PMOS Treatment Is Built for One Phenotype. It's Failing the Other Half.

If you've been told you have PMOS (formerly PCOS, renamed in May by an international consensus group after eleven years of work), you've probably been handed some version of the same protocol. Metformin, or myo-inositol if you'd rather avoid the side effects. Maybe the two together. A lower-glycemic way of eating. Spearmint tea. Maybe berberine.

The underlying logic is consistent everywhere you look: bring insulin down, and the androgens driving your symptoms come down with it.

For a lot of women, that logic holds, and the protocol works, sometimes remarkably well. For a lot of other women, it doesn't move a single symptom, and nobody circles back to ask why. They just get told to try harder: eat cleaner, exercise more, give the same protocol six more months.

I want to make the case that "try harder" is the wrong instruction. Why? Because the protocol was never built for what's actually driving their case in the first place.

I see a particular pattern constantly in clinic, and it usually arrives looking the same way. A woman has been diligent, sometimes for years. Tracking her carbohydrates, cutting sugar, walking every day, taking the inositol religiously. Her cycle is still irregular. Her skin and hair symptoms haven't shifted. Her labs come back with a normal fasting glucose and a normal HOMA-IR, which she's told means she's "just one of the harder cases," rather than a sign that the entire premise of her treatment plan might be wrong.

Nobody re-examines the premise, because the premise was never stated out loud in the first place. It was baked into the prescription pad before she walked in the door.

Two Androgen Pathways, One Diagnostic Label

PMOS is diagnosed on the Rotterdam criteria. Two out of three createria must be met: irregular ovulation, high androgens, and/or polycystic ovarian morphology. What that framework doesn't ask is where the androgens are coming from.

Androgens in the body get made in two different places - the ovaries and the adrenal glands. And the pathway matters enormously for what will actually treat it.

The far more common route is ovarian. Insulin resistance drives the ovarian theca cells to overproduce androgens. A large systematic review of clamp studies (the gold-standard way of measuring insulin sensitivity directly, rather than inferring it from a fasting number) confirms insulin resistance is present in the substantial majority of women with the condition, though how large a majority depends heavily on body composition. One comparison of lean and overweight women with PCOS found insulin resistance in roughly 57% of the overweight group, against roughly 9% of the lean group, while other cohorts have put both figures considerably higher.

The range in the literature is wide enough that I'd be doing you a disservice quoting a single clean percentage as gospel. What's consistent across the data is the direction: heavier body composition tracks with a much higher likelihood of insulin resistance sitting underneath the androgen picture, though plenty of lean women carry it too, and plenty of larger-bodied women don't.

The less common, far less discussed route is adrenal. This is where I see DHEA-S elevated, testosterone unremarkable, ovarian morphology sometimes present and sometimes not, insulin sensitivity often completely normal, occasionally even higher than average. I wrote about my own case landing in this category in the piece on PMOS testing. I won't repeat that story here beyond saying it took years to get an accurate read on my own bloodwork, simply because nobody was looking for an adrenal pattern in a diagnosis everyone assumes is ovarian by default. Adrenal androgen excess has been shown to track with a more favourable insulin and lipid profile than presentations built around normal adrenal androgens, which is close to the opposite of what most PMOS treatment assumes walking in the door.

Yet, the clinical picture the two produce can look almost identical from the outside. Irregular cycles, acne along the jaw, unwanted hair growth, hair thinning at the temples. That overlap is exactly why it's so easy for a clinician working from symptoms alone to assume ovarian by default and reach for the standard protocol without a second thought. But the two pictures usually separate once you look past the symptom list and into the labs themselves.

Ovarian-driven PMOS tends to come with the metabolic markers you'd expect from insulin resistance. These include a raised fasting insulin relative to glucose, a less favourable lipid pattern, and often (though not always) higher body weight. Adrenal-driven PMOS tends to travel without them, sometimes alongside a history that reads as more psychological or circumstantial than metabolic. Think chronic stress, disrupted sleep, a nervous system that's been switched to a survival setting for years rather than a diet or activity pattern that would explain elevated androgens on its own.

Neither history is more legitimate than the other. They're just pointing at different organs.

Why the Standard Protocol Is Built the Way It Is

None of this makes metformin, inositol, or a lower-glycemic diet bad advice. They're good advice for the mechanism they're built to treat. Head-to-head trials of myo-inositol against metformin both work by the same route. They improve how efficiently cells respond to insulin, which lowers circulating insulin, which removes the signal telling ovarian theca cells to keep manufacturing androgens. Spearmint tea's anti-androgenic effect and a lower-glycemic diet's benefit both run through variations on the same insulin-mediated pathway. It's a coherent, evidence-based approach to the ovarian phenotype, and I'd recommend most of it to a woman I'd confirmed was insulin resistant.

The problem isn't the protocol.

It's that it gets handed out by default to every PMOS diagnosis, insulin resistant or not, because the diagnostic criteria never asked the question that would tell you which one you're looking at.

What Happens When You Treat the Wrong Mechanism

If your androgens are adrenal rather than ovarian, and insulin was never the driver, an insulin-focused protocol has nothing to act on. At best, it does nothing, and you're told to keep trying. At worst, it actively adds load. A woman who is already insulin-sensitive, restricting carbohydrates hard on a low-glycemic protocol that was never built for her, on top of the caloric tightening that so often accompanies a PMOS diagnosis, can end up recreating exactly the energy-scarcity signal I wrote about in the piece on cycles shortening before they stop. Instead of correcting the adrenal picture, the protocol adds another stressor for an already taxed HPA axis to absorb.

That's because adrenal androgen production doesn't run through insulin at all. It runs through the hypothalamic-pituitary-adrenal axis. ACTH, released from the pituitary in response to the brain's read on stress, stimulates the adrenal cortex to produce cortisol and DHEA together, and under sustained load, the adrenal glands adapt their steroid output in ways that can persistently favour androgen production alongside cortisol. Chronic, unresolved stress, not carbohydrate intake, is the upstream driver. Bringing insulin down further doesn't touch that axis at all. You can lower a hormone that was never elevated and leave the actual mechanism running exactly as it was. In fact, as I learnt when I tried a ketogenic diet back in 2017, sometimes it can leave it worse. In my case, and in many of the clients I’ve worked with who have had similar experiences, the intervention itself became one more demand on a nervous system that was already over capacity.

It's worth being honest about how messy this gets in a real body rather than a textbook diagram. Cortisol and DHEA don't always move together in a clean, predictable ratio. The relationship between ACTH stimulation and the two hormones can dissociate under pathological or sustained-stress conditions, which is part of why a single cortisol reading is a poor stand-in for a genuine adrenal workup. A woman can have a morning cortisol that reads as unremarkable and still be running an adrenal androgen pattern driven by an axis that's stopped behaving predictably under years of load. This is a reason for more careful testing, not a reason to abandon the adrenal hypothesis the moment one number looks fine.

This is the same blind spot I wrote about in the RED-L piece: a framework built to catch one mechanism (there, energy availability and here, insulin resistance) simply doesn't see a different mechanism producing an overlapping clinical picture. Two women can walk in with the same Rotterdam-defined diagnosis and be running two almost entirely separate physiologies underneath it.

What Actually Helps Adrenal PMOS

If your DHEA-S is elevated, your testosterone is unremarkable, and your insulin markers already look fine, here's where I'd actually put the effort, in rough order of leverage.

Get the right panel before changing anything. DHEA-S, testosterone (total and free), SHBG, and prolactin drawn in the early follicular window, alongside fasting insulin and glucose, so you have an honest baseline rather than an assumption. If ACTH and a morning cortisol are accessible, they help confirm the adrenal axis is genuinely the active driver rather than one part of a more mixed picture. I've written the full logic of correct cycle-day and diurnal timing in the testing piece, so I won't repeat it here, but it applies just as much to this panel as to the standard one.

Address exercise intensity directly, because more isn't automatically better here. Exercise above roughly 60% of maximal oxygen uptake, sustained for more than ten to fifteen minutes, reliably raises cortisol above resting levels, and for a nervous system already running an overactive HPA axis, daily high-intensity training can be adding to the exact load driving the androgen picture, not counteracting it. This doesn't mean stop moving. It means matching intensity to what your system can currently absorb. Activities like walking, strength work at a moderate pace, yoga, and swimming are often better options, rather than defaulting to HIIT seven days a week because that's the fitness-industry answer to "PMOS."

Support the adrenal cortex's actual nutrient requirements. The adrenal cortex holds one of the highest concentrations of ascorbic acid of any tissue in the body, and vitamin C functions as a required cofactor in adrenal steroidogenesis, with ACTH stimulation shown to deplete adrenal ascorbic acid stores by roughly a quarter to a third. Magnesium plays a comparable structural role: magnesium deficiency has been shown to increase CRH transcription and drive HPA axis hyperactivity, and magnesium status is now understood to sit in a genuinely bidirectional relationship with chronic stress. What I mean by this is that each depletes the other. Food-first where you can, supplementing where testing or symptoms suggest a gap.

Consider adaptogens with actual trial evidence behind them, not the whole aisle. A randomised, placebo-controlled trial of ashwagandha in adults with chronic stress found a significant reduction in serum cortisol relative to placebo, a finding a subsequent meta-analysis of nine trials has since supported. Holy basil (Ocimum tenuiflorum) produced measurable reductions in hair cortisol, a marker of cumulative HPA output over weeks rather than a single moment, over an eight-week trial against placebo. These aren't interchangeable with each other or with rhodiola. I also want to point out that none of them substitutes for addressing whatever is actually generating the chronic stress load in the first place. They're support for the system while you do that work, not the work itself. I rarely use them in clinic for this reason, other than for very short-term support - I’m more interested in addressing why your stress is a 9/10 every day than suppressing your body’s natural response to it.

Keep blood sugar reasonably stable anyway, for a different reason than the standard protocol assumes. This isn't a contradiction of everything above. Even when insulin resistance isn't the driver of your androgens, sharp glucose swings are themselves a stressor the HPA axis has to respond to. A body already running high cortisol output doesn't need another reason to release more of it. The goal here isn't the aggressive carbohydrate restriction built for the ovarian phenotype. It's ordinary, unglamorous blood sugar steadiness. Think protein and fibre alongside carbohydrates and regular meals rather than long stretches undereaten followed by a big evening meal. We’re aiming to prevent the adrenal axis from fielding a glucose emergency on top of whatever else it's managing.

Protect sleep and circadian rhythm specifically, not just generically. Cortisol output follows its own daily rhythm, highest shortly after waking and tapering across the day, and that rhythm is one of the more sensitive things chronic stress and poor sleep disrupt. Late nights, inconsistent wake times, and screen exposure that delays the evening cortisol decline all keep the HPA axis firing at times it should be winding down. A consistent sleep and wake window does more for this pattern than most people expect from something that costs nothing and requires no lab work to start.

Do the nervous system work the labs can't show you. This is the piece I think gets skipped most often because it doesn't fit neatly into a supplement protocol. I wrote in more depth about the vagus nerve's relationship to hormonal regulation here. The same logic applies directly to an overactive HPA axis: a nervous system that never gets a genuine signal of safety keeps the ACTH-driven cascade running regardless of what's on your plate or in your supplement drawer. Sleep, downregulation practices, and, where relevant, the kind of deliberate stress-reduction work I described going through myself in the cycle-changes piece tend to move adrenal androgens more than any single nutrient or herb does on its own.

The Diagnosis Was Never Supposed to Be the End of the Investigation

PMOS covers at least two distinct mechanisms wearing the same label, and a diagnostic framework built around ovarian morphology was never going to reliably tell you which one you've got.

If metformin, inositol, and a low-glycemic diet have done nothing for you after a genuine trial, that isn't evidence you need to try harder on the same plan. It's a reasonable signal to ask whether anyone has actually looked at where your androgens are coming from.

If you'd like a proper look at your own pattern, insulin markers and adrenal markers read together rather than assumed from the diagnosis alone, I work with women one-on-one to figure out which mechanism is actually driving their case before building a plan around it.

I also put together a guide that walks through how to read a full hormone and metabolic panel properly, including the adrenal markers most standard PMOS workups leave out entirely.