The endometriosis numbers worth tracking — and what they can and can't tell you

To make sense of why these particular markers, it helps to hold the model of the disease in mind. An endometriotic lesion is sustained by a handful of reinforcing inputs — estrogen that the lesion is exquisitely sensitive to, the insulin and IGF-1 that amplify its growth, the liver's capacity to clear estrogen once the body has used it, and the systemic inflammation that surrounds active disease. The useful thing about that model is that each of those inputs leaves a fingerprint in the blood. The panel that follows operationalizes the biology across four domains, and it is best understood not as a checklist of targets to hit but as a set of dials you and a clinician can watch turn over time. That framing is not a soft caveat; it is the entire correct use of the panel, and the reasons will recur because they have to.

The first domain is estrogenic drive and its progesterone opposition, and the cardinal error here is reading either hormone alone. Estradiol is the primary estrogenic signal, and in endometriosis the relevant problem with estrogen is that it is proliferative and pro-inflammatory for ectopic tissue when nothing opposes it. What opposes it is progesterone, the physiologic brake on estrogen-driven proliferation. So the marker that actually matters is the relationship between the two — the balance, read in the luteal phase where progesterone should be ascendant — not estradiol in isolation and not progesterone in isolation. This is the precise meaning of the phrase "estrogen dominance," which is widely misused: it denotes a relative imbalance, estrogen insufficiently opposed by progesterone, not an absolute excess of estrogen. A person can have an unremarkable estradiol and still sit in a functionally estrogen-dominant state because the progesterone opposing it is low. As representative functional ranges, a framework like this might describe mid-cycle estradiol in the region of 80 to 150 pg/mL and luteal progesterone in the region of 10 to 20 ng/mL — but those figures are non-diagnostic functional targets, deliberately narrower than population reference ranges, and stated here only to make the pairing concrete, not as numbers to chase.

Sitting beside the estrogen-progesterone pair is SHBG, sex hormone-binding globulin, which sets how much of the circulating estrogen is actually free and bioavailable. A high-binding-protein state keeps more estrogen bound and quiet; a low one releases more free estrogen into tissue. And SHBG is where the first and second domains connect, because the thing that most reliably lowers SHBG is insulin. This is why SHBG is not read alone either — it is read against fasting insulin. When insulin runs high, SHBG falls, free estrogen rises, and the same hyperinsulinemia independently drives the PI3K–AKT–mTOR proliferative axis that the lesion exploits. So the pairing of SHBG with fasting insulin tells a single story about free-estrogen exposure and insulin resistance together, which neither marker tells on its own. A framework of this kind might point toward SHBG somewhere around 70 to 90 nmol/L and fasting insulin in the low single digits, roughly 3 to 5 µIU/mL — and the same warning applies in full: these are functional-optimization targets used within a model, not validated thresholds, and several of the associations behind them are drawn more firmly than the formal evidence yet supports.

The third domain is the one most people never think to look at — whether the body is clearing estrogen efficiently once it has done its work. Estrogen is not only produced; it is metabolized and excreted, and that clearance runs substantially through the liver, through phase I and phase II conjugation and through methylation. Two markers act as proxies for that machinery. Homocysteine stands in for methylation capacity: when it rises, it signals that the B-vitamin-dependent, MTHFR-influenced methylation pathways are strained, and methylation is what allows the COMT enzyme to inactivate the catechol-estrogens the liver produces. Free T3 stands in for the thyroid contribution: adequate active thyroid hormone supports hepatic conjugation and clearance, and when it is low, clearance slows and estrogen is more likely to be reabsorbed through the gut. Read as a pair, homocysteine and Free T3 describe two arms of the same clearance system — methylation on one side, thyroid-supported hepatic processing on the other. A framework might flag homocysteine below roughly 7 µmol/L and Free T3 toward the upper part of its range, but once again these are non-diagnostic functional markers, and the link between, say, a homocysteine value and estrogen clearance in an individual is a directional rationale, not a validated diagnostic equation.

The fourth domain is systemic inflammation, the background tone against which active disease burns, and here the instructive pairing is hs-CRP with ferritin. High-sensitivity CRP is a direct readout of systemic inflammatory tone; values drifting upward reflect the inflammation characteristic of active disease. Ferritin is more slippery, and that is exactly why it is paired with CRP. Ferritin is usually read as an iron-status marker, and at the lower end it does reflect iron sufficiency for tissue and mitochondrial repair. But ferritin is also an acute-phase reactant, which means that when it climbs high it may be reporting inflammation rather than iron stores at all. Without CRP beside it, a high ferritin is genuinely ambiguous; with CRP beside it, the ambiguity resolves — a high ferritin alongside a high CRP is inflammation wearing an iron-marker's clothing. Representative functional framing might place hs-CRP below about 0.5 mg/L, with values above roughly 1 mg/L taken to indicate systemic inflammation, and ferritin around 60 to 90 ng/mL at the sufficiency end, with high readings above roughly 140 reflecting inflammation rather than stores. The non-diagnostic caveat holds here as everywhere: these are optimization targets within a framework, not laboratory cut-offs.

Vitamin D sits a little apart from the four pairs because it plays two roles at once. It is a modifiable input — repletion is straightforward and within a person's control — and it is also, separately, a severity marker, in that lower levels associate with greater disease severity. That dual character is worth naming precisely because it is the kind of association that is easy to overstate: a correlation between low vitamin D and more severe disease is real enough to track, but it should not be read as proof that the deficiency caused the severity or that correcting it will reverse anything. It is a dial to watch and an input to address, held with appropriate humility about what the association does and does not establish.

That humility is the through-line, and it has to be stated more than once because the temptation to treat these numbers as authoritative is strong precisely when a person is frustrated and looking for something concrete to fix. So, plainly: every range mentioned above is a functional-optimization target used inside a particular model, not a validated diagnostic reference interval and not a cut-off that defines disease. They are deliberately narrower than the population ranges your lab prints, and several of the biological associations behind them are stated, in frameworks like this one, more confidently than the formal evidence has yet earned. The hormone values shift with the cycle — estradiol around mid-cycle, progesterone in the mid-luteal phase — so a single mistimed draw can mislead entirely. And none of this is a self-diagnosis tool. A blood panel does not diagnose endometriosis; clinicians and surgeons do, with clinical evaluation and, where needed, imaging and laparoscopy. The panel cannot replace that, and reading it as if it could is the failure mode to avoid.

What the panel can defensibly do is narrower and more honest. Drawn as a baseline and repeated at intervals, interpreted in the pairs described — estradiol with progesterone, SHBG with insulin, homocysteine with Free T3, hs-CRP with ferritin — and read by someone who can hold the cycle timing and the clinical picture together, it lets a person and their prescribing provider watch the direction of change over time in that one body. Falling free estrogen, falling insulin, a normalizing CRP, improving clearance markers: those trajectories are the biochemical correlates of biology moving in a favorable direction. The word that matters there is direction. The panel is a way of asking whether the underlying biology is shifting over months, not a scoreboard to optimize in isolation and not a set of numbers to chase on your own. Used that way — in pairs, with cycle timing, under clinician guidance, as a record of trend rather than a verdict — it earns its place.

Which leaves the real value of looking at these markers at all, and it is less about any single result than about a change in how the disease is understood. To track estrogen against progesterone, insulin against SHBG, methylation against thyroid, and inflammation against iron is to stop treating endometriosis as a fixed anatomical fact and start treating it as a biology with inputs that move — some of them modifiable, all of them worth watching with a clinician who knows what the trend means. That reframing is the quiet utility here. The numbers were never going to be the answer; what they offer is a way of asking better questions about a body over time, which is a different and more durable thing than a target to hit.