The promises and pitfalls of personalized health

This is Optimizer, a weekly newsletter sent every Friday from Verge senior reviewer Victoria Song that dissects and discusses the latest gizmos and potions that swear they’re going to change your life. Opt in for Optimizer here.

A few days ago, my esthetician was smearing hot wax on my face. The two caterpillars I call eyebrows were in desperate need of taming — as was my lady ’stache. I hate this monthly ritual, but facial hair is a sore spot. Hirsutism is perhaps one of the few visual indicators of a condition that’s plagued me for the past decade. Until this week, I’ve always known it as polycystic ovary syndrome (PCOS).

Normally, I spend waxing sessions chattering away about the weather (it’s sort of hard to have deep conversations when someone is ripping hair off your face). But that day, we spent the entire session talking about how the global medical establishment decided this week to rename PCOS to polyendocrine metabolic ovarian syndrome, or PMOS.

There are several reasons why. Despite the original name, PMOS — which affects roughly 170 million, or one in eight, women worldwide — often doesn’t result in ovarian cysts. The updated name more accurately reflects how it’s both a hormonal and metabolic condition, not purely a reproductive one. The reality is that the condition can impact multiple organs and is associated with other health conditions, like insulin resistance, Type 2 diabetes, obesity, cardiovascular disease, and obstructive sleep apnea. According to The New York Times, focusing the name on one symptom of the condition — ovarian cysts — has led to inadequate clinical training, poorer research funding, delays in diagnosis, and fragmented care for people suffering with PMOS. In my experience, doctors have often told me that PMOS is benign and decline to offer treatment unless I want to actively pursue pregnancy.

As it turned out, my esthetician also has PMOS. Except where she has ovarian cysts, I don’t. I have insulin resistance; she doesn’t. I struggle a bit with hirsutism, while she lamented about cystic acne. We both put on roughly 60 pounds out of nowhere, but she was able to get it under control through intermittent fasting, a specialized diet, and supplementing with milk thistle and myo-inositol, a type of carbohydrate that helps improve insulin sensitivity. Metformin — a diabetes drug that’s used off-label to treat PMOS — did absolutely nothing for her, while it’s an effective part of my treatment along with a GLP-1.

I’ve had this conversation with so many fellow PMOS sufferers over the years. We always marvel at how the same condition can manifest in such wildly different ways. And while we often swap tips, I’ve never heard of a single treatment working for everyone I’ve ever met with PMOS. Personally, I’ve spent roughly 12 years at this point trying to get mine under control. It’s the reason I first turned to wearables and health tech.

And it’s the first thing I think about whenever a health tech company or wellness influencer pitches the idea of personalizing health.

May is usually when my calendar starts filling up with meetings with the companies on my beat. It’s the start of several conversations about what new products and features are on the docket, where they see this category evolving, and how they’re thinking about current wellness trends. Of the half dozen I’ve had so far this year, there’s one concept that keeps coming up time and time again: personalized health.

All the companies I’ve spoken with agree that teasing out useful insights from a mountain of health data is confusing. But I’m repeatedly told that by being smart about data, the holy grail is personalized health — meaning recommendations based on your individual health metrics versus generalized advice. A simple example might be if your heart rate variability metric shows good recovery, but you had a lousy night of sleep, a wearable device might recommend a moderate 20 minutes of vinyasa yoga as opposed to a high-intensity interval training workout. If you diligently log your meals, perhaps a health AI will identify that the roasted broccoli in your lunch is interacting with a medication and is the likely culprit behind your daily battle with sulfur burps. Depending on your CGM and blood test data, maybe your fitness tracker will one day tell you what supplements to take. And then, I’m often told, you’d have a lot more agency over your health.

It’s an alluring premise. Health is highly individualized. Your genetics play a large role in what medications work for you, what conditions you’re predisposed to, and even what sports your body might be best suited for. I can understand why health tech companies, big and small, are excited by this concept. But again, I have to point to my journey with PMOS.

For better or worse, most doctors will tell someone with PMOS to try losing weight. The logic is that weight loss generally improves, or potentially reverses, symptoms. The only problem there is that it’s a catch-22. Weight gain with PMOS is often related to insulin resistance, which in turn makes the simple “calories in, calories out” logic much harder to enact because it triggers a vicious cycle. Higher insulin levels trigger excess androgen production, which in turn causes PMOS sufferers to store more abdominal fat. Studies have also shown that PMOS sufferers tend to have lower basal metabolic rates. Meaning, everything else equal, they burn fewer calories a day than people without PMOS. The condition may also contribute to complications in building lean muscle mass.

I have yet to find any fitness or nutrition feature that takes these factors into account. There’s no button I can press to adjust recommendations or calorie burn estimates. (This is one reason why I don’t recommend folks put too much stock into calorie tracking with fitness trackers.) The same often holds for reproductive health tech features. When it comes to predicting fertile windows, I usually can’t even test these features because the algorithms aren’t equipped to deal with hormonal birth control — another common treatment for PMOS (and for people with all sorts of other conditions!). Do these features account for how oral hormonal contraceptives can alter body temperatures? Each health tech maker has their own proprietary algorithm, so you’d have to ask each one.

Perhaps this will be possible one day. Wearable tech is helping researchers discover new correlations between biometric data points, particularly in reproductive health. But as of right now, there are few personalized modes for people who deviate from the “norm” that algorithms are built for. If you happen to fall into that category, “personalized health” often ends up being a series of ad hoc solutions you’ve cobbled together yourself.

Some of this is a timing issue. Generative AI is still relatively new, and companies are figuring out the limitations in real time. Plus, the human body remains, in large part, a mystery. It’s hard to offer truly personalized health tech if even the medical experts are baffled by certain conditions. Perhaps most damning, good science takes a long time, whereas the push is for technology to move fast. PMOS was first officially identified in 1935, and only in 2026 does the name now accurately reflect the scope of the condition. (Even the renaming took 14 years and over 50 professional medical groups.) And yet, personalized health features are rolling out now. For some people, that might mean a relatively straightforward experience because they don’t have particularly complicated health. For many others, it’s an experience built on a foundation that’s constantly shifting.

For me, personalized health tech currently means a lot of elbow grease. I have to train the AI coach. I have to think deeply about the parameters of my PMOS and non-alcoholic fatty liver disease. I have to decide which metrics I’ve collected over the years are relevant to my needs. I have to do the independent research and subject myself to the torture of consulting multiple doctors. I have to do the trial and error of testing out available treatments. I spent an inordinate amount of time asking experts a lot of questions and then fact-checking the answers.

I’m not trying to rain on personalized medicine’s parade. I’m cautiously optimistic about some health features I’ve gotten sneak previews of. And there have been many times in the past 12 years when I would’ve loved this sort of guidance. (Who knows, it might’ve saved me from listening to that one quack doctor who insisted I take desiccated pig thyroid extract because they’d misdiagnosed me with hypothyroidism.) I’m hopeful that down the line, health tech companies may explore “algorithmic modes” to account for different diagnoses, medications, and other health idiosyncrasies.

All I’m saying is that health tech companies are going to start pitching this concept as it’s possible now. As if it’s just as simple as wearing a device and consulting an AI. And it just isn’t. At least, not in the way you’d likely want — in the background, with minimal effort on your part.

And in the meantime, I guess I’m going to research the milk thistle supplement my esthetician recommended.