Women are, for the most part, their own epidemiologists.
If you watch what actually happens when a woman develops a hormonal condition, what you're watching is an enormous amount of unpaid research, conducted in parallel, on phones and in notes apps and in Reddit threads at two in the morning, by people who never volunteered for the job. They track what they eat. They switch shampoos. They rotate through supplements. They log symptoms, usually somewhere they know they'll lose them. They form sharp, specific hypotheses and then try to test those hypotheses against a body that does not come with a control group.
Almost none of this gets captured anywhere. It doesn't make it into journals. It doesn't shape the next version of the medical guidelines. It usually doesn't even make it across the kitchen table to the next woman who'll need it, because by the time she needs it the first woman has moved on, or stopped talking about it, or is tired.
I find this idea unreasonably interesting, in part because I spent three years being one of them.
The thing that brought me into it was pain. I'd been diagnosed with PCOS in my early twenties, told to come back in a year, and through what I would now describe as accidental lifestyle intervention, mostly got better. A few years later a different kind of pain started, one that did not respond to any of the things that had fixed the first one. It took, depending on which appointment you count, six or seven years to be handed the word "endometriosis." Before that, at various points, it was IBS, and stress, and normal, and maybe try yoga.
I was, by the point someone said it, reasonably well read and working in tech. If I'd found it hard, the women I couldn't stop thinking about were the ones who'd been told the same things for twice as long, or who never got in front of a search engine to start sorting through the mess in the first place.
What I did in the meantime is what most women in this situation do. I kept a notes app. When the notes got too long I started taping pages of a notebook together because I'd run out of space. I spent an amount on supplements that I would now describe, being generous, as a small car. I read peer reviewed papers I was not really qualified to read. I kept the tabs open. I joined the Reddit threads. I compared myself to strangers, and at some point decided that what I was actually doing was a part time research job, unpaid, whose only deliverable was me.
The thing about hormonal conditions that is hard to explain to people who don't have one is that the dominant cost isn't the cost of the condition. It's the cost of the research you have to do in order to live around it.
The conventional story about why this hasn't been fixed is that the research isn't there. That is mostly wrong. Most of what you'd want to know about how a given ingredient affects a given condition is already on PubMed, for free. The problem is that it's in the wrong format. It's written for researchers, scattered across thousands of studies, often contradictory on the surface, and essentially impossible for a tired woman at ten in the evening to read. Translating that research into something usable is an enormous piece of work. The reason it hasn't been done is that the people equipped to do it and the people who need it done have, until recently, been almost entirely disjoint populations.
This is not a dramatic claim. It shows up in the data. Until about five years ago most major trials in cardiology, endocrinology and pharmacology under-represented women to the point of structural uselessness. The endocrine disruptor literature is full of studies that compared two men to two other men. The shape of the whole field is, roughly, research done by and mostly on people who do not have the bodies the research is supposed to be about. It's hard for that not to produce blind spots.
The useful consequence of having been in this situation, as both a patient and a founder, is that you end up with a very specific picture of what to build.
Not a scanner app. Scanning is the front door. Behind the front door is the thing that was actually missing, which is an instrument that captures what women are already doing by hand and makes the research they can't read available to them at the level of a single bottle in a single basket. What did I eat. What did I put on. What did my cycle do this week. What does the evidence say about any of this for a body like mine. Put like that it sounds obvious. The reason nobody has built it isn't that it isn't obvious. It's that the people it's obvious to and the people who build software have, until recently, not been the same people.
Embyr is the attempt. It's evidence linked, condition personalised, and accountable to you in the specific sense that every flag on every product ties back to the study it came from. We scan food, skincare and supplements. Underneath, a research team does the translation work I used to try to do on my own, at night, badly.
I think the class of problems where the evidence is public and the translation is missing is one of the more interesting places to be working right now. It looks, from the outside, already solved, because all the raw material is lying around. You can Google it. You can find the studies. The gap between "exists" and "usable" seems, at first glance, small. It isn't. It's the whole problem, and it's a gap outsiders systematically underestimate, because the size of a gap is usually only visible to the people trying to cross it.
If any of this sounds like something you have been quietly thinking about, or if you have your own version of the notes app, I'd like to hear from you. catie@embyr.health.
Catie