Eden Openly

Hormone Monitoring & Target Ranges

Evidence Review · ER-008

Identifier: ER-008 · Version: 1.0 · Status: Current · Published: July 2026 · DOI (all versions): 10.5281/zenodo.21274735 · DOI (this version): 10.5281/zenodo.21274736

This review does something the guidelines mostly don’t: it grades the target numbers rather than restating them. The ranges in everyday use — a physiological-female oestradiol window, a testosterone threshold — are widely cited and confidently applied, yet rest on remarkably little outcome evidence. What follows separates the part of monitoring that is on reasonably firm ground (how to measure, and when) from the part that is not (what number to aim at, and why). For the short version, see the companion Evidence Note → (5 min read).

Purpose

To appraise the certainty behind the hormone-monitoring targets used in feminising and masculinising therapy — not to set them. This is not a prescribing or monitoring protocol. It should not be used to start, stop, dose-adjust or change monitoring without clinical supervision; monitoring should follow local standards. Numbers are cited here to be weighed, never to be prescribed.

One distinction runs through the whole review, because collapsing it is where most guidance turns muddy. Monitoring — what to measure, when, and how well the assay performs — is a largely analytical and pharmacokinetic question, and much of it is firm. Targets — what number to aim at, and whether any outcome study justifies it — is a different and far weaker evidence base. The review is built in three parts on exactly that seam: the measurement layer (Part A), the appraisal of the targets (Part B), and the safety surveillance that sits on firmer ground than either (Part C). Ranges quoted are for appraisal; assay- and lab-specific reference intervals always take precedence in practice.

Executive summary

What this covers. Both directions of care, organised by the monitoring question rather than by hormone: what monitoring is for; what to measure (total versus free/bioavailable, SHBG, LH, and the assay problem); when to measure (peak, trough and interval); the SHBG mechanism that makes a single total a weak instrument; the feminising and masculinising targets graded against outcome data; and the safety surveillance that is genuinely useful.

What is reasonably firm. The measurement layer. Free testosterone must be calculated, not measured directly; mass spectrometry outperforms immunoassay at low concentrations; injectable formulations swing across the dosing interval so timing of the draw changes the number; and haematocrit is a real, monitorable safety signal.[5,8,9,14]

What is weak — and confidently asserted anyway. The targets. A 2025 systematic review found that existing data do not support the guideline oestradiol range of 100–200 pg/mL for feminising outcomes or safety; a narrative review found no randomised trials of optimal concentrations at all; and different guidelines pick materially different numbers.[10,11,12] Higher oestradiol, once testosterone is adequately suppressed, has not been shown to improve measured feminising outcomes such as breast development or body composition — testosterone suppression is the better-supported feminising lever, though other endpoints such as bone may differ.[11,13]

The mechanism at the centre. A single total level is a weak instrument because it ignores protein binding: oestrogen raises SHBG, testosterone lowers it, so the same total can mean different bioavailable hormone depending on the direction of care. The “free” figure that would fix this is itself a calculation resting on assumptions — not a hard measurement.[9,17,21]

The honest bottom line. We often know how to measure these hormones better than we know how to validate the target ranges attached to them. This review grades accordingly.

Key take-home messages

  1. Separate monitoring from targets: how and when to measure is largely firm; what number to aim at is not.[10]
  2. The feminising oestradiol target of 100–200 pg/mL is not supported by outcome data; a systematic review concluded exactly this.[12]
  3. Measured feminising outcomes studied to date track testosterone suppression more consistently than higher oestradiol; a higher oestradiol has not been shown to improve outcomes like breast development.[11,13]
  4. Free testosterone must be calculated from total testosterone, SHBG and albumin — direct free-testosterone immunoassays are inaccurate and should not be used.[7,8]
  5. Oestrogen raises SHBG and testosterone lowers it, so a total level can misrepresent bioavailable hormone in opposite directions.[9,17]
  6. Route matters for safety more than the exact level: transdermal oestradiol appears to carry the lowest observed thrombotic risk and ethinylestradiol should be avoided.[16,18]
  7. Haematocrit is the clearest routinely monitorable hormone-linked safety signal in masculinising therapy; most other “targets” are consensus reference intervals, not validated thresholds.[8,14]

Part A — Monitoring: the measurement layer

1. What monitoring is for — and what it can’t do

Monitoring in hormone therapy does two jobs: safety surveillance (catching effects like a rising haematocrit before they cause harm) and titration to effect (adjusting dose toward a goal). The honest complication is that the second job assumes we know the goal number, and for most of what is measured we do not. A laboratory review of the field put it plainly: no study has evaluated optimal serum sex-steroid concentrations in feminising therapy, and the incremental outcome benefit of routine concentration-monitoring once a regimen is stable is uncertain — although many guidelines still recommend periodic monitoring, and levels remain genuinely useful for adherence, absorption, pharmacokinetic timing, unexpectedly high or low exposure, safety context, and a discordant clinical response.[10]

So the useful framing is not “measure everything often.” It is: measure the things that change management (a safety signal, or a level that resolves a genuine clinical question), interpret them against the right reference interval, and resist the temptation to chase a number toward a target that may not mean what it appears to. Titration is ultimately to the person — clinical effect and goals — with levels as one supporting data point, not the destination.[2,10]

Confidence: 🟡 Moderate (safety surveillance) / 🔴 Low (that level-chasing improves outcomes).

2. What to measure — total, free, SHBG, LH, and the assay problem

Total versus free. Free hormone is immediately bioactive — for testosterone, roughly 2–3% of the total — while albumin-bound testosterone is usually counted as bioavailable too, because it dissociates readily; SHBG-bound hormone is the least readily available.[9] Total testosterone is reliable and cheap but does not describe that bioavailable fraction. Guidelines diverge here: the Endocrine Society recommends monitoring total testosterone and oestradiol only (as in cisgender hypogonadism, reserving free/bioavailable for borderline cases), whereas UCSF routinely adds SHBG and albumin to calculate bioavailable testosterone.[1,2,5] Our reading: total as the default, with SHBG and calculated bioavailable testosterone when the total and the clinical picture disagree, or when SHBG is likely shifted (Section 4).

The free-testosterone trap. If free or bioavailable testosterone is wanted, it must be calculated from total testosterone, SHBG and albumin (for example by the Vermeulen equation), which tracks the equilibrium-dialysis gold standard closely.[7] It must not be read off a direct analog immunoassay: the Endocrine Society states those assays are inaccurate and should not be used.[8] This is one of the firmest, most actionable facts in the whole review — and one of the most commonly ignored.

Free oestradiol barely exists as a routine test. Free oestradiol is measured only by equilibrium dialysis in special cases, so total oestradiol remains the practical feminising measurement. The free-versus-total distinction is therefore largely a testosterone story. Two caveats on what an oestradiol number does not capture: it cannot monitor synthetic or conjugated estrogens at all, and oral oestradiol drives oestrone up through first-pass metabolism, which is part of why people report feeling different on different routes at the “same” level.[9,25]

LH as a cheap axis readout. Suppressed gonadotrophins (LH and FSH) can serve as a surrogate that sex-hormone exposure is adequate — particularly relevant for bone in people not on full replacement (for example after gonadectomy), where under-replacement risks bone loss. It is an under-used, inexpensive readout that the fixation on a target oestradiol number tends to ignore.[2,13]

The assay problem. Tandem mass spectrometry is more specific and accurate than immunoassay across a wider range, and is preferable when low concentrations are expected; immunoassay is acceptable and cheaper at higher concentrations.[5] Even SHBG assays drift: one reagent reformulation shifted calculated free and bioavailable testosterone by 12–16% with no reference-range adjustment.[26] The number is only as good as the method behind it.

Confidence: 🟢 High (the analytical facts) / 🟡 Moderate (which fraction to measure in practice).

3. When to measure — peak, trough and interval

Injectables swing, so timing changes the number. This is true in both directions and is routinely underplayed on the feminising side. For injectable testosterone, peak levels can be measured 24–48 hours after injection and trough immediately before the next; for enanthate/cypionate a suggested concentration is roughly 400–700 ng/dL, and for undecanoate the interval should lengthen if the trough is below 400 ng/dL.[5,6] Injectable oestradiol (valerate, cypionate) behaves the same way: levels rise and fall across the interval, and guideline-based dosing can overshoot into supraphysiologic peaks — so when the sample is drawn materially changes what “in range” means.[24] A level without a timing is hard to interpret.

How often. The guidelines openly disagree. The Endocrine Society recommends measuring testosterone and oestradiol every 3 months (and testosterone every 3 months in masculinising care until it sits within the adult male reference interval), with haematocrit at baseline, 3-monthly in the first year, then once or twice a year.[1,5] UCSF counters that 3-monthly hormone monitoring is not realistic and unlikely to add value once dosing is stable, with annual visits sufficient on a stable regimen.[2] We present that tension rather than resolve it silently: intensive early, pragmatic once stable, and driven by change (new symptoms, dose changes, intercurrent illness) rather than the calendar.

Confidence: 🟢 High (the pharmacokinetic timing facts) / 🟡 Moderate (the optimal interval).

4. The mechanistic core — SHBG and the free fraction

This is the hinge of the whole review: why a single total level is a weak instrument. Total hormone is the sum of free, albumin-bound and SHBG-bound fractions, and SHBG is not a constant — it is moved by the very hormones being given.[9]

The direction flips between the two treatments, and that is the teaching point. Oestrogen — especially oral — raises SHBG, which binds more testosterone and lowers the bioavailable fraction, so in an oestrogen-treated trans woman with high SHBG a female-range total testosterone can overestimate bioavailable testosterone; UCSF’s own advice is to calculate bioavailable testosterone when the clinical androgen effect and the total disagree.[2,17] Tissue androgen effect also depends on receptor sensitivity, 5α-reduction and local metabolism, so the calculated fraction informs rather than settles the picture. Testosterone runs the other way — it tends to lower SHBG — so in a trans man the free fraction can be higher than an “in-range” total implies. Same lab value, opposite interpretation, depending on the direction of care.

The same binding biology is a further reason to treat total oestradiol as an imperfect exposure marker — a total can read “in range” while binding shifts what the bioactive fraction actually is — although, unlike testosterone, the clinical utility of a free-oestradiol measurement in gender-affirming care is not established, so this is a caveat on the total rather than a case for measuring free oestradiol.

And the fix is itself soft. The bioavailable or free figure that would resolve this is a calculation resting on assumptions (an equation, plus an SHBG assay that drifts), not a hard measurement. So the honest position is not “measure free instead of total” but “hold the total lightly, reach for calculated bioavailable when binding is likely shifted or the picture is discordant, and treat every number as an estimate.”[7,26]

Confidence: 🟢 High (the binding physiology) / 🟡 Moderate (how cleanly it translates to a management action).

Part B — Targets: the appraisal layer

5. Feminising targets, graded

The headline target — oestradiol 100–200 pg/mL with testosterone below 50 ng/dL — is one of the most confidently cited numbers in the field and one of the least outcome-validated.[1]

The specific 100–200 pg/mL range is not validated against outcomes. A 2025 systematic review of 49 studies set out to test whether the 100–200 pg/mL range delivers better feminising outcomes or fewer adverse events, and concluded that existing data do not support that range as a marker of adequacy, insufficiency, testosterone suppression, toxicity or adverse events — oestradiol can serve as one data point in an overall assessment, but not as a validated target.[12] A companion narrative review found no randomised trials of optimal concentrations, and found that higher oestradiol, given adequate testosterone suppression, did not enhance breast development or produce more feminine body composition.[11] UCSF is blunt that the original target derives from expert opinion, may be overly conservative, and that there is no evidence higher oestradiol adds feminisation.[2] The guidelines can’t even agree on the number: 100–200 pg/mL (Endocrine Society), above ~54 pg/mL for bone (Amsterdam), 68–163 pg/mL (Australia).[10]

Testosterone suppression is the better-supported lever. What the outcome data do support is that suppressing testosterone matters for feminisation — and that suppression is frequently not achieved: in one cohort 7.9% did not reach target, and even at a looser threshold of below 100 ng/dL, a third of another cohort fell short.[11,13] So the more defensible feminising posture is: prioritise adequate testosterone suppression, keep oestradiol in a physiological range mainly to limit risk rather than to chase feminisation, and avoid supraphysiologic peaks. The physiological “female range” itself is a cisgender reference interval adopted by consensus, not a trans-outcome-derived target.

Two guardrails on this critique. First, this is an appraisal of target validation, not an argument against monitoring: oestradiol and testosterone measurements remain clinically useful for adherence, absorption, pharmacokinetic timing, unexpectedly high or low exposure, safety context and a discordant clinical response. Second, testosterone suppression is not the same as full clinical feminisation — voice, facial and body hair, and skeletal structure frequently need non-hormonal interventions regardless of the numbers.

Confidence: 🔴 Low (the oestradiol target as an outcome-validated number) / 🟡 Moderate (that testosterone suppression aids feminisation).

6. Masculinising targets, graded

The masculinising target is more coherent but no better validated against long-term outcomes. The practical aim is a total testosterone in the adult male range — roughly 300–1000 ng/dL, with effective-dosing studies often using 320–1000 — reached and then maintained, with the injectable timing caveats of Section 3.[5,6] Bioavailable testosterone can be calculated in complex or discordant cases, and male-range reference intervals for related labs (such as alkaline phosphatase and creatinine) become appropriate over time.[3]

The honest grading: reaching the male range is a reasonable, internally coherent goal supported by pharmacology and cohort experience, but the specific bounds are cisgender male reference intervals, and there is no outcome study establishing that any particular point within that range optimises virilisation, mood, bone or cardiovascular safety. Above-range testosterone is not known to add virilisation and may add risk (see the Masculinising Hormone Therapy review), so the target is a range to sit within, not a number to exceed. Male-range targets are cisgender-derived and applied to trans men without trans-specific outcome validation.

Confidence: 🟡 Moderate (male-range as a reasonable goal) / 🔴 Low (any specific point as outcome-optimal).

7. What are the targets even optimising?

Step back and the appraisal resolves into one question the targets rarely state: optimising for what? Feminisation, bone, cardiovascular safety, mental health and breast development are different endpoints that may not share an optimum, and the number-to-outcome link is weak across all of them.

For feminisation, the better-supported lever is testosterone suppression — measured outcomes track it more consistently than oestradiol concentration.[11] For bone, the evidence points the other way — higher oestradiol is associated with higher bone density, which is why some centres target a floor for bone rather than a physiological ceiling for feminisation.[10,11] For cardiovascular and thrombotic safety, what matters is route and formulation far more than hitting a precise level (Section 8). For wellbeing, one cross-sectional study found higher oestradiol produced a statistically detectable but clinically negligible difference in distress and life satisfaction.[28] So a single target number is being asked to serve outcomes that pull in different directions — which is precisely why “your level is in range” can be true and still not mean what a person was told it means.

Confidence: 🔴 Low. The link from any specific target number to hard clinical outcomes is weak in both directions.

Part C — Safety surveillance: firmer ground

8. Safety monitoring and screening

Surveillance for harm rests on firmer evidence than target-chasing, because here the thing measured maps onto a real risk.

Erythrocytosis (masculinising). The clearest routinely monitorable hormone-linked safety signal in masculinising therapy. Testosterone raises haematocrit; check it at baseline, 3-monthly in the first year, then once or twice yearly, with a haematocrit above 0.50 a warning and above 0.54 the level at which intervention is generally indicated — though that 0.54 threshold is largely extrapolated from testosterone-treated cisgender men, and in trans men the exact cut-off and its management remain debated. The 0.54 threshold is extrapolated from cisgender men.[5,8,14] Management is dose or formulation adjustment and, occasionally, phlebotomy (see the Masculinising Hormone Therapy review).

Venous thromboembolism and route (feminising). This is where route beats level. The transdermal route appears to carry the lowest observed thrombotic risk among common oestrogen routes — the evidence is observational and confounded, so “lowest” is not “zero”; oral oestrogen carries a first-pass procoagulant shift, and ethinylestradiol carries the strongest thrombotic signal — one long-term cohort linked oral ethinylestradiol to a roughly threefold rise in cardiovascular death.[16,18,29] Current guidance therefore favours transdermal 17β-oestradiol, avoids ethinylestradiol and high-dose cyproterone, does not recommend routine thrombophilia screening before starting (clinical risk factors predict better), and does not use primary anticoagulant prophylaxis.[16,17,18] Route selection should also weigh personal risk factors — smoking, older age, obesity, thrombophilia and prior VTE all raise the stakes and push toward the transdermal route. Two honest caveats: the oral-versus-transdermal comparisons use non-equivalent doses, so some of the “route” effect may be dose; and the absolute risk remains low. Trans men on testosterone appear to carry little or no excess VTE risk on current data.[16]

Metabolic. Lipids at baseline and follow-up in both directions; the most consistent testosterone change is a fall in HDL; blood pressure and glucose as part of general cardiovascular care.[6]

Prolactin (feminising). Contested. Oestrogen stimulates lactotrophs and prolactinomas have been reported, but the early cases were largely on oestrogen formulations and cyproterone doses no longer recommended, and prolactinomas are rare — so routine prolactin monitoring is of doubtful value and risks unnecessary investigation. A defensible position is to check prolactin for symptoms (new persistent headache, visual disturbance, galactorrhoea) rather than reflexively, interpreted against the female reference interval.[19,27]

Potassium (feminising). If spironolactone is used, monitor potassium — with extra caution in renal impairment or alongside other potassium-raising drugs.[6]

Bone. DEXA at baseline for those at risk of osteoporosis, otherwise from around age 60 or earlier if sex-hormone levels run persistently low; a very low oestradiol on either regimen is a bone red flag, and suppressed LH/FSH can act as a surrogate that exposure is adequate.[2,6]

Screening while organs are in situ. Cancer-screening guidance for trans people is largely extrapolated from cisgender protocols on limited data, so recommendations are broad.[21,22] The workable rule is organ-based: people with a cervix and breast tissue should access cervical and breast screening per local schedules while those organs are present (with careful speculum technique given atrophy, and self-examination of residual chest tissue after masculinising chest surgery); trans women retain the prostate and should be screened by shared decision (palpable via the anterior neovaginal wall after vaginoplasty).[21,23] One drug-specific flag: long-term high-dose cyproterone carries a meningioma risk and warrants a switch or dose reduction.[21] All of this is extrapolated from cisgender screening evidence; trans-specific screening outcome data barely exist.

Context and scope. A few treatment-context points sit alongside the labs. Anti-androgen-specific monitoring — cyproterone, spironolactone, GnRH analogues and bicalutamide — belongs to the Anti-androgens Compared review; pregnancy potential, contraception and fertility preservation in testosterone users are covered in the Masculinising Hormone Therapy review. This review is adult-framed: adolescent monitoring differs, and low-dose or nonbinary goals may make physiological binary targets inappropriate.

Confidence: 🟢 High (erythrocytosis) / 🟡 Moderate (VTE-route, prolactin, bone) / 🔴 Low (screening thresholds, which are extrapolated).

9. Evidence quality

The review’s own grading, stated plainly. The measurement layer is firm: the analytical and pharmacokinetic facts (calculated-not-measured free testosterone, mass spectrometry over immunoassay, injectable timing, the SHBG mechanism) are well established. The target layer is weak: no randomised trials of optimal concentrations exist, a systematic review explicitly rejects the flagship oestradiol range, and guidelines pick different numbers — the targets are cisgender reference intervals adopted by consensus, not thresholds validated against trans outcomes.[10,11,12] The safety layer is intermediate: erythrocytosis monitoring is solid, route-based VTE risk is directionally clear but confounded by non-equivalent dosing, and screening is extrapolated. Ratings use the series scheme — 🟢 reasonably robust, 🟡 moderate, 🔴 limited — as editorial judgements, not formal GRADE. See Reading the Evidence.

Claim Confidence Why
Free testosterone must be calculated, not measured 🟢 Robust Endocrine Society: direct immunoassays inaccurate; calculated tracks dialysis.[7,8]
Mass spectrometry > immunoassay at low concentrations 🟢 Robust Method-comparison consensus.[5]
Injectable levels swing; timing changes the number 🟢 Robust Pharmacokinetics of esters/valerate.[5,24]
SHBG shifts (oestrogen up, testosterone down) alter bioavailable fraction 🟢 Robust Established endocrinology.[9,17]
Oestradiol 100–200 pg/mL as an outcome-validated target 🔴 Low Systematic review found no supporting data.[12]
Higher oestradiol adds feminisation 🔴 Low Not shown; testosterone suppression the lever.[11]
Testosterone suppression aids feminisation 🟡 Moderate Cohort-supported; often not achieved.[11,13]
Male-range testosterone as a specific optimal point 🔴 Low Cisgender reference interval; no outcome optimum.[5]
Haematocrit as a monitorable safety signal 🟢 Robust Quantified, thresholded.[8,14]
Transdermal oestradiol: lowest observed thrombotic risk; avoid ethinylestradiol 🟡 Moderate Directionally clear; dose-confounded.[16,18]
Routine prolactin monitoring is useful 🔴 Low Contested; old-regimen cases; rare.[19]
Trans-specific cancer-screening thresholds 🔴 Low Extrapolated from cisgender protocols.[21,22]

10. Frequently asked questions

My level is “in range” — does that mean my dose is right? Not necessarily. The ranges are consensus reference intervals, not validated targets, and a total level can miss the bioavailable picture. (Sections 4, 5, 7)

Should I push my oestradiol higher to feminise faster? No evidence supports it; measured feminising outcomes track testosterone suppression, and pushing oestradiol above physiological targets has not been shown to help and may add risk depending on route, formulation, dose and personal risk factors. (Sections 5, 7)

Is a “free testosterone” blood test better than total? Only if it’s calculated — direct free-testosterone immunoassays are inaccurate and shouldn’t be used. (Section 2)

Does the type of oestrogen matter more than the level? For safety, yes — route and formulation drive thrombotic risk more than hitting a precise number; avoid ethinylestradiol. (Section 8)

Why do two labs give me different “normal” ranges? Assays differ, SHBG assays drift, and reference intervals are lab-specific — always interpret against the reporting lab’s range. (Sections 2, 4)

11. Clinical bottom line

  1. Hold the monitoring/target distinction: how and when to measure is firm; what number to aim at is not.[10]
  2. Don’t treat the oestradiol range as validated — it isn’t; prioritise testosterone suppression for feminisation and physiological oestradiol mainly for safety.[11,12]
  3. If you need a free/bioavailable figure, calculate it; never use a direct free-testosterone immunoassay.[7,8]
  4. Interpret any total against SHBG and the direction of care; treat every number as an estimate tied to its assay and draw timing.[9,17]
  5. For safety, choose route over level: transdermal oestradiol, no ethinylestradiol; monitor haematocrit, and screen organs that are in situ.[8,16,18,21]

Sidebar — Targets at a glance, and what they’re worth

Target in use Source Outcome support
Oestradiol 100–200 pg/mL (feminising) Endocrine Society consensus 🔴 Systematic review found none
Testosterone <50 ng/dL (feminising) Cisgender female interval 🟡 Suppression aids feminisation; exact cut-off not validated
Oestradiol floor for bone (~>54 pg/mL) Amsterdam cohort 🟡 Higher E2 ↔ higher BMD (association)
Testosterone 300–1000 ng/dL (masculinising) Cisgender male interval 🟡 Reasonable goal; no optimal point shown
Haematocrit <0.54 (masculinising) Erythrocytosis threshold 🟢 Real, quantified risk signal

Sidebar — What the total number hides

A single total hormone level is a weak instrument for three reasons:

Sidebar — Common myths

Myth Evidence Strength
“A level in the target range means the dose is optimal.” Ranges are unvalidated consensus intervals; totals miss the free fraction. 🔴
“Higher oestradiol means faster feminisation.” Not shown; testosterone suppression is the lever.[11] 🔴
“A direct free-testosterone test is more accurate than total.” Direct free-T immunoassays are inaccurate; calculate instead.[8] 🟢
“The exact oestradiol level drives clot risk.” Route/formulation drive thrombotic risk more than level.[16,18] 🟡

Evidence appendix — the key studies at a glance

Study Design What it establishes Key limitation
Winston-McPherson 2025[12] Systematic review (49 studies) Data do not support the 100–200 pg/mL oestradiol target Observational inputs; heterogeneity
Nolan narrative review[11] Narrative review No RCTs; higher E2 doesn’t add feminisation; T suppression does Non-systematic; cohort data
Vermeulen 1999[7] Method validation Calculated free T tracks equilibrium dialysis Assumption-based equation
Endocrine Society (Bhasin) 2018[8] Guideline Direct free-T immunoassays inaccurate; Hct>54% erythrocytosis Largely cisgender male data
ESTHER / route-VTE cohorts[16,18] Cohort/case-control Oral (esp. ethinylestradiol) more thrombogenic than transdermal Non-equivalent doses; low absolute risk
Clinical Chemistry lab-monitoring review 2025[10] Narrative review Guidelines pick different targets; utility of level-monitoring unclear No optimal-concentration studies exist

What we know / What we don’t know

What we know

What we don’t know

Related Eden Openly reviews

A note on support

This article is health information, not personal medical advice, and it is not a substitute for assessment by a qualified clinician. It should not be used to start, stop, dose-adjust, change the route of, or change the monitoring of any medication without clinical supervision, and monitoring should follow local standards. Laboratory reference ranges are assay- and lab-specific; always interpret results against the reporting laboratory’s ranges. If a result or symptom concerns you, discuss it with your own clinician.

Revision history

Version 1.0 — July 2026. First published.

References

Population flags mark sources whose primary evidence is from cisgender populations and is extrapolated to trans people.

  1. Hembree WC, Cohen-Kettenis PT, Gooren L, et al. Endocrine Treatment of Gender-Dysphoric/Gender-Incongruent Persons: An Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2017;102(11):3869–3903. doi:10.1210/jc.2017-01658. (Targets: oestradiol 100–200 pg/mL, testosterone <50 ng/dL; monitoring intervals.)
  2. Deutsch MB, ed. Guidelines for the Primary and Gender-Affirming Care of Transgender and Gender Nonbinary People (UCSF Gender Affirming Health Program) — Overview of feminizing hormone therapy. (Target derives from expert opinion, may be overly conservative; no evidence higher oestradiol adds feminisation; LH/FSH suppression as bone surrogate; prolactin only if symptomatic; annual monitoring once stable.)
  3. UCSF Gender Affirming Health Program — Overview of masculinizing hormone therapy. (Total testosterone monitoring; bioavailable testosterone calculation; male-range reference intervals.)
  4. Coleman E, et al. Standards of Care for the Health of Transgender and Gender Diverse People, Version 8 (WPATH). Int J Transgend Health. 2022. (Transdermal preferred; avoid ethinylestradiol.)
  5. ARUP Consult — Endocrine Testing in Transgender Adults. (Monitoring intervals; enanthate/cypionate 400–700 ng/dL; undecanoate lengthen interval if trough <400; tandem mass spectrometry vs immunoassay; SHBG in oestrogen users; free/bioavailable calculated.)
  6. Gardner IH, Safer JD. Progress on the road to better medical care for transgender patients. Curr Opin Endocrinol Diabetes Obes. 2013;20(6):553–558. (Practical targets: male 300–1000 ng/dL; female testosterone 30–100, oestradiol <200; prolactin/triglycerides; potassium on spironolactone; BMD screening.)
  7. Vermeulen A, Verdonck L, Kaufman JM. A Critical Evaluation of Simple Methods for the Estimation of Free Testosterone in Serum. J Clin Endocrinol Metab. 1999;84(10):3666–3672. (Calculated free testosterone; correlates with equilibrium dialysis.)
  8. Bhasin S, Brito JP, Cunningham GR, et al. Testosterone Therapy in Men With Hypogonadism: An Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2018;103(5):1715–1744. [Population flag: cisgender men.] (Direct free-testosterone immunoassays inaccurate — use algorithm or equilibrium dialysis; erythrocytosis at haematocrit >54%.)
  9. Laboratory Monitoring in Transgender People. ADLM (formerly AACC) Clinical Laboratory News. 2020. (Free testosterone ~2–3% of total; total testosterone = free + albumin-bound + SHBG-bound; oestradiol cannot monitor synthetic or conjugated estrogens.)
  10. Nolan BJ, Cheung AS. Laboratory Monitoring in Transgender and Gender-Diverse Individuals. Clin Chem. 2025;71(3):358–377. doi:10.1093/clinchem/hvaf001. (Endocrine Society, Amsterdam and Australian guidelines set different oestradiol and testosterone targets; no study has evaluated optimal concentrations in feminising therapy.)
  11. Nolan BJ, Cheung AS, et al. Relationship Between Serum Estradiol Concentrations and Clinical Outcomes in Transgender Individuals Undergoing Feminizing Hormone Therapy: A Narrative Review. Transgend Health. 2021. (No RCTs; higher oestradiol does not enhance breast development or body composition; testosterone suppression the key lever; higher oestradiol ↔ higher BMD.)
  12. Winston-McPherson GN, Thomas TA, Krasowski MD, et al. Estradiol Concentrations for Adequate Gender-Affirming Feminizing Therapy: A Systematic Review. LGBT Health. 2025. doi:10.1089/lgbt.2024.0407. (49 studies; existing data do not support the 100–200 pg/mL oestradiol range for feminising outcomes or reduced adverse events.)
  13. Slack DJ, Krishnamurthy N, Garrity M, Chen D, Kyweluk M, Trakhtenberg E, Kirkley J, Safer JD. Choice of Hormone Assay to Monitor Feminizing Gender-Affirming Hormone Therapy. Endocr Pract. 2025;31(12):1521–1529. PMID: 40945660. (Cross-sectional analysis of 9,916 transfeminine individuals; favour target testosterone over oestradiol in patients with testes, and consider LH in those without testes not on GnRH agonists; testosterone suppression frequently not achieved in cohorts — ENIGI 7.9%; 33% of 136 above 100 ng/dL.)
  14. Madsen MC, van Dijk D, Wiepjes CM, et al. Erythrocytosis in a Large Cohort of Trans Men Using Testosterone. J Clin Endocrinol Metab. 2021;106(6):1710–1717. doi:10.1210/clinem/dgab089. (Haematocrit thresholds and determinants.)
  15. Okano SHP, Braga GC, Cantelli DAL, et al. Effect of testosterone formulations on haematocrit in transgender individuals: a systematic review. Andrology. 2025;13(3):422–430. doi:10.1111/andr.13695.
  16. Goldstein Z, Khan M, Reisman T, Safer JD. Managing the risk of venous thromboembolism in transgender adults undergoing hormone therapy. J Blood Med. 2019;10:209–216. doi:10.2147/JBM.S166780. (Review: oral oestrogen’s first-pass procoagulant shift; avoid ethinylestradiol; progestins add risk; transdermal the lowest observed risk; absolute risk low. Route-VTE mechanism derives from ESTHER — ref 29.)
  17. Assessing and Addressing the Risk of Venous Thromboembolism Across the Spectrum of Gender-Affirming Care: A Review. PMC10081942. 2023. (Transdermal lowest observed thrombotic risk; non-equivalent dose comparisons; VTE higher than cisgender controls; ~half within first 12 months; peri-surgical considerations.)
  18. VTE risk in transgender women on estrogen therapy: formulation-specific risks and management. Blood (ASH). 2025;146(Suppl 1):1344. [Conference abstract — lower certainty.] (Transdermal 17β-oestradiol preferred; avoid ethinylestradiol and cyproterone; routine thrombophilia screening not recommended; no primary anticoagulant prophylaxis.)
  19. Nolan BJ, Accatino MI, Angus LM, Cheung AS. Approach to prolactin monitoring and hyperprolactinaemia in transgender and gender-diverse individuals undergoing gender affirming hormone therapy. Front Endocrinol (Lausanne). 2025. doi:10.3389/fendo.2025.1608108. PMCID: PMC12148852. (Oestrogen stimulates lactotrophs; the early prolactinoma cases were on oestrogen and cyproterone regimens no longer recommended; routine prolactin monitoring is of doubtful value; use the female reference interval.)
  20. Wiepjes CM, et al. Bone Mineral Density Increases After 1 Year of Hormonal Treatment; Bone safety during the first ten years of gender-affirming hormonal treatment. J Bone Miner Res. 2017/2019. (Higher oestradiol ↔ higher BMD; Amsterdam bone-floor target.)
  21. The implications of hormone treatment for cancer risk, screening and treatment in transgender individuals. Best Pract Res Clin Obstet Gynaecol. 2024. (Screening extrapolated from cisgender protocols; breast and cervical screening while organs are in situ; long-term high-dose cyproterone → meningioma risk; no proven endometrial-cancer increase on testosterone.)
  22. Sterling J, Garcia MM. Cancer screening in the transgender population: a review of current guidelines, best practices, and a proposed care model. Transl Androl Urol. 2020;9(6):2771–2785. doi:10.21037/tau-20-954. (Very few trans-specific screening recommendations; guidance defaults to cisgender protocols applied via small case series.)
  23. De Haan G. Adapting cancer screening after gender-affirming transition. Presented at the FIGO World Congress of Gynecology and Obstetrics, 2023. [Conference presentation — lower certainty.] (Careful speculum technique or sedation for cervical screening; prostate palpable via the anterior neovaginal wall after vaginoplasty.)
  24. Kanin M, Slack M, Patel R, Chen KT, Jackson N, Williams KC, Grock S. Injectable Estradiol Dosing Regimens in Transgender and Nonbinary Adults Listed as Male at Birth. J Endocr Soc. 2025;9(5):bvaf004. doi:10.1210/jendso/bvaf004. PMID: 40170698; PMCID: PMC11957913. (Guideline-based injectable oestradiol dosing can produce supratherapeutic levels; EV/EC dose ranges.)
  25. Nolan BJ, Brownhill A, Bretherton I, et al. Relationships between body mass index, oral oestradiol dose and serum oestradiol in transgender adults undergoing feminising hormone therapy. Ther Adv Endocrinol Metab. 2020;11:2042018820924543. (Oral oestradiol raises oestrone through first-pass metabolism; route affects the oestrone/oestradiol ratio.)
  26. Clinical Pathology Laboratories — SHBG reagent reformulation notice, 2019. (Reformulation shifted calculated free/bioavailable testosterone by 12–16% without reference-range adjustment.)
  27. Autonomous prolactin secretion in two male-to-female transgender patients using conventional oestrogen dosages. BMJ Case Rep. PMCID: PMC3029513. (Prolactinomas rare; long-latency low-dose cases.)
  28. Cheung AS, et al. Estradiol Concentrations and Wellbeing in Trans People Using Estradiol Hormone Therapy. Cross-sectional study, 2020–2021. (Higher oestradiol associated with a statistically detectable but clinically negligible difference in distress and life satisfaction; no significant association with thrombosis, malignancy, stroke or myocardial infarction — underpowered.)
  29. Canonico M, Oger E, Plu-Bureau G, et al. (ESTHER Study Group). Hormone therapy and venous thromboembolism among postmenopausal women: impact of the route of estrogen administration and progestogens. Circulation. 2007;115(7):840–845. [Population flag: postmenopausal cisgender women.] (Oral, but not transdermal, oestrogen raises VTE risk — the basis for the route effect extrapolated to feminising care.)