Gender-Affirming Hormone Therapy in Adults: What the Research Shows
In October, a new federal rule takes effect that restricts Medicaid and CHIP funding for gender-affirming care for people under 19. Most of my patients aren't affected. The rule covers pediatric and adolescent care, and at Aurum Haven Medical I see adults.
But it seemed like the right moment to write about what some of the current research says for adults, because that is a separate body of evidence. As an FYI, I'm also going to include some parts where the evidence is weak, or where it hasn't been held to the same methodological standard as other areas of medicine.
What the rule does, and what it doesn't
The Centers for Medicare and Medicaid Services issued the final rule on August 11, 2026.(1) It bars federal matching funds for puberty suppression, cross-sex hormone therapy, and related surgery for Medicaid beneficiaries under 18 and CHIP enrollees under 19. It takes effect October 13, 2026.
A few details worth stating precisely, because they get flattened in coverage:
Patients already receiving cross-sex hormone therapy as of the effective date can continue under federal funding for a tapering period of up to six months. This doesn't apply to new starts, and it doesn't apply to puberty blockers or surgery at all.
Federal Medicaid and CHIP funding remains available for psychotherapy and mental health care for gender dysphoria.
States can still cover these services with state-only funds, or patients can use private insurance.
The rule itself states that state Medicaid agencies are not prohibited under federal law from covering these services for beneficiaries 18 and older.
That last point is the one I'd underline for anyone reading this as an adult patient. Adult care sits outside the scope of this rule, and the rule says so.
In California, the Department of Health Care Services issued a provider notice in May 2025 reaffirming that all medically necessary gender-affirming care remains covered under Medi-Cal and that Medi-Cal policy had not changed.(2) DHCS and the Department of Managed Health Care have said they are working with managed care plans to monitor compliance with California's protections and to help members who hit barriers.(3)
What I tell patients is that this is a frustrating political attempt at medical regulation. I disagree with it, and California has moved the other way — codifying or making laws that protect patients and that line up with current medical practice.
"Low certainty" doesn't mean what it sounds like
The federal reasoning leans on an HHS umbrella review of the pediatric literature, the United Kingdom's Cass Review, and systematic reviews from Sweden and Finland, all of which concluded that certainty of benefit was very low.(1)
Start with the scope problem. The Cass Review examined NHS gender identity services for children and young people.(4) Its evidence base was a set of commissioned systematic reviews of the pediatric and adolescent literature. And the review's own summary makes the distinction I'd make: it says that masculinizing and feminizing hormones in those under 18 present many unknowns despite their longstanding use in the adult transgender population, and attributes the gap to a lack of long-term follow-up in people who start treatment young.
That's the review, in its own words, separating pediatric uncertainty from adult practice.
I'll note the counterargument before someone else raises it. Cass used "children and young people" to mean under 25 in places, and one recommendation asked NHS England to build follow-through services for 17-to-25-year-olds. That recommendation was about continuity of care across the boundary between children's and adult services, not about the hormone evidence in adults. But it's there, and worth acknowledging.
The questions driving the pediatric uncertainty also don't transfer. Puberty suppression is an intervention adults don't usually receive. Bone mineral accrual during the pubertal window, neurocognitive development in adolescence, diagnostic trajectory in prepubertal children, capacity to consent as a developing minor — none of these are live questions for a 34-year-old starting estradiol.
Now the term itself. Certainty ratings under GRADE describe how confident we are in the size of an effect. A low rating means further research would probably change the estimate. It does not mean the effect isn't there, and it does not mean the studies were done badly.
In this field, the ceiling on certainty is structural. Cooney and colleagues say this directly in their 2025 systematic review: the risk-of-bias ratings across this literature are attributable to inherent limitations of studying this particular intervention.(5) You cannot blind anyone to whether they're receiving hormones, because the intervention produces visible physical change. And for outcomes like dysphoria, mood, and quality of life, there is no serum marker — self-report is the only instrument available.
The same review addresses the obvious follow-up question. Equipoise doesn't permit placebo-controlled trials where hormones are already available as standard of care. What can be done ethically is quasi-experimental and waitlist-controlled design. Which is exactly what produced the best trial we have.
One more thing: the federal rule also argues that the WPATH and Endocrine Society guidelines are untrustworthy. I cite both below, and I'd rather name that than have you find it later. I'm citing them for their descriptive tables on the expected time course of physical changes, not as authority for whether hormones work. The case for that relies on the primary studies.
What the adult research found
The adult evidence base
What the studies looked at
Five sources carry most of the weight in the adult literature. They differ in design and size, and they point the same direction.
Randomized trial
Nolan et al., 2023–24
Immediate testosterone against the standard three-month wait. Dysphoria fell 7.2 points, depression 5.6, suicidality 6.5. Depression improved past the clinical threshold in 61% on testosterone versus 13% on standard care. Suicidal thoughts resolved in 52% versus 5%.
The 7.2-point dysphoria change sits below the instrument's 11-point clinical threshold. The authors attribute this to three months being too short.
Controlled cohort
Foster Skewis et al., 2021
The only prospective study to separate masculinizing from feminizing effects. Dysphoria fell in both groups — 6.80 points with testosterone, 4.22 with estrogen. Emotional well-being and social functioning improved with testosterone; no quality-of-life domain moved with estrogen.
Both dysphoria changes also fall below the 11-point threshold, with the same explanation: six months isn't long enough.
Real-world cohort
Corman et al., 2025
A high-symptom population in ordinary telehealth care rather than a trial. Among those starting with elevated depression, 42% saw their score halve and 27% reached remission. Suicidal thoughts resolved in 60% of those who reported them.
Just over half the original cohort completed follow-up, and the sample skews young, white, and uninsured.
Systematic review
Doyle et al., 2023
Consistent reductions in depressive symptoms and psychological distress across the literature. Quality-of-life findings trended positive without being consistent.
Risk of bias varied widely, with small samples and unadjusted confounders throughout.
Systematic review
Cooney et al., 2025
Graded every outcome formally. Rated the trial evidence low-certainty and the uncontrolled studies at serious to critical risk of bias — and found no substantive harms in any domain where data existed.
Eight prespecified outcomes had no comparative data at all, including physical health, mortality, and discrimination.
Read together, these studies agree on direction and disagree on magnitude. That distinction is the whole argument: we know hormone therapy helps, and we are still working out how much.
Full citations appear at the end of this article. Figures come from the published reports and reflect the populations studied, which skew younger and are concentrated in Europe, Australia, and the United States.
The randomized trial
Nolan and colleagues ran a three-month open-label randomized trial in Melbourne.(6) The design solves the ethical problem head-on: because standard care there already meant roughly a three-month wait, participants were randomized either to start testosterone immediately or to that existing wait. Nobody waited longer than they otherwise would have.
Sixty-four transgender and gender-diverse adults, ages 18 to 70. Everyone met the threshold for clinical gender dysphoria at baseline, and the depression burden was heavy — 34% scored moderately severe on the PHQ-9 and 23% scored severe.
Compared with the standard-care group, immediate testosterone produced:
A reduction in gender dysphoria of 7.2 points on the GPSQ (95% CI, −8.3 to −6.1; P<.001)
A reduction in depression of 5.6 points on the PHQ-9 (95% CI, −6.8 to −4.4; P<.001)
A reduction in suicidality of 6.5 points on the SIDAS (95% CI, −8.2 to −4.8; P<.001)
The responder numbers say more than the averages do. An improvement of at least 5 points on the PHQ-9 — the threshold considered clinically significant — occurred in 19 people (61%) on testosterone and 4 (13%) on standard care. Among those reporting suicidal thoughts at baseline, those thoughts resolved in 11 people (52%) on testosterone and 1 person (5%) on standard care.
Adverse events at three months were minor: seven people had injection-site pain or discomfort, one had a transient headache. Nobody developed polycythemia.
A prespecified secondary analysis looked at quality of life.(7) Self-rated health improved by 11.6 points on the EQ-VAS (95% CI, 4.9 to 18.3; P=.02), which clears the 7-to-10-point range previously treated as clinically meaningful. The utility index moved 0.07 points but didn't reach statistical significance (95% CI, −0.07 to 0.21; P=.06) — though 0.07 sits at the top of the range previously defined as clinically meaningful, which tells you something about how much weight these thresholds can bear. Cooney's summary of this is the fair one: hormones may improve quality of life or have no effect on it, depending on which instrument you use.(5)
The controlled cohort
Foster Skewis and colleagues followed 77 adults starting hormones against 103 cisgender comparators matched on presumed sex at birth, over six months.(8) This is the only prospective controlled study to separate masculinizing from feminizing effects.
Dysphoria fell in both groups. In the masculinizing group (n=42), it fell 6.80 points (95% CI, −8.68 to −4.91; P<.001), and two quality-of-life domains improved past their clinical threshold: emotional well-being by 7.48 points and social functioning by 12.50. In the feminizing group (n=35), dysphoria fell 4.22 points (95% CI, −6.21 to −2.24; P<.001), and no quality-of-life domain moved significantly.
Real-world data
Corman and colleagues followed 342 adults starting hormones through a telehealth clinic across 43 states — 192 on estrogen, 150 on testosterone, mean age 26.5.(9) This is not a trial population. A quarter reported suicidal thoughts at baseline, 40% had elevated depression scores, and 36% had elevated anxiety scores.
At three months, among those who started with elevated depression, 42% saw their score cut in half and 27% reached remission. Among those with elevated anxiety, 40% responded and 28% reached remission. Of the 83 people reporting suicidal thoughts at baseline, 50 no longer reported them.
The caveat here is attrition. Of 669 adults who started hormones during the study window, 342 completed follow-up — 52%. The authors report no baseline mental health differences between those who finished and those who didn't, which helps, but half the group is unaccounted for. The sample also skews young, white, and uninsured, and comes from a single telehealth clinic. There was no control group, which the authors describe as both infeasible and unethical.
The systematic reviews
Doyle and colleagues synthesized 46 studies and found consistent reductions in depressive symptoms and psychological distress, with quality-of-life findings that trended positive without being consistent.(10) Risk of bias varied widely.
Cooney and colleagues graded each outcome formally across 13 studies.(5) The trial evidence came out low-certainty, downgraded once for the blinding problem and once for imprecision given the small sample. The uncontrolled studies rated serious to critical risk of bias. And across all seven outcome domains where data existed, no study identified substantive harms. A companion review covering gender-affirming care more broadly reached the same conclusion across 28 studies and four randomized trials.(11)
Why the numbers look modest
Let’s look a little closer.
The dysphoria reduction in the Nolan trial — that 7.2 points — falls below the GPSQ's threshold for clinically significant change, which is 11 points. The authors say so themselves, and they attribute it to the three-month follow-up being too short for masculinizing physical changes to develop.(6) Twelve participants (39%) did clear the threshold individually. No one in the control group did.
The same thing happens in Foster Skewis. Dysphoria fell 4.2 to 6.8 points depending on the group, and the authors point out that both figures sit below the same 11-point threshold. Their explanation is the same: six months isn't long enough.(8)
Two independent studies, two different designs, the same finding, the same explanation. However, the effect is real and it is measured before it has finished happening.
This is what I tell patients, and it's the least surprising thing in the world once you've watched it: for better or worse, these changes don't happen overnight. They're on a hormonal timeline.
Some things move fast. On testosterone, people notice increased hair growth, shifts in mood, changes in libido. On estrogen, many of my patients notice the emotional effects first, along with some softening of the skin and the beginning of breast bud development.
But none of that immediately translates into physical congruence. And congruence is what the dysphoria instruments are ultimately measuring.
Expected time course
Hormone therapy runs on a hormonal timeline
Each bar begins when an effect typically starts and extends to when it typically reaches its maximum. Faded ends mean the published range is unknown or variable. Individual response varies considerably.
Testosterone
Most effects begin within the first six months. Hair and muscle take the longest to finish developing.
Estrogen
The earliest changes are the ones other people can't see. Breast growth and fat redistribution take years to mature.
Estrogen's earliest effects are largely internal — mood, libido, skin. The changes other people notice arrive later and mature over years. That gap is why voice work, hair removal, and surgical planning belong in the conversation from the start rather than afterward.
Fewer than one in five patients reach Tanner breast stage 4 or 5 after two years of estrogen. Knowing that early prevents a lot of unnecessary dose changes.
Onset and maximum ranges adapted from WPATH Standards of Care, Version 8, Appendix C, itself adapted from Hembree et al., 2017. Breast development figure from de Blok et al., 2021, as reported in SOC-8. These are clinical observations rather than guarantees, and vary with genetics, age at initiation, and overall health.
The published time courses back this up.(12,13) Facial and body hair on testosterone starts somewhere between six and twelve months and keeps developing past five years. Breast growth on estrogen starts at three to six months and doesn't reach its maximum for two to five years. Fewer than one in five patients reach Tanner stage 4 or 5 breast development after two years of estrogen.(12) I'd rather someone hear that number from me at the first visit than discover it themselves at month eighteen.
This also explains the split in the Foster Skewis data. Testosterone produces visible, externally legible change relatively quickly, and the quality-of-life measures follow. Estrogen produces its earliest effects internally, and the visible changes arrive later and mature slower. So at six months, dysphoria has fallen in both groups, but only the testosterone group shows movement on emotional well-being and social functioning.
Looking again at what the authors said: people seeking feminization sometimes need multidisciplinary support — speech pathology, surgical input — for those quality-of-life gains to show up.(8) That matches what I see. Voice work, hair removal, and surgical planning aren't always part of the process only to be considered later. For some patients they're the difference between hormones working chemically and hormones working in someone's actual life.
Where the evidence is thin, or missing
Quality of life is the least consistent endpoint in this literature. It trends positive and it doesn't reliably reach significance, and it's weakest for feminizing therapy specifically. Part of that is the timeline; part of it is that these instruments weren't built for this population. The SF-36 asks about mobility and bodily pain. The EQ-5D-5L has never been validated in trans and gender-diverse people, and its own developers note that minimal important differences haven't been established here at all. When a tool designed to detect functional disability fails to register that someone feels more like themselves, that's a fact about the tool.
Long-term mortality data exists and it is uncomfortable. The Amsterdam cohort followed people on hormone treatment from 1972 to 2018 — 2,927 transgender women and 1,641 transgender men.(14) Among transgender women there were 317 deaths, a standardized mortality ratio of 1.8 (95% CI, 1.6–2.0) against general-population men and 2.8 (2.5–3.1) against general-population women. Elevated causes included cardiovascular disease, lung cancer, HIV-related disease, and suicide. Among transgender men there were 44 deaths, with an SMR of 1.8 (1.3–2.4) against general-population women but 1.2 (0.9–1.6) against general-population men — a confidence interval that crosses 1. The risk didn't decline across five decades.
The authors' own interpretation is the one to hold onto: the pattern of causes gives no indication of a specific effect of hormone treatment, and points instead to the importance of monitoring and treating medical comorbidities and lifestyle factors in trans health care. Smoking, HIV, and minority stress are likely confounding in those numbers, and this study design cannot separate them from hormone exposure. Read correctly, this is an argument for closer primary care, not against hormones.
And some outcomes have no comparative data at all. Across the 13 studies in Cooney's hormone-specific review, not one measured access to health services, physical health, stigma, discrimination, utilization of health services, mortality, satisfaction with care, or violence.(5) That's not weak evidence, that's an absence of evidence.
Hormones are not a cure-all
Separate but intertwined
Hormone therapy is not a cure-all
Many patients live with gender dysphoria, depression, and anxiety at the same time. Treating the dysphoria often eases the other two. It rarely eliminates them.
Gender-affirming hormone therapy acts on dysphoria. Because these conditions overlap, relief there often lifts mood and eases anxiety as well, and the research shows this consistently.
But depression and anxiety are their own conditions and need their own treatment. They don't disappear because the dysphoria improved.
Figures from Corman et al., 2025: of 83 adults reporting suicidal ideation at baseline, 50 no longer reported it at three months. Areas of overlap are illustrative and not drawn to scale.
Many of my patients have depression and anxiety alongside their gender dysphoria. If you picture a Venn diagram, treating the dysphoria often improves the depression, or improves the anxiety. It doesn't eliminate them. These are separate conditions that happen to be intertwined, and they have to be treated that way.
The research says the same thing. In the Corman cohort, suicidal ideation resolved in 60% of the people who reported it at baseline.(9) Which means 40% still had it at three months, while on hormone therapy.
What this means in practice
In practice, this means an individualized approach — the same as any other part of primary care. I don't guarantee that hormone therapy will improve someone's depression or mood.
That said, the large majority of my patients who've gone through hormone therapy as part of their transition have told me their emotional health changed significantly. I've had very few report negative effects on their emotional health, particularly with adjunct therapy like progesterone, and many patients find progesterone beneficial.
It's also worth naming that this is a pubertal and transitional process. Along with it comes the stress of growth and new beginnings.
None of the uncertainty in this literature is an argument for withholding treatment. It's an argument for paying attention — for regular labs, for follow-up that's long enough to see changes when they happen, and for appointments with enough room to talk about the parts that don't show up on a questionnaire. That's the practice I’ve been building at AHM.
If you're considering starting hormone therapy, or you're already on it and something hasn't felt right, you can read about how care works here.
If you're struggling right now, Trans Lifeline is available at 877-565-8860. The Trevor Project offers 24/7 crisis support for LGBTQ+ young people. The 988 Suicide and Crisis Lifeline can be reached by call or text.
References
Centers for Medicare & Medicaid Services. Medicaid Program; Prohibition on Federal Medicaid and Children's Health Insurance Program Funding for Sex-Rejecting Procedures Furnished to Children. Final rule. Fed Regist. August 11, 2026.
California Department of Health Care Services. Provider notice: gender-affirming care coverage. May 2025. https://mcweb.apps.prd.cammis.medi-cal.ca.gov/news/33457
California Health & Human Services Agency. Joint statement from California Health & Human Services leaders on the federal government's attack on health care for transgender Americans. December 18, 2025.
Cass H. Independent Review of Gender Identity Services for Children and Young People: Final Report. NHS England; April 2024.
Cooney EE, Yeh PT, Kennedy KS, Kaptchuk RP, Wong B, Kennedy CE. Provision of gender-affirming hormones for trans and gender-diverse adults: a systematic review of health and quality of life outcomes, values and preferences, and costs. eClinicalMedicine. 2025;88:103460. doi:10.1016/j.eclinm.2025.103460
Nolan BJ, Zwickl S, Locke P, Zajac JD, Cheung AS. Early access to testosterone therapy in transgender and gender-diverse adults seeking masculinization: a randomized clinical trial. JAMA Netw Open. 2023;6(9):e2331919. doi:10.1001/jamanetworkopen.2023.31919
Nolan BJ, Zwickl S, Locke P, Zajac JD, Cheung AS. Testosterone and quality of life in transgender and gender-diverse adults seeking masculinization: a secondary analysis of a randomized clinical trial. JAMA Netw Open. 2024;7(10):e2443466. doi:10.1001/jamanetworkopen.2024.43466
Foster Skewis L, Bretherton I, Leemaqz SY, Zajac JD, Cheung AS. Short-term effects of gender-affirming hormone therapy on dysphoria and quality of life in transgender individuals: a prospective controlled study. Front Endocrinol (Lausanne). 2021;12:717766. doi:10.3389/fendo.2021.717766
Corman J, et al. J Med Internet Res. 2025;27:e64017. doi:10.2196/64017
Doyle DM, Lewis TOG, Barreto M. A systematic review of psychosocial functioning changes after gender-affirming hormone therapy among transgender people. Nat Hum Behav. 2023;7(8):1320-1331. doi:10.1038/s41562-023-01605-w
Cooney EE, Muschialli L, Yeh PT, et al. Provision of gender-affirming care for trans and gender-diverse adults: a systematic review of health and quality of life outcomes, values and preferences, and costs. eClinicalMedicine. 2025;88:103458. doi:10.1016/j.eclinm.2025.103458
Coleman E, Radix AE, Bouman WP, et al. Standards of Care for the Health of Transgender and Gender Diverse People, Version 8. Int J Transgend Health. 2022;23(suppl 1):S1-S259. doi:10.1080/26895269.2022.2100644
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
de Blok CJM, Wiepjes CM, van Velzen DM, et al. Mortality trends over five decades in adult transgender people receiving hormone treatment: a report from the Amsterdam cohort of gender dysphoria. Lancet Diabetes Endocrinol. 2021;9(10):663-670. doi:10.1016/S2213-8587(21)00185-6
