Testosterone and hematocrit: what changes by formulation

Testosterone and hematocrit are linked in every delivery form, but by different amounts. A Bayesian network meta-analysis published in The Journal of Urology in 2022, pooling 29 randomized placebo-controlled trials and 3,393 men, reported mean hematocrit increases against placebo ranging from 1.4 percent on the patch to 4.3 percent on oral testosterone undecanoate.

Key takeaways

• The network meta-analysis published in The Journal of Urology in 2022 covered 29 placebo-controlled randomized trials and 3,393 men, and reported mean hematocrit increases against placebo of 3.0 percent for gel, 4.3 percent for oral testosterone undecanoate, 1.4 percent for the patch, 4.0 percent for intramuscular enanthate or cypionate and 1.6 percent for intramuscular undecanoate.
• When the formulations were compared against each other in that analysis, only one difference reached significance: intramuscular cypionate or enanthate against the patch. The authors wrote that “the clinical concern of this increase remains questionable.”
• A meta-analysis published in the Journal of Clinical Endocrinology and Metabolism in 2018, pooling four trials at low risk of bias and 1,779 patients, found a relative risk of developing erythrocytosis of 8.14 (95 percent CI 1.87 to 35.40).
• A systematic review published in Blood Advances in 2025 reported erythrocytosis rates “of up to 66.7%” and stated that “up to 2.7% of men on testosterone therapy developed thromboembolic events.”
• In the T4DM trial, published in Lancet Diabetes and Endocrinology in 2021, hematocrit above 54 percent occurred in 106 of 491 men (22 percent) on testosterone versus 6 of 484 (1 percent) on placebo.
• The European Association of Urology guidelines on male hypogonadism, 2026 edition, list a hematocrit of 54 percent or above among the contraindications and a baseline hematocrit of 48 to 50 percent among the relative contraindications.

Who does this page apply to?

This page applies to men with confirmed low testosterone. The European Association of Urology guidelines, 2026 edition, make a strong recommendation to use a total testosterone of 12 nmol/L (3.5 ng/mL) as the diagnostic threshold, measured fasting between 07:00 and 10:00 and confirmed on two separate occasions. It does not apply to men whose testosterone is normal, and no trial below tested them. The diagnostic steps are in how low testosterone is diagnosed in Canada. What testing your own case needs is a decision for your doctor or nurse practitioner.

Why does the link between testosterone and hematocrit matter?

Hematocrit is the share of blood volume made up of red cells, and testosterone raises it. It matters because it is the laboratory value most likely to interrupt treatment: guidelines set a hard threshold on it, and the trials that record it show the increase happening in every form tested. It is also the outcome where the route of delivery makes the clearest difference, which is why it belongs in any comparison between injections, gels, capsules and the less common forms.

What did the network meta-analysis find for each formulation?

The comparison most often cited is a systematic review and Bayesian network meta-analysis of randomized trials published in The Journal of Urology in 2022, covering 29 placebo-controlled randomized trials and 3,393 men. It is the first analysis of its kind on this question, and the authors explain why they built one: it was needed “given the absence of head-to-head trials.”

Read the table with its intervals. Oral undecanoate has the highest point estimate, 4.3 percent, and also the widest interval, 0.7 to 8.0, which means the underlying data are thin. The short-acting intramuscular esters sit close behind at 4.0 percent with a much tighter interval of 2.9 to 5.1.

Which formulation differences were statistically significant?

Only one. When the formulations were compared against each other rather than against placebo, intramuscular cypionate or enanthate was associated with a significantly higher increase than the patch, and no other between-formulation difference was detected. That is a narrower result than the table suggests at a glance: the point estimates rank the forms, but the evidence does not establish most of those rankings as real differences.

The authors also put a limit on the clinical meaning of the whole finding, writing that “the clinical concern of this increase remains questionable.” A page that quotes the table without that sentence is quoting half of the paper.

How common is erythrocytosis on testosterone therapy?

The estimates vary widely because the populations and definitions do. A meta-analysis published in the Journal of Clinical Endocrinology and Metabolism in 2018, restricted to four trials at low risk of bias covering 1,779 patients, found a relative risk of developing erythrocytosis of 8.14 (95 percent CI 1.87 to 35.40). The same analysis found small significant gains in sexual desire, erectile function and sexual satisfaction, and no effect on energy or mood.

A systematic review published in Blood Advances in 2025 pooled 45 studies, 35 of them on testosterone and other androgens, and reported erythrocytosis rates “of up to 66.7%,” with intramuscular formulations, higher doses and older age associated with increased risk. It also states that “up to 2.7% of men on testosterone therapy developed thromboembolic events.”

A narrative review published in Sexual Medicine Reviews in 2015 puts the size of the effect in one sentence: men undergoing testosterone therapy have “a 315% greater risk for developing erythrocytosis (defined as Hct > 0.52) when compared with control.” The same review states that “all TRT formulations cause increases in Hb and Hct, but injectables tend to produce the greatest effect.”

What do the injection studies show?

They show the highest rates, and the longest record. A two-year study published in the Journal of Clinical Endocrinology and Metabolism in 1997 followed 45 men on 200 mg of enanthate or cypionate every two weeks. Eleven of them, 24 percent, developed polycythemia requiring phlebotomy or temporary withholding of treatment, and a third of those cases occurred within the first year. About a third of the subjects discontinued therapy.

That study is small and old, but it is the one that quantifies the practical consequence rather than the laboratory change: about one man in four needed the treatment interrupted or blood removed. The wider injection evidence is set out in what the evidence says about testosterone injections.

Does subcutaneous differ from intramuscular?

In the one study that compared them, yes. A comparison published in The Journal of Urology in 2021 followed 234 hypogonadal men on 100 mg weekly, either intramuscular cypionate or subcutaneous enanthate, for 12 weeks. Trough total testosterone rose in both groups, from 313.6 to 536.4 ng/dL and from 246.6 to 552.8 ng/dL respectively, both with p less than 0.001, and after linear regression the modality was not associated with the total testosterone level (p = 0.057).

The difference was elsewhere. Subcutaneous administration was independently associated with a lower post-therapy hematocrit (p less than 0.001) and a lower estradiol (p less than 0.001), while neither modality raised PSA significantly (p = 0.965). The study was not randomized, and the route and the ester are confounded, so it cannot separate the effect of going under the skin from the effect of using a different ester.

What about nasal gel and pellets?

A matched-cohort analysis published in the Canadian Urological Association Journal in 2023 is the only study to place three routes side by side on this outcome. In 78 men matched for age, body mass index and sleep apnea over 16 weeks, hematocrit rose 4.4 percent on cypionate injection and 1.7 percent on pellets, and fell by 0.8 percent on nasal gel. Testosterone rose on all three routes, with p less than 0.001. The study is small, retrospective and not randomized.

Pellets have the longest follow-up. An analysis published in the European Journal of Endocrinology in 2010 followed 158 men treated for a mean of eight years and up to 21 years, and reported an independent odds ratio for polycythemia of 15.0 per log testosterone (95 percent CI 2.5 to 90.8) after adjusting for smoking and age, with duration of therapy not altering the risk. Neither pellets nor buccal testosterone is available in Canada, as set out in nasal, buccal and pellet testosterone.

Is the hematocrit rise linked to blood pressure?

In one oral testosterone study, the two moved together. A study published in the Journal of Cardiovascular Pharmacology and Therapeutics in 2021 used ambulatory blood pressure monitoring in 138 men over four months on oral testosterone undecanoate. Twenty-four-hour, awake and sleep systolic pressure rose 3.8, 5.2 and 4.3 mm Hg, hematocrit rose 3.2 percent and hemoglobin rose 0.9 g/dL.

The link is in the subgroup. Among men in the top quartile of hematocrit change, a rise of 6 to 14 percent, the systolic increase averaged 8.3 mm Hg, against 1.9 to 3.3 mm Hg in the lower three quartiles. That is an association within a single trial, not a demonstrated causal chain. The oral evidence is covered in oral testosterone and what the new capsules changed.

What do guidelines say about hematocrit thresholds?

The European Association of Urology guidelines on male hypogonadism, 2026 edition with a limited update in March 2026, list a hematocrit of 54 percent or above among the contraindications to testosterone therapy. The same guideline gives relative contraindications: an IPSS above 19, a baseline hematocrit of 48 to 50 percent, and a family history of venous thromboembolism. A value in that 48 to 50 percent band is treated as a reason for closer attention rather than an automatic bar.

That guideline also sets the timing of checks at 3, 6 and 12 months after starting treatment, then annually. What is checked and when is covered in testosterone therapy monitoring in Canada.

The 54 percent figure has a trial behind it. In the T4DM trial, published in Lancet Diabetes and Endocrinology in 2021 and randomizing 1,007 men aged 50 to 74, hematocrit above 54 percent occurred in 106 of 491 men (22 percent) on testosterone versus 6 of 484 (1 percent) on placebo, and the authors describe increases in hematocrit as potentially treatment limiting. Those participants were selected to be without pathological hypogonadism, which limits how far the rate transfers.

The same mechanism reads in two directions. The Testosterone Trials anemia analysis, published in JAMA Internal Medicine in 2017 in 788 men aged 65 and older, reports that 6 men who were not anemic at baseline reached a hemoglobin above 17.5 g/dL.

What the evidence does not show

• It does not show that the hematocrit rise causes harm. The authors of the 2022 network meta-analysis state directly that “the clinical concern of this increase remains questionable,” and no trial has tested whether lowering hematocrit on treatment changes outcomes.
• It does not establish most of the formulation rankings. In that analysis only intramuscular cypionate or enanthate against the patch reached significance; every other between-formulation comparison did not.
• The oral figure is imprecise. Oral testosterone undecanoate had the highest point estimate at 4.3 percent but a credible interval of 0.7 to 8.0, wide enough that its position in the ranking is not settled.
• The three-way route comparison rests on 78 men over 16 weeks in a retrospective matched cohort, and no randomized trial has compared nasal, injection and pellets on this outcome.
• The subcutaneous versus intramuscular comparison was not randomized and confounds route with ester, so the lower hematocrit on subcutaneous administration cannot be attributed to the route alone.
• The pellet estimate is imprecise, with an odds ratio of 15.0 and a 95 percent confidence interval of 2.5 to 90.8.
• The blood pressure association comes from quartiles within one 138-man trial and does not establish that a rising hematocrit raises blood pressure.

Frequently asked questions

Which testosterone formulation raises hematocrit the most?

In the network meta-analysis published in The Journal of Urology in 2022, covering 29 randomized placebo-controlled trials and 3,393 men, the highest mean increases against placebo were oral testosterone undecanoate at 4.3 percent (95 percent CI 0.7 to 8.0) and intramuscular enanthate or cypionate at 4.0 percent (2.9 to 5.1). Only the comparison of intramuscular short esters against the patch reached statistical significance.

How common is erythrocytosis on testosterone therapy?

A 2018 meta-analysis in the Journal of Clinical Endocrinology and Metabolism, pooling four trials at low risk of bias and 1,779 patients, found a relative risk of 8.14 (95 percent CI 1.87 to 35.40). A 2025 systematic review in Blood Advances reported rates “of up to 66.7%” across 45 studies, with intramuscular formulations, higher doses and older age associated with increased risk.

What hematocrit level is treated as too high?

The European Association of Urology guidelines on male hypogonadism, 2026 edition, list a hematocrit of 54 percent or above among the contraindications to testosterone therapy, and a baseline hematocrit of 48 to 50 percent among the relative contraindications alongside an IPSS above 19 and a family history of venous thromboembolism. The same guideline sets monitoring at 3, 6 and 12 months, then annually.

Do injections raise hematocrit more than gels?

The point estimates say so and the statistics do not confirm it. The 2022 network meta-analysis reported 4.0 percent for intramuscular enanthate or cypionate against 3.0 percent for gel, but the only between-formulation difference that reached significance was intramuscular short esters against the patch. A 2015 review in Sexual Medicine Reviews states that injectables “tend to produce the greatest effect.”

Does subcutaneous injection lower the hematocrit rise?

In a 12-week comparison of 234 men published in The Journal of Urology in 2021, subcutaneous enanthate was independently associated with a lower post-therapy hematocrit than intramuscular cypionate at the same weekly dose (p less than 0.001), along with a lower estradiol. The study was not randomized and the route and the ester are confounded, so the finding cannot be assigned to the route alone.

Does a higher hematocrit mean testosterone is dangerous?

That is unknown. The authors of the 2022 network meta-analysis in The Journal of Urology wrote that “the clinical concern of this increase remains questionable.” A 2025 Blood Advances review states that “up to 2.7% of men on testosterone therapy developed thromboembolic events,” but no trial has tested whether reducing hematocrit during treatment changes clinical outcomes. Discuss your own blood work with your doctor or pharmacist.

Related reading: testosterone gels and creams and testosterone esters explained.

References

1. Nackeeran 2022. The Effect of Route of Testosterone on Changes in Hematocrit: A Systematic Review and Bayesian Network Meta-Analysis of Randomized Trials. The Journal of Urology. PMID 34445892. DOI: https://doi.org/10.1097/JU.0000000000002188
2. Ponce 2018. The efficacy and adverse events of testosterone replacement therapy in hypogonadal men. Journal of Clinical Endocrinology and Metabolism. PMID 29562341. DOI: https://doi.org/10.1210/jc.2018-00404
3. Liu 2025. Diagnosis, management, and outcomes of drug-induced erythrocytosis: a systematic review. Blood Advances. PMID 39913688. DOI: https://doi.org/10.1182/bloodadvances.2024015410
4. Reddy 2023. Prevalence of secondary erythrocytosis: a matched-cohort analysis of intranasal gel, injections, and pellets. Canadian Urological Association Journal. PMID 37068153. DOI: https://doi.org/10.5489/cuaj.8210
5. Hajjar 1997. Journal of Clinical Endocrinology and Metabolism. PMID 9360543. DOI: https://doi.org/10.1210/jcem.82.11.4387
6. Ip 2010. Trough testosterone predicts polycythemia over up to 21 years of pellets. European Journal of Endocrinology. PMID 19903801. DOI: https://doi.org/10.1530/EJE-09-0717
7. Choi 2021. Comparison of Outcomes for Hypogonadal Men Treated with Intramuscular Testosterone Cypionate versus Subcutaneous Testosterone Enanthate. The Journal of Urology. PMID 34694927. DOI: https://doi.org/10.1097/JU.0000000000002301
8. Jones 2015. Sexual Medicine Reviews. PMID 27784544. DOI: https://doi.org/10.1002/smrj.43
9. White 2021. Effects of a Novel Oral Testosterone Undecanoate on Ambulatory Blood Pressure. Journal of Cardiovascular Pharmacology and Therapeutics. PMID 34191621. DOI: https://doi.org/10.1177/10742484211027394
10. Wittert 2021. Testosterone treatment to prevent or revert type 2 diabetes in men enrolled in a lifestyle programme (T4DM): a randomised, double-blind, placebo-controlled, 2-year, phase 3b trial. Lancet Diabetes and Endocrinology. PMID 33338415. DOI: https://doi.org/10.1016/S2213-8587(20)30367-3
11. Roy 2017. Association of Testosterone Levels With Anemia in Older Men: A Controlled Clinical Trial. JAMA Internal Medicine. PMID 28241237. DOI: https://doi.org/10.1001/jamainternmed.2016.9540
12. European Association of Urology. Guidelines on Sexual and Reproductive Health, male hypogonadism chapter. 2026 edition, limited update March 2026. https://uroweb.org/guidelines/sexual-and-reproductive-health/chapter/male-hypogonadism

This page is educational information, not medical advice. Testosterone is a prescription medication and a controlled substance in Canada. Talk to your doctor or pharmacist about your own situation.