Do Steroids Cause High Hematocrit? 7 Critical Facts

Data in brief

Three figures frame the whole question: how much androgens move red cells, how often it crosses into a diagnosis, and where the number that everyone quotes as the danger line actually comes from.

What this guide covers

Do steroids cause high hematocrit? The direct answer

Yes, and the relationship is dose dependent rather than incidental. In a dose response study, men received a monthly GnRH agonist plus one of five weekly doses of testosterone enanthate, 25, 50, 125, 300 or 600 mg intramuscularly, for 20 weeks. Hemoglobin and hematocrit rose in a linear, dose dependent fashion in both the 61 young men and the 60 older men, and the rise was significantly greater in the older group (PMID 18160461).

At a therapeutic dose the change is moderate. Testosterone gel for six months in older men raised hemoglobin by about 7 per cent and hematocrit by about 10 per cent (PMID 24158761). At the doses used outside medicine there is no equivalent controlled trial, which is worth saying plainly: the dose response line above stops at 600 mg a week, and nobody has run the experiment beyond it.

How often that rise crosses into a diagnosis of erythrocytosis depends heavily on the preparation. The 2025 systematic review reports rates up to 66.7 per cent in cisgender and transgender men on prescription testosterone, and names intramuscular formulations, higher doses and older age as the risk factors (PMID 39913688).

What actually drives the rise, and what only gets repeated

Four of the rows below rest on controlled data. Three of them are things this audience says constantly and nobody has measured.

Why it happens, and the part that is still disputed

The accepted outline is that androgens push red cell production two ways at once: erythropoietin rises, and hepcidin, the protein that restricts iron availability, falls, which frees more iron for the marrow. In the testosterone gel trial, erythropoietin rose while ferritin and hepcidin fell at one and three months. By six months both had drifted back toward baseline even though hemoglobin and hematocrit stayed high, which the authors read as a new set point rather than a continuing push (PMID 24158761).

Here is the part most pages skip. In the dose response study, there was no significant difference in the percent change in erythropoietin across the five dose groups, and changes in erythropoietin did not correlate with changes in total or free testosterone (PMID 18160461). Two well conducted studies, two different pictures of the same mechanism. A 2024 review sets out the fuller version, with an initial erythropoietin rise, a new set point, a parallel fall in hepcidin, and several other candidate mechanisms alongside them (PMID 39212549).

What follows from that honestly is narrow: the direction is certain, the size is dose related, and the exact pathway is not settled. Anyone explaining this in one sentence is overstating what has been shown.

Where the danger line actually comes from

Ask why 45, or 50, or 54 per cent is the number that matters and the answer almost always traces back to one trial. It is worth looking at directly, because what it shows is strong and what it does not cover is important.

CYTO-PV randomized 365 adults with JAK2 positive polycythemia vera to a target below 45 per cent or a target of 45 to 50 per cent. Over a median of 31 months, death from cardiovascular causes or major thrombosis occurred in 5 of 182 patients in the tighter group, 2.7 per cent, against 18 of 183, 9.8 per cent, in the looser one, a hazard ratio of 3.91 with a confidence interval of 1.45 to 10.53 (PMID 23216616).

That is a real result, and it is the reason clinicians take a rising hematocrit seriously rather than shrugging at it. It is also a trial in a myeloproliferative disease, in patients being treated with phlebotomy or hydroxyurea, at a median age far above that of most people reading this. No trial has randomized androgen users to a hematocrit target. Treating the 45 per cent figure as a proven line for a lifter on cycle is an extrapolation, and the honest framing is that it is the best available anchor rather than a measured threshold for this group.

What has been observed directly in this population is narrower and still serious: thromboembolic events in up to 2.7 per cent of men on testosterone therapy across the studies reviewed in 2025 (PMID 39913688), and a documented case literature of testosterone use producing erythrocytosis (PMID 29038321). The Endocrine Society guideline lists elevated hematocrit among the conditions against starting testosterone therapy at all, and puts hematocrit in the standard monitoring plan for the first year (PMID 29562364).

When it is not the steroids

A high hematocrit on a report is a ratio of cells to plasma, which means half the ways of raising it do not involve making a single extra red cell. Before anyone concludes that a compound did it, these belong on the list.

The practical consequence is that a single high reading taken the morning after a hard session in the heat is a weak basis for any decision, and a repeat draw under ordinary conditions settles more than an argument about compounds does.

Androgens against every other drug that does this

One more thing that review settled, at least for a neighbouring drug class: patients who stopped the SGLT-2 inhibitor showed improvement or resolution of the erythrocytosis. For androgens the equivalent question, how completely and how fast the hematocrit comes down, is handled on the management page, where the honest answer includes that the evidence supporting therapeutic phlebotomy in this setting is described as lacking (PMID 39212549).

Six claims about high hematocrit that do not survive

Each of these is standard gym advice. Each one is either contradicted or simply unmeasured.

Where people go wrong reading this

• Treating a percentage as a count. Hematocrit is the share of blood volume made of red cells, so plasma volume moves it without any change in red cell production (PMID 10694114).
• Acting on one draw. A repeat under ordinary conditions, not after a hard session or a hot day, settles more than any argument about compounds.
• Importing a threshold without saying where it came from. The evidence behind it is strong in the disease it was tested in, and absent in this one.
• Ignoring the rest of the panel. Pressure and lipids move in this population too. See the blood pressure page and the cholesterol page.
• Skipping the baseline. Without a pre-cycle reading there is nothing to compare against, which is the argument made in the baseline guide.

British spellings of these markers, haematocrit and haemoglobin, refer to exactly the same measurements and appear on reports outside the United States.