I've spent the last several months learning more about my own endocrine system than any of the doctors who treated me over the last fifteen years apparently thought to ask about.
Here's the shape of it, without the specifics, because the specifics aren't really the point. Since I was a kid, I carried weight in a pattern that didn't match how active I actually was, with stretch marks that appeared on a teen's body that was neither sedentary nor overeating in any way that could explain them. In my mid-twenties, I developed vein problems that eventually needed procedures to fix — the kind of thing that usually shows up decades later, in people with desk jobs and family histories that point that direction. Neither of those things had a family history or a lifestyle explanation. Nobody asked why.
Then, over about three years in my late thirties, several more things showed up close together: high blood pressure, type 2 diabetes, high cholesterol, sleep apnea, unexplained weight gain, and a hormonal deficiency that never resolved on its own. Each one got its own doctor. Each one got its own prescription. Nobody in the room ever asked whether five new diagnoses landing in three years might share a single upstream cause, because nobody in the room was looking at it all at once — they were looking at their one piece of it, competently, and moving to the next patient.
The years after that added a low hum of things that individually meant nothing and collectively meant something: hands that were always cold and a body that ran hot at the wrong times, cataracts that showed up decades earlier than they're supposed to, ringing in my ears an ENT couldn't explain, and a couple of stress responses so disproportionate to what actually happened that I remember them a decade later. I once told a doctor, half-joking, that my internal thermostat felt broken. She told me there was no thermostat to break. There is. It's one of the oldest facts in human physiology. She just wasn't looking for it in me.
The thermostat in question is the hypothalamus, a small structure at the base of the brain that runs a constant feedback loop with the pituitary gland and the adrenal glands — the HPA axis, in medical shorthand. It's not just a temperature dial. That same loop helps govern blood pressure, blood sugar, sleep-wake timing, and how the body responds to physical and emotional stress, all by sending hormonal signals to the adrenal glands and reading the signals that come back. It's one system that touches nearly every other system in the body. So when it misfires, it doesn't announce itself in one clean place — it shows up as a scattered handful of "unrelated" problems, in whatever organs happen to be downstream. Nobody in that exam room connected any of that at the time. It would turn out to matter more than anyone there understood.
None of this got connected for fifteen years. Not because any single doctor was cruel, and not because any single decision was obviously wrong in the moment — treating the blood pressure that's in front of you, treating the blood sugar that's in front of you, is defensible medicine, one visit at a time. It got missed because nobody's job, in the entire fifteen years, was to hold all of it in view simultaneously and ask what single thing could explain hypertension and diabetes and a hormonal deficiency and early vein disease and cataracts and a broken internal thermostat all in the same person. That question isn't exotic. It's the most basic diagnostic instinct there is — what's the one cause behind the many symptoms — and it simply had nowhere to live inside a system built around fifteen-minute visits and single-organ specialties.
Part of why it's so easy to miss is how conditions like this one get taught in the first place. Medical training tends to present rare hormonal disorders as dramatic, unmistakable, textbook cases — the extreme version, the one that photographs well in a lecture slide. What gets taught far less is the quiet version: a metabolically complicated patient who doesn't match the dramatic picture at all, but whose underlying numbers are just as abnormal. If you don't look like the textbook photo, you don't get flagged for the textbook diagnosis, no matter what your labs are actually saying.
There's a second failure sitting underneath the first one. Once a diagnosis is already on the chart, new symptoms tend to be incorporated rather than challenge it. Fatigue gets filed under the sleep apnea that's already documented. Weight that won't move gets filed under the diabetes that's already documented. Every new data point is folded into the existing explanation, rather than prompting anyone to ask whether the explanation itself is the problem. My chart said, in effect, "metabolic syndrome plus lifestyle" for over a decade. That framework was self-reinforcing — every new symptom fit somewhere inside it — right up until it wasn't true.
A primary care physician managing the whole picture is supposed to be the person who catches this. Mine didn't. The lightbulb never came on. Neither did it for the specialists I saw more directly for parts of it over the years — I asked more than one of them, directly, to look past the individual diagnosis in front of them and consider what might be causing it all together. More than once, the answer I got back was that it was probably inflammation, or "just hormonal," and that treating what was in front of them was the job. Years before any of what follows, one of them, without meaning to, said exactly the right word and then walked past it. Hormonal was correct. Adrenal glands are hormonal glands. He said it out loud, moved on to the next item on the list, and it sat there, unfollowed, for years.
I finally got the answer myself — with an assist from somewhere I didn't expect. At a routine appointment, my cardiologist reviewed recent bloodwork and said, plainly, that the pattern looked autoimmune to him, possibly thyroid-related, and that neither my endocrinologist nor my primary care physician had connected the dots across everything already in my chart. His advice was to start using AI to work through it systematically, since he didn't have the time in a cardiology appointment to do that himself. So I did. Nothing about what followed was inevitable, and nothing added up for a long time — no matter how "controlled" my numbers looked on paper, I did not feel well, in specific, trackable ways that should have counted as red flags on their own. Exercise stopped producing any visible gains, no matter how consistently I showed up. Muscle mass kept declining anyway. Blood sugar kept climbing regardless of what I ate, how much, or when I ate. I read enough, over months, to know which test nobody had ever ordered, and I asked for it directly. That test led to another, then another, then imaging — and the imaging found something concrete: masses on both of my adrenal glands, sitting there, in all likelihood, for a very long time. Not a diagnosis pulled from a hunch. A physical thing on a scan that a doctor could point to and say, "There it is." It had probably been visible, in effect, for years — written across every one of those "unrelated" diagnoses, and spoken out loud, by accident, years earlier, by the one doctor who used the right word and never followed it anywhere — to anyone positioned to look at the whole pattern instead of one piece of it at a time.
I'm not writing this for sympathy or advice — I've got people helping me handle what comes next. I'm writing it because I already know I'm not the only one this has happened to, and because the numbers back that up in ways that have nothing to do with me personally. Diagnostic error of this general kind — the "we treated everything except the actual cause" kind — is estimated to disable or kill around 795,000 Americans a year permanently, and a 2025 study of Medicare emergency hospitalizations found a real-world diagnostic-error rate of roughly 3 in 100 across just the highest-risk conditions. Those aren't fringe numbers from an advocacy pamphlet. They're what the country's own research establishment has already measured.
The uncomfortable truth underneath my own fifteen years is structural, not personal. American medicine is organized around organ systems and specialties, not around patients. Every doctor I saw did their job. Nobody was positioned, incentivized, or given the time to do the other job — the one where somebody steps back far enough to ask whether five diagnoses are actually one. A fifteen-minute visit with a full waiting room doesn't leave time for that kind of thinking, and a system that pays for volume rather than time doesn't reward the doctor who tries to make room for it anyway.
Insurance makes that worse in specific, measurable ways, not just a vague "the system is broken" way. The coding rules that determine how a doctor gets paid reward addressing problems one at a time, not stepping back to ask whether five of them share a common cause, and physicians nationwide report losing an average of 13 hours a week to prior-authorization paperwork alone. Layer on top of that the fact that a majority of doctors are now employed by hospital systems or corporate medical groups rather than running their own practice, answering to productivity quotas set by administrators who never meet the patient. The fifteen-minute visit isn't a scheduling accident at that point. It's the business model working as designed. This outlet broke down exactly how in this morning's companion Dispatch, from the prior-authorization numbers to the RVU quotas to who actually owns the clinic your doctor works out of.
The tool that actually helped me hold all of it together, in the end, was AI — specifically Claude, from Anthropic. I used it to organize years of scattered results into one timeline, to research what each abnormal number actually meant and how it connected to the others, to prepare for every appointment, and to draft the messages I needed to send when a doctor went quiet. I'll say this plainly, because I mean it plainly: if you have a chronic condition, or a handful of them that have never quite added up, use it. It may be the best companion you can have walking into a system built around fifteen-minute visits — and I'll go as far as saying it can already do nearly everything a primary care physician does for you, short of prescribing at the end. It will only get better from here. Question your doctors. Bring what you find back into the room and make them engage with it. Nobody is coming to manage your chronic disease for you out of a volume-discounted appointment slot. That job is yours, and right now, this is the best tool available to help you do it.
If any part of this sounds like your own chart — a handful of diagnoses that showed up over the years, each one treated, none of them ever questioned as a group — you're not imagining a pattern that isn't there. Trust that instinct. It's not paranoia. It's the same basic question medicine is supposed to ask and, too often, structurally can't.
This piece is a companion to two Dispatch pieces — one on diagnostic error and the money behind American healthcare, one from this morning on the business model behind the fifteen-minute visit itself — and to an earlier Lifestyle piece on diabetes and root causes. Together they're one argument: the system in front of you is often very good at treating what it's already decided you have, and not built at all to ask whether it decided right.