Article draft · testosteroneinwomen.com · 25 July 2026

A Patch That Reads Your Estradiol in Sweat. So Far, Two Women Have Worn One.

You've had the appointment where your labs come back and the word used is "normal," and you leave with nothing. Maybe an offer of an antidepressant. Maybe a suggestion that this is just your age.

So the idea of a patch on your arm, streaming your actual hormone levels to your phone, lands somewhere between appealing and overdue. Diabetics got that fifteen years ago. Why are we still guessing from one tube of blood drawn on whatever Tuesday the lab had an opening?

Researchers are working on it. Three groups have made real progress in the past two years, and one of them has measured estradiol in human sweat and matched it against blood.

I went through the actual papers, the EU grant records, and the FDA device databases, because the press coverage of this field is running well ahead of the evidence. What I found is more interesting than the hype, and also more sobering.

Here's the short version. The technology is real. The sample sizes are tiny. And the number these devices would show you is not the number your doctor orders, which turns out to be the most important sentence in this whole piece.

What has actually been built

Three efforts are worth knowing about. They're at genuinely different stages, and coverage tends to flatten them into one story.

The Caltech sweat patch. In Wei Gao's lab, a finger-worn sensor measured estradiol in sweat and transmitted it wirelessly. It detected down to 0.14 picomolar, and readings correlated with blood at 0.837.

That study involved two women, across two menstrual cycles, 51 paired measurements.

The microneedle platform. Published this month, a four-channel array that reads estradiol, progesterone, luteinizing hormone and testosterone from the fluid just under your skin.

The continuous four-hormone tracking was done in rats. Validation used 45 banked human serum samples in a tube. The human on-body portion, five days of wear, measured testosterone only.

EstroPatch. A European fellowship at KU Leuven, funded at about €321,000, to build a microneedle patch reading estrone, estradiol and estriol from interstitial fluid.

That's one postdoc with a grant. It's a real project, not a product.

There's also a startup, FemSync, spun out of Anne Andrews' lab at UCLA, marketing a 30-day patch that measures four hormones from passive sweat. I'll come back to that number, because it deserves scrutiny.

That's the entire human evidence base for wearable reproductive-hormone sensing. Not a criticism of the science, which is genuinely elegant. Just the size of it.

Why hormones are so much harder than glucose

The comparison to continuous glucose monitors is the reason this field gets funded, and it's also what makes people expect a product next year.

The analytical gap is enormous, and worth seeing in real numbers.

Glucose sits in your blood at about 5 millimolar. Estradiol after menopause sits at roughly 20 picomolar.

Read the highlighted rows. For a woman past menopause, a sensor has to find a target a quarter of a billion times more dilute than what a glucose monitor looks for. In sweat, half a billion.

And here's a wrinkle the coverage misses entirely. Sensitivity isn't the only problem. Range is.

Across one healthy cycle, your estradiol moves about a hundredfold. A single sensor has to stay accurate across two or three decades of concentration. Detecting a very small number is hard. Being accurate from 20 to 2,900 picomolar is harder.

The part that matters most if you're on hormone therapy

Now the centerpiece, and the reason I wanted to write this at all.

Both sweat and interstitial fluid contain only the free hormone. The unbound fraction. Not the total.

Here's why. Most of your testosterone and estradiol travels through your bloodstream stuck to a carrier protein called sex hormone binding globulin. Bound hormone can't cross into sweat, and the capillary wall largely keeps SHBG out of the fluid between your cells.

To put a number on it: in one study, interstitial cortisol measured 0.80 against free plasma cortisol of 1.7 and total of 8.8. About nine percent of total.

So a wearable would be reading something your lab report doesn't show you.

Why that's a bigger deal than it sounds. SHBG has a nearly five-fold normal range in women, 18 to 86. And it moves.

Oral estrogen raises SHBG by roughly 132 percent and drops free testosterone by about 33 percent. Transdermal estradiol moves SHBG about 12 percent and free testosterone barely at all.

Sit with that for a second, especially if you take testosterone alongside estrogen.

If you're on oral estrogen, your total testosterone can hold steady or even climb while the fraction that actually reaches your receptors falls by a third. Your labs look fine. You feel nothing. Nobody's lying to you and nothing is broken. The carrier protein went up and the free hormone went down, and total testosterone, the number almost every clinician orders, cannot show you that.

Which is the catch, and it's a serious one.

Every clinical threshold in menopause and androgen therapy is anchored to a single-timepoint total measurement. There's no reference interval for free estradiol sampled every fifteen minutes. No traceable standard. No number that means "too low" or "enough."

Even if the hardware worked perfectly tomorrow, we'd have a stream of data and no framework for reading it. That's a harder problem than the chemistry, and it's barely being discussed.

Five things worth knowing before you get excited

The devices make you sweat

The Caltech patch doesn't wait for you to perspire. It uses a small electrical current to draw sweat out at rest, then corrects the reading using simultaneously measured pH, salt concentration and skin temperature.

It also needs 60 to 90 minutes of incubation per measurement. Call it one reading a day.

That's a stimulated, calibrated, repeated spot check. The authors are honest about it and describe continuous operation as future work.

Measuring the hormone destroys the sensor

This is my favorite detail in the whole literature, and it has no equivalent in glucose monitoring.

These sensors drift for two reasons. One is ordinary fouling: proteins and lipids gumming up the surface. The other is that the act of measuring strips away the sensing layer, because the voltage pulse used to take a reading gradually desorbs the DNA coating from the electrode.

So sampling rate trades directly against sensor lifetime. A monitor that checks you often kills itself faster.

Which brings us to that 30-day claim.

The 30-day number is a fundraising figure

The best published on-body result for a hormone-sensing patch of this type is five days. The typical published record for the sensor class, in a living body, is under six hours. One human microneedle study ran five hours with what the authors called high rates of device failures.

Thirty days is six times the best result anyone has published, and about a hundred times the norm, for a molecule millions of times more dilute than glucose.

It might happen. There's real work on tougher sensor chemistries. But right now it's a number on an investor page, not a specification, and it should be read that way.

Luteinizing hormone may be the hardest target of all

Estradiol and testosterone are small molecules, around 272 and 288 daltons. Luteinizing hormone is a glycoprotein at roughly 28,000 to 30,000.

About a hundred times heavier.

Size determines how fast something diffuses to a sensor. In one modeling study, glucose reached 90 percent of its true tissue concentration in 20 minutes, while a 21,000-dalton protein reached only 31 percent after two hours.

LH sits above that. And nobody has detected LH in sweat at all. It's been read in interstitial fluid, in rats, more slowly than the steroids.

That matters for ovulation prediction specifically, because the LH surge is the signal people most want to catch.

Nothing is FDA cleared, and the reason is stark

No device is cleared to continuously measure circulating estradiol or testosterone. That's not a backlog. There is no FDA product category for such a device. It doesn't exist as a regulatory concept yet.

All 56 cleared estradiol tests and all 68 cleared testosterone tests are Class I laboratory instruments. Every one of them.

One precision worth having, because it's the kind of thing that gets challenged: over-the-counter urine LH ovulation kits are cleared. So "there's no cleared hormone device" would be wrong. The accurate sentence is that nothing cleared measures circulating hormone continuously.

What I'd actually want one of these for

None of the above means the idea is bad. It means the useful version is narrower than the marketing.

If a device like this worked and were properly validated, the questions I'd want answered are unglamorous and specific:

• Is my patch still delivering on day three, or dropping before I change it?
• Does absorption change when I apply it somewhere different?
• Does heat or a hard workout change how much I'm getting?
• Where does my testosterone actually sit between injections or doses?
• Do the days I feel flat line up with anything measurable?
• Does that pharmacy switch I didn't ask for behave the same in my body?

Notice these are all questions about variability over time in one person. That's exactly what a stream of data is good for, and exactly what a single blood draw can never answer.

You don't need a validated population reference range to learn that your own level reliably craters on day six. You need consistency, not accuracy in the absolute sense.

That's probably where this technology arrives first: not as a diagnostic, but as a way to see your own pattern. Which is genuinely valuable and much easier to validate.

The honest summary

The technology has crossed a real threshold. Direct measurement of reproductive hormones on a human body is no longer hypothetical.

The evidence, stated plainly:

• Sweat estradiol tracking blood in two women
• Sweat testosterone in one woman
• Four-hormone continuous monitoring in rats
• Reproductive-hormone sweat sensing at UCLA demonstrated in artificial sweat, on a bench
• The strongest confirmed on-body result is a different hormone entirely, cortisol, from 2022

Every one of those is a real accomplishment. None of them is a product, and none should change what you do this year.

For now, the things that actually move how much hormone reaches your receptors are already known and already fixable. The route of your estrogen. Whether your testosterone is measured by mass spectrometry rather than an immunoassay that can read more than 100 percent high at female concentrations. Whether anyone looked at your SHBG at all.

None of that requires a patch. All of it requires a clinician who'll engage with the details.

I'm a medical technologist, not a physician. I read the primary literature and I check the study design, and what I've laid out here is the evidence as it stands in July 2026, not a recommendation. Decisions about your own hormones belong with a clinician who knows your history.

But when someone tells you a continuous hormone monitor is nearly here, you now know the right question. Not whether the sensor can detect the hormone. Whether anyone has established what the reading means.

Send this to someone whose labs came back "normal."

Sources

Draft written 25 July 2026 for testosteroneinwomen.com. Rewrite of a ChatGPT-authored article, incorporating every correction from the accompanying fact-check. Nothing here is medical advice.