Testosterone GAHT and Blood Clots: What the Data Show

For trans men and other AFAB people using testosterone for gender affirmation, the best available studies have not found the procoagulant blood-test pattern seen with estrogen-based therapy. Fibrinogen, activated partial thromboplastin time (aPTT), and plasminogen activator inhibitor-1 (PAI-1) did not change in a 2024 meta-analysis, and no venous thromboembolic events were reported. The catch is size: the pooled analysis included only 312 people. That is enough to describe common laboratory changes, but not enough to rule out a rare blood clot outcome.

This evidence applies to AFAB people using exogenous testosterone as gender-affirming hormone therapy (GAHT). It does not establish the risk of testosterone treatment in cisgender men or women, whose ages, indications, baseline hormone levels, and cardiovascular risks may differ.

No measured procoagulant shift, but more red blood cells

A 2024 systematic review and meta-analysis by Tienforti et al. pooled seven studies—six prospective and one retrospective—of 312 AFAB people taking testosterone GAHT. Participants had mean ages between 23 and 30 years. The analysis found no significant change in fibrinogen, aPTT, or PAI-1, and it reported zero thromboembolic events.

Those are useful findings, especially because estrogen-based GAHT has a well-described procoagulant profile. But they are biomarker and event-observation findings, not proof that testosterone prevents clots or carries zero clot risk. A normal fibrinogen result does not erase every pathway that can contribute to thrombosis, and zero events in a small, young cohort can occur even when a rare risk exists.

The most consistent testosterone-related blood change was erythrocytosis: hemoglobin rose by 1.48 g/dL (95% CI, 1.17 to 1.78) and hematocrit by 4.39 percentage points (95% CI, 3.52 to 5.26). Heterogeneity was low for hemoglobin (I²=9%) and modest for hematocrit (I²=23%). Higher hematocrit can increase blood viscosity through a physical, rheological mechanism. That is distinct from activating clotting factors, but it is still clinically relevant.

Longer-term clinical literature has likewise identified erythrocytosis as the main hematologic issue during testosterone treatment. A 2021 review by Schönauer et al. describes the need for ongoing monitoring and occasional dose adjustment during long-term therapy. An American Society of Hematology education review also treats hematocrit monitoring as a practical part of testosterone GAHT care.

The tiny INR increase probably reflects the test tube

Tienforti and colleagues did find a statistically significant rise in international normalized ratio (INR): +0.02 (95% CI, 0.01 to 0.03; p=0.0001). Statistical significance answers whether an observed difference is unlikely under a null model; it does not tell us whether the difference changes a patient's health. An INR change of 0.02 is analytically interesting and clinically tiny.

The authors proposed a plausible explanation: rising hematocrit reduces the plasma fraction in a fixed-volume citrate collection tube. If the tube contains too much citrate relative to plasma, clotting-time measurements can lengthen even when clotting-factor activity has not changed. The unchanged aPTT and fibrinogen results fit that explanation better than a broad impairment of coagulation does.

Still, the explanation remains a hypothesis. None of the included studies directly measured vitamin K-dependent clotting factors II, VII, IX, and X. The result should be read as a small INR shift with a credible laboratory-artifact explanation—not as direct evidence that testosterone changes, or does not change, every component of coagulation.

A useful comparison with estrogen, with limits

The clearest head-to-head evidence comes from a 12-month prospective cohort study by Scheres et al. The researchers followed 100 trans men and 98 trans women in parallel. Testosterone use in the trans men cohort produced no apparent procoagulant changes, while estrogen-based therapy in trans women produced procoagulant shifts that could be related to thrombotic risk.

Because the study measured both groups within one design and at the same 12-month endpoint, it reduces some of the problems that arise when comparing separate studies. It does not convert the comparison into a long-term VTE trial. The study followed biomarkers over one year, not enough people or time to precisely estimate a rare clinical event.

The distinction between hormone regimens nevertheless matters. A 2022 meta-analysis by Totaro et al. found elevated VTE prevalence among transgender women receiving feminizing hormone therapy. That does not prove that every hormone regimen has the same risk, and it should not be used to import estrogen's risk profile into testosterone GAHT. It does show that clinically meaningful hormone-associated thrombotic effects can become visible when enough data accumulate.

Why "no events observed" is not a safety verdict

The current AFAB-specific evidence is reassuring about the measured coagulation markers, but it is underpowered for rare VTE outcomes. The participants were young, an age group with low baseline VTE rates. The pooled studies also had limited follow-up, no duration-stratified analysis, and did not report baseline thrombophilia data such as Factor V Leiden, prothrombin G20210A, or antiphospholipid syndrome. They also did not control or report key acquired risk modifiers including BMI, smoking status, and prior oral-contraceptive exposure.

A 2021 research agenda by Safer et al. identifies the lack of adequately powered prospective data on long-term cardiovascular and thromboembolic outcomes as a major gap in transgender medicine. That gap matters most for people whom the small studies barely represent: older adults initiating GAHT, people with prior VTE, known thrombophilia, active smoking, or other established risk factors. Evidence on GAHT through midlife remains limited as well, as reviewed by Mehta et al. in 2024.

The cisgender-men signal cannot settle this question

A 2024 review by Kowalczyk Mullins et al. reports a VTE rate ratio of 1.63 (95% CI, 1.12 to 2.37) during the first six months of exogenous testosterone use in cisgender men. The confidence interval excludes 1.0, so this estimate was statistically significant in that population. It is the clearest quantitative VTE signal attached to exogenous testosterone in the available evidence.

It cannot be numerically transferred to young AFAB people starting testosterone GAHT. Cisgender men beginning testosterone treatment are often older, hypogonadal, and exposed to a different baseline cardiovascular and hormonal environment. But the finding also cannot be waved away: it shows that exogenous testosterone is not automatically VTE-neutral in every population. The 312-person AFAB meta-analysis cannot confirm or refute whether any comparable risk exists in its population.

Preparation and individual biology may matter

The meta-analysis pooled injectable esters, transdermal formulations, and subcutaneous preparations without route-specific subgroup analyses. That is a meaningful limitation because a 2022 prospective observational study by Schutte et al. found that transdermal GAHT had differential effects on inflammatory and hemostatic biomarkers. The study does not prove that one testosterone route causes more VTE than another, but it makes a route-agnostic conclusion less secure.

Population averages may also conceal individual variation. A 2023 prospective cohort study by Pallotti et al. found that androgen and estrogen receptor polymorphisms modified clinical outcomes in AFAB people receiving testosterone GAHT. It did not establish a personalized clot-risk algorithm, but it adds another reason to avoid treating a pooled mean as a guarantee for every person.

What to do with this evidence

For an AFAB person using testosterone GAHT without a personal or family history of VTE, known thrombophilia, active smoking, or other major risk factors, the available data offer reasonable reassurance: studies have not found the estrogen-like procoagulant pattern, and they have reported no VTE events. The better-established issue is rising hemoglobin and hematocrit, which supports periodic complete blood count monitoring.

For someone with a prior clot, known thrombophilia, high BMI, or active smoking, the evidence does not provide a reliable risk estimate because these groups were not adequately characterized in the studies. Decisions about baseline thrombophilia assessment, formulation, dose, and monitoring should be individualized with a clinician who can weigh the person's clot history against the benefits of GAHT.

Bottom line: Testosterone GAHT in AFAB people has not shown measurable procoagulant cascade activation in the small studies available. It consistently raises hematocrit, and the evidence remains too sparse to exclude a rare VTE risk—especially across formulations, longer treatment durations, and higher-risk patients.