has a name, SULT1A1, and there’s a test for it. This article is about what the research shows, what the test can and can’t tell you, and what I’d do if I were a non-responder, which, as it happens, I have been.

The 1990 experiment that settled what minoxidil is

Buhl and colleagues at Upjohn (the company that made Rogaine) took whisker follicles from newborn mice, grew them in culture, and asked a simple question: is it minoxidil or minoxidil sulfate that makes hair grow?^[1] They measured the follicles’ own sulfotransferase activity, confirmed the follicles convert minoxidil to minoxidil sulfate, and then ran dose-response curves. Minoxidil sulfate was 14 times more potent than minoxidil at stimulating the follicles.

Then they did the part that makes it conclusive. They blocked the sulfotransferase enzyme with diethylcarbamazine. Minoxidil stopped working. Minoxidil sulfate kept working. They blocked the cell’s supply of activated sulfate with chlorate. Same result. The paper’s own summary: sulfation is a critical step, and it’s the sulfated metabolite that acts on the follicle.

This is why I keep saying minoxidil is a prodrug, in the same sense that codeine is a prodrug for morphine. Some people lack the enzyme to convert codeine and get no pain relief. Some people lack scalp sulfotransferase and get no hair.

Thirty to forty percent, and a test that predicts it

Response rates for topical minoxidil in men have hovered around 30 to 40 percent for as long as the drug has existed, which is a strange number for a first-line treatment. In 2014 Andy Goren’s group published a retrospective study asking whether SULT1A1 activity in plucked hair follicles tracked with who responded.^[2] It did: the assay predicted response with 95 percent sensitivity and 73 percent specificity. Roberts and colleagues then replicated it prospectively in women treated with 5% minoxidil for six months, and got 93 percent sensitivity and 83 percent specificity.^[3]

Two things to say about these numbers. Sensitivity of 93 to 95 percent means that if the test says you have the enzyme, you almost certainly have the capacity to respond. Specificity of 73 to 83 percent means the test calls some men non-responders who would have responded, so a negative result is a strong hint, not a verdict. And both studies were small, from one research group, in a journal that is not JAAD. I take the finding seriously because it rests on Buhl’s 1990 mechanism, not because the clinical samples were large. A mechanism plus two consistent small studies is a different kind of evidence from one big trial, and it’s worth knowing which kind you’re looking at.

A negative enzyme test is a strong hint. A positive one is close to a promise that the drug can work for you, if you can tolerate it long enough.

What the dose does, and doesn’t, fix

The obvious move for a weak converter is more drug. Olsen’s 48-week trial in 393 men compared 5% to 2% minoxidil and found 45 percent more regrowth with the higher strength, and an earlier response.^[4] More substrate does push more product through a limited enzyme. But the relationship isn’t linear, and a scalp with very little SULT1A1 can be flooded with minoxidil and still make little minoxidil sulfate. The 5% versus 2% gain is real and it’s the cheapest thing to try. It’s also the smallest lever on this list.

The second lever is absorption. Ferry’s crossover study in 19 men found that applying tretinoin cream alongside 2% minoxidil nearly tripled the amount of minoxidil absorbed through the skin, by increasing stratum corneum permeability.^[7] That’s a bigger multiplier than the bottle strength gives you, and it’s the pharmacological basis for the minoxidil-plus-tretinoin combinations I’ve written about separately. It still only moves more minoxidil toward an enzyme that may or may not be there.

The third lever changes the route entirely. Oral minoxidil puts the drug in the bloodstream, and the whole body’s sulfotransferase capacity, liver included, takes part in activating it. Panchaprateep’s open-label study gave 30 men 5 mg a day for 24 weeks and measured a mean gain of 35 hairs per square centimetre, with 93 percent of them growing hair elsewhere on the body as well.^[5] The 2022 JAMA Dermatology network meta-analysis ranked oral minoxidil 5 mg as the single most effective option for terminal hair count at 24 weeks, ahead of both topical forms and ahead of finasteride 1 mg.^[6] Many men who “failed” topical minoxidil respond to the pill. Whether that’s because oral dosing bypasses a weak scalp enzyme, or simply because the systemic dose is larger, nobody has cleanly separated. The clinical observation holds either way.

What the big minoxidil trial can and can’t tell you about yourself

It’s worth being precise about what a “response rate” is, because the number does a lot of quiet work in this article. Olsen’s trial randomized 393 men aged 18 to 49 to 5% minoxidil, 2% minoxidil or vehicle, twice a day for 48 weeks, and measured hair counts in a fixed target area.^[4] The 5% arm was superior to 2% and to placebo on hair count, on patient ratings of coverage, and on investigator ratings, and it responded earlier. It also produced more itching and irritation than the 2% strength. All of that is a statement about the average of 157 men.

Inside that average sit men who gained a great deal and men who gained nothing, and the trial wasn’t designed to tell them apart in advance. That’s the gap the enzyme work fills. A population mean tells you the drug works; a sulfotransferase assay tries to tell you whether it will work in your scalp. The two kinds of evidence answer different questions, and the mistake I see most often is treating a positive trial as a promise to each individual who buys the bottle. It isn’t. It’s a promise about the group, and the group includes the third of men for whom the bottle was never going to do much.

One more thing the trial shows: the 5% arm separated from placebo within the first months, not at the end.^[4] If you are a responder, the photos usually say so by month four to six. If month six looks like month zero, more months rarely change the verdict, and that is exactly the point at which the enzyme explanation, and the escalation ladder below, become the useful next thought.

What I’d do as a non-responder

Here’s the order I’d work through, and it’s the order I used on myself before deciding minoxidil wasn’t my drug.

One. Give it a fair trial before calling it. Six months of twice-daily 5% solution or foam, with standardized photos at zero and six months. Not a mirror. Not a feeling. The dread shed in months one to three fools almost everyone into quitting right before the drug shows itself, and I’d rather lose six months than mislabel myself.

Two. If the photos are flat, consider the enzyme test. It’s commercially available through a few labs, it’s a plucked-hair sample, and a positive result tells you to persist or escalate rather than abandon. A negative result, given the 73 to 83 percent specificity, tells you to escalate rather than persist.

Three. Escalate the route, not the frequency. Tretinoin co-application is the cheap experiment. Oral minoxidil is the stronger one, and it’s a prescription conversation because the side effect profile is systemic: hypertrichosis in most men at 5 mg, fluid retention and heart rate effects in some.^[5][8] I’ve covered the real numbers in the oral versus topical piece. The point here is narrower: the pill is the one form of minoxidil whose success does not hinge on your scalp’s enzyme.

Four. Remember what minoxidil is for. It’s a growth stimulant. It does nothing about the androgen signal that’s miniaturizing the follicle in the first place. A non-responder who skips minoxidil and runs a 5-alpha-reductase inhibitor plus a topical antiandrogen is not giving up. He’s spending his effort on the mechanism that’s causing the loss. In my own protocol, that’s where the weight sits, and minoxidil is the optional accelerator, not the foundation.

The takeaway

If minoxidil never did anything for you, the most likely reason isn’t that you applied it wrong. It’s that your scalp doesn’t switch it on. That’s a measurable thing, there’s a route around it, and there’s a whole other class of treatment that doesn’t depend on it at all. Stop blaming yourself for a drug that was never going to work in your skin.