TRT Does Not Cause Heart Attacks (But Blasting Is Different)

Menno Henselmans posted a clean version of the point last week on Instagram: TRT does not raise cardiovascular risk when it restores testosterone from low to normal, and there is a world of difference between that and taking androgens to get jacked. He is right. The problem is that the reassuring headline coming off TRAVERSE, and the FDA dropping the cardiovascular boxed warning off testosterone labels in February 2025, has been read across as "testosterone is safe," full stop. That reading does not survive contact with either the trial's exclusion criteria or the 1,189-man Danish AAS cohort published in Circulation six months later.

Here is what an athlete actually needs to take away from the last three years of testosterone-and-CV literature. The good news is real. The bad news is separate, and neither cancels the other.

What TRAVERSE actually showed

TRAVERSE is the trial (Lincoff et al., NEJM 2023, PMID 37326322) that finally put the "testosterone causes heart attacks" panic to rest for hypogonadal men on replacement doses. Understanding what it did and did not test is the whole game for anyone using this data to justify their own protocol.

The trial that settled the Vigen and Finkle scare

Two poorly-designed papers built the boxed warning. Vigen et al. 2013 in JAMA (PMID 24193080) reported a 5.8% absolute risk increase for the composite of death, MI and stroke in VA men who started testosterone after coronary angiography. The paper had to publish a correction after critics found statistical errors and 29 women in what was reported as an all-male cohort (JAMA 2014;311(9):961-2). Finkle et al. 2014 in PLoS ONE (PMID 24489673) reported a 36% relative risk increase in non-fatal MI in the 90 days after a testosterone prescription, using a design that could not separate the drug from the doctor visit that led to the prescription.

TRAVERSE was designed to litigate those signals with a proper randomised controlled trial. 5,246 hypogonadal men aged 45-80 with cardiovascular disease or high CV risk, two fasting total T measurements under 300 ng/dL, randomised to transdermal 1.62% testosterone gel titrated to a total T of 350-750 ng/dL versus matching placebo. Mean treatment 21.7 months, mean follow-up 33 months. It is the biggest cardiovascular outcomes trial testosterone has ever seen, and it will stay the biggest for the foreseeable future.

The MACE result, and the signals nobody talks about

The primary endpoint was a composite of cardiovascular death, non-fatal MI and non-fatal stroke. 182 of 2,596 in the testosterone arm (7.0%) and 190 of 2,602 in the placebo arm (7.3%). Hazard ratio 0.96 (95% CI 0.78-1.17), P less than 0.001 for non-inferiority. TRT was not worse than placebo for major cardiac events. That is the number every headline ran and it holds up cleanly.

Three secondary signals appeared alongside it, and they matter more than the coverage suggested.

Atrial fibrillation: 3.5% on testosterone versus 2.4% on placebo, P = 0.02. The paper reported percentages and P values but no hazard ratio, which several 2024-2026 reviews have flagged. The independent TriNetX replication in the Journal of Sexual Medicine (Greenberg et al. 2024, PMID 39487489) put the AF relative risk at 1.48 (95% CI 0.93-2.37, non-significant when other studies were pooled). A pooled meta-analysis by Corona et al. 2024 across 106 RCTs gave an OR of 1.44 for AF, non-significant.

Pulmonary embolism: 0.9% versus 0.5%. A 2025 secondary analysis in the American Heart Journal (Gagliano-Jucá et al., PMID 40246046) linked TRT-induced rises in neutrophils and monocytes to the VTE signal, with per-standard-deviation odds ratios of 1.32 and 1.39.

Acute kidney injury: 2.3% versus 1.5%, P = 0.04. The Greenberg TriNetX replication put AKI relative risk at 1.53 (95% CI 1.07-2.18), significant.

None of these individually derailed the primary conclusion. Together they are a reminder that even at replacement doses, TRT is not neutral on every axis. At bodybuilder doses, every mechanism plausibly behind these signals (erythrocytosis, sympathetic tone, aromatisation to oestradiol, endothelial stress) is amplified.

The FDA's February 2025 label change removed the cardiovascular boxed warning on the strength of TRAVERSE. The same label change added a blood pressure warning based on postmarketing ambulatory BP data. The reassuring headline was more complicated than most people read.

Who was actually in TRAVERSE (this matters)

Baseline median total T was 227 ng/dL, deep in the hypogonadal range. On-treatment median rose 148 ng/dL to sit around 375 ng/dL at 12 months, at the very low end of the 350-750 ng/dL target band. Multiple critical reviews (Morgentaler in JACC Advances 2023, Krishnan et al. in Progress in Cardiovascular Diseases 2024) noted TRAVERSE participants were effectively undertreated by clinical standards.

For this article's purpose the under-treatment is a feature, not a bug. It means the trial cleared testosterone at total T concentrations well below the reference-range mid-point. Anything above the reference ceiling is off the map the trial drew.

What TRAVERSE does not tell you

The inclusion criteria that quietly exclude every athlete

TRAVERSE excluded men with haematocrit above 50%, uncontrolled hypertension, recent CV events, and PSA above 3.0 ng/mL. Read those four exclusions against the enhanced-athlete population: a large fraction of TRT-using bodybuilders sit at haematocrit 52-56%, many have systolic BP above 135 mmHg on cycle, and PSA rises are common on cycle. The trial screened out exactly the people whose CV risk profile would be modified most by testosterone.

The other silent exclusion is duration. TRAVERSE followed men for a mean 33 months. Career-long enhanced athletes are looking at 10 to 30 years of exposure. Nothing in TRAVERSE speaks to cumulative-dose CV remodelling over decades.

On-treatment total T versus what a cruise actually looks like

TRAVERSE's target ceiling was 750 ng/dL. Median achieved was 375 ng/dL. Bhasin's 2001 dose-response study in healthy young men (PMID 11701431) is the cleanest map from weekly injectable testosterone to trough total T, in a design that suppressed endogenous production:

A 200 mg per week cruise interpolates to around 800-1,000 ng/dL at trough, with peaks 24-48 hours post-injection typically 1.6 to 1.8 times higher. A 500 mg per week blast projects to roughly 2,000 ng/dL at trough. TRAVERSE tested testosterone at a total T around 375 ng/dL. The gap between the trial and a typical cruise is not a small extrapolation. It is 2 to 5 times the on-treatment concentration the trial cleared.

The 2025 Danish cohort, the counter-signal

The dataset that closes the loop is Windfeld-Mathiasen et al. in Circulation 2025 (PMID 39945117). 1,189 men sanctioned by the Danish anti-doping authority for AAS use in fitness centres between 2006 and 2018, matched against 59,450 age-matched general-population controls, followed a median 11 years through mid-2023. Adjusted hazard ratios in the AAS users:

• Acute myocardial infarction: 3.00
• Cardiomyopathy: 8.90
• PCI or CABG: 2.95
• Heart failure: 3.63
• Venous thromboembolism: 2.42
• Arrhythmias: 2.26

This is where the "3x MI, 9x cardiomyopathy" number circulating on podcasts actually comes from. Real cohort, real follow-up, real magnitudes. It measures a different population from TRAVERSE, on different doses, over a longer time-horizon. Both papers are true. Both apply to their own populations. Neither cancels the other.

The line is not milligrams per week

The most common mistake in this whole conversation is treating a milligram number as a proxy for cardiovascular risk. The honest number is not on your syringe. It is on your bloodwork.

Why 200 mg per week is not one thing

A 60 kg lean man on 200 mg per week with high SHBG might sit at 1,100 ng/dL at trough. A 100 kg man on the same 200 mg per week with low SHBG might trough at 700 ng/dL. Same dose, different labs, different position relative to the reference-range ceiling. The variables that push these two men apart are body mass, SHBG, injection frequency and ester. None of them are on the label of the vial.

Injection frequency is the biggest lever most protocols ignore. Once-weekly IM cypionate produces a peak-to-trough ratio around 1.6 to 1.8. Split the same weekly dose across two or three injections and that swing roughly halves, moving the trough up by 200-400 ng/dL and the peak down by 400-600 ng/dL. The same milligram number can look like replacement at trough and clearly supraphysiological at peak, depending on when the blood draw happened.

The lab-value threshold worth caring about

The harmonised LC-MS/MS reference range for total testosterone in non-obese men aged 19-39 is 264-916 ng/dL, established across four cohorts including Framingham and EMAS (Travison et al., 2017, PMID 28324103). This is the range LabCorp adopted in 2017 and the Endocrine Society Clinical Practice Guideline (Bhasin et al., 2018, PMID 29562364) endorses as the reference for CDC-traceable assays.

Two numbers matter:

Above ~916 ng/dL (LC-MS/MS) or ~1,000-1,100 ng/dL (most immunoassays): your bloodwork will flag "H" for high. This is the honest first threshold. A protocol that produces sustained trough total T above this ceiling is not replacement in the clinical sense, regardless of what the milligram number says.

Above ~750 ng/dL: TRAVERSE's dose-titration protocol would have reduced your dose. The trial data does not clear anything above this. If a reader is leaning on "TRAVERSE cleared TRT" to justify their own numbers, 750 ng/dL is the trial ceiling to compare against.

Reading your own bloodwork for the "still TRT" boundary

Draw at trough. Immediately before the next injection, not two days after it. The peak-versus-trough gap is often the difference between "this looks fine" and "this is clearly outside the trial evidence." The Endocrine Society explicitly recommends trough draws for exactly this reason.

Use LC-MS/MS on a CDC-traceable assay if you can. Immunoassays are cheaper and more common but lose specificity above the physiological range, which is where the number matters most if you are above the reference ceiling.

Mechanism, how supraphysiological doses damage the heart

The TRAVERSE-versus-Windfeld-Mathiasen split is not two contradictory papers. It is one biology behaving differently at two exposure levels. Four mechanisms scale with dose. Each has its own bloodwork signature.

Erythrocytosis and viscosity

Coviello 2008 (PMID 18160461) is the cleanest dose-response for haematocrit. In GnRH-suppressed men on 25, 50, 125, 300 or 600 mg of weekly testosterone enanthate over 20 weeks, both haemoglobin and haematocrit rose linearly and dose-dependently. The response was steeper in older men. The rise was not explained by erythropoietin or transferrin-receptor changes, meaning the mechanism is androgen-direct, not classic hypoxia-driven.

The threshold that matters is haematocrit at or above 54%, where AUA and Endocrine Society guidance both recommend holding or de-escalating testosterone. Ory et al. 2022 (PMID 35050717) followed more than 50,000 hypogonadal men on testosterone therapy and showed that secondary polycythaemia raised first-year MACE and venous thromboembolism risk versus TRT users who did not develop it. This is the mechanism that can kill someone this month rather than this decade.

LVH from androgen-driven cardiac remodelling

The HAARLEM study (Smit et al., 2021, PMID 36304014) followed 100 amateur bodybuilders through a self-directed cycle with 3D echocardiography before, at the end, and roughly a year post-cycle. LV mass rose 28.3 grams (95% CI 14.2-42.4, P less than 0.001) across a median 13-week cycle at a mean 901 mg per week AAS-equivalent. Systolic and diastolic function both worsened. In this amateur cohort, LV mass fully normalised by a median 8 months post-cessation.

That is the amateur-short-cycle reassuring finding. Baggish et al. 2017 (PMID 28533317) is what happens further out. In 86 long-term AAS users versus 54 non-user weightlifters (age 34-54), LVEF was 52 ± 11% in users versus 63 ± 8% in non-users (P less than 0.001). Diastolic function was worse. Coronary plaque volume was 3 mm³ (users) versus 0 mm³ (non-users), and plaque burden scaled with cumulative lifetime AAS exposure. HAARLEM measured the reversible part. Baggish measured the part that does not un-form.

HDL crash and ApoB drift

Thompson 1989 (PMID 2915439) is old but the mechanism has not changed. In 11 male weightlifters running crossover 6-week protocols, testosterone enanthate at 200 mg per week dropped HDL-C by 9%. Oral stanozolol at 6 mg per day dropped HDL-C by 33% and the HDL2 subfraction by 71%. The 17-alpha-alkylated orals (winstrol, anavar, dianabol, superdrol) induce hepatic lipase, which strips HDL2 and collapses reverse cholesterol transport in a way injectable testosterone alone does not.

ApoB is the marker that separates "LDL looks fine" from "particle count says you are cooking your arteries." HAARLEM 2022 showed LDL-C rising about 0.45 mmol/L during a cycle while ApoB rose about 18 mg/dL. When androgens shift particles toward small, dense, cholesterol-poor LDL, the LDL-C number under-reports the atherogenic burden. ApoB catches it because every atherogenic particle carries exactly one ApoB. Detail on the compound-by-compound lipid picture is in the cholesterol on steroids article.

Blood pressure

HAARLEM (Smit et al., 2022, PMID 35014715) documented mean systolic BP rising 6.87 mmHg (95% CI 4.34-9.40) and diastolic BP rising 3.17 mmHg (1.29-5.04) during self-administered cycles. 41% of subjects were hypertensive by clinic BP at some point during the cycle. Clinic BP is a single snapshot; home ambulatory BP monitoring detects the sustained cycle-related elevation the clinic misses. The management side is covered in how to lower blood pressure on cycle with telmisartan and nebivolol.

The reversibility question

This is where the honest article separates from the reassuring one. What comes back after cessation depends on what you measured and how long you were exposed.

Reversible in short-cycle amateur cohorts (HAARLEM):

• HDL-C rebounds within weeks to a few months, fastest after injectable-only cycles.
• Haematocrit returns to baseline within 3-6 months (RBC lifespan is roughly 120 days).
• Blood pressure normalised by 3 months post-cycle in HAARLEM.
• LV mass fully normalised by a median 8-12 months in the amateur cohort.

Not fully reversible in long-term users:

• Coronary plaque accumulated during cycles does not regress. Baggish 2017 plaque volume scaled with cumulative lifetime AAS exposure; each 10-year increment added burden. This is the substrate for MACE 10 to 20 years after the last cycle.
• Coronary microvascular dysfunction persists in ex-users. Bulut et al. 2024 in JAMA Network Open (PMID 39680410) showed impaired myocardial flow reserve in both current and former AAS users, with cumulative duration independently predicting reduced flow reserve.
• Systolic LV dysfunction in long-term users. Abdullah et al. 2024 in European Journal of Preventive Cardiology (DOI 10.1093/eurjpc/zwad362) followed 101 users (mean 11 accumulated years of AAS) versus 71 non-users. Severely reduced LVEF at or below 40% appeared in roughly 10% of current users. Former users showed persistently impaired systolic function and elevated blood pressure years after stopping.

The pattern is not "reversible" or "not reversible." It is a variable that scales with cumulative years of exposure. Everything you can measure on a routine panel comes back after a short cycle. What does not come back is what you cannot see without a coronary CT and a stress PET. The first cycle behaves like a different drug from the tenth year of use.

Bloodwork that matters if you are above the reference range

If your on-treatment total T is above 916 ng/dL, the TRAVERSE data does not cover you. The panel below is what does. Order matters, because time and money are finite.

Home blood pressure, first

The HAARLEM 41% hypertensive figure came from clinic readings. Home BP monitoring catches the sustained cycle-related elevation the clinic misses. Validated automated upper-arm cuff, correct cuff size, seated with back supported and feet flat for 5 minutes rest, arm at heart level. Two readings 1 minute apart, morning and evening, on 7 consecutive days. Discard day 1. Average the remaining readings. Report the mean.

ACC/AHA home BP threshold for hypertension is 130/80. Target on-cycle is under 130/80. Sustained at or above 135/85 is where dose reduction or an antihypertensive becomes the conversation.

Haematocrit and haemoglobin

The marker that could kill you this month rather than this decade. Haematocrit above 54% is the AUA and Endocrine Society intervention threshold. Ory 2022 tied polycythaemia to first-year MACE and VTE in a cohort of more than 50,000 men, at both the 52% and 54% cut-offs.

Test at baseline, mid-cycle around week 6-8, end of cycle, and post-cycle 12 weeks. For chronic cruise, quarterly at minimum. Detail on the hematocrit management side (donation, dose splitting, ester choice) is in how to lower hematocrit on TRT without blood donation, with the reverse-side risk of over-donation in the iron paradox on TRT.

ApoB, and Lp(a) once

ApoB replaces LDL-C as the atherogenic-particle number on cycle. 2026 ACC/AHA dyslipidaemia guideline tier: under 90 mg/dL for lower risk, under 70 mg/dL for higher risk, under 55 mg/dL for very-high-risk secondary prevention. Test every cycle.

Lp(a) is roughly 70-90% genetic and does not move on cycle. Test once, ever. If elevated, the whole lifetime CV strategy shifts toward tighter ApoB targets and earlier statin consideration. NLA 2024 defines above 125 nmol/L as high risk, under 75 nmol/L as low.

hs-CRP, two readings apart

hs-CRP in chronic AAS users runs roughly 3-4 times higher than in drug-free lifters (Grace and Davies 2004, PMID 14751958). Within-subject coefficient of variation is roughly 44%, so single readings do not represent basal level. Two readings at least 2 weeks apart, both in the absence of acute illness or recent hard training, is the convention. Discard any reading over 10 mg/L (acute inflammation, not vascular risk) and retest.

hs-troponin-I and NT-proBNP, only for cause

Neither is a routine on-cycle screen. There is no strong evidence that resting hs-troponin-I is chronically elevated in asymptomatic AAS users, and NT-proBNP stayed within the normal range in Rasmussen's 16-year follow-up despite users having larger LV mass and lower LVEF. Order these markers if there are red-flag symptoms (chest pain, unexplained dyspnoea, palpitations, syncope), if an echo shows LVH or reduced ejection fraction, or in the workup of any acute cardiac event. Ordering them quarterly as a "am I damaging my heart" check will produce normal results even when subclinical structural change is developing.

Testing intervals in one paragraph

Baseline 2-4 weeks before a cycle: full panel plus Lp(a) once. Mid-cycle week 6-8: haematocrit, lipids with ApoB, LFTs, home BP week-average. End of cycle: repeat full panel. Post-cycle 8-12 weeks: repeat CBC, lipids, hs-CRP, BP, and the HPTA panel (total T, LH, FSH, E2). If BP, HDL and haematocrit have not normalised by 12 weeks post-cycle, persistent remodelling is on the table and an echo is worth ordering. Comprehensive panel structure for enhanced athletes is in the complete haematology panel for enhanced athletes.

Practical recommendations

If you are on replacement TRT

You are the population TRAVERSE actually studied. Keep your trough total T inside 500-800 ng/dL, haematocrit under 52%, ApoB under 90 mg/dL, home BP under 130/80. Recheck every 3-6 months. The TRAVERSE MACE result is your evidence, and the AF, PE and AKI secondary signals are real but small at replacement doses. This article's argument is not against you.

If you are considering starting TRT, the starting TRT in Australia article covers the clinic and monitoring side, and the TRT haemoglobin and haematocrit article covers the erythrocytosis question in more depth.

If you are on a cruise dose above the reference range

Your on-treatment total T is likely 900-1,500 ng/dL. TRAVERSE does not apply directly to you, and the Windfeld-Mathiasen cohort's 3x MI and 9x cardiomyopathy signals come from users at your dose range and above. The question worth asking is not "is this safe" but "what is my exposure trajectory over the next 10 years."

Two levers matter. Split your dose across at least two injections per week to bring your trough up and your peak down, so the labs read what you actually live at. And run the bloodwork panel above every 3-4 months rather than every 6-12. Erythrocytosis, hypertension and ApoB drift are the parts you can catch early and modify.

If you are blasting

Cycles above 500 mg per week, especially stacked with 19-nors (nandrolone, trenbolone) or 17-alpha-alkylated orals (oxymetholone, stanozolol), put you well outside every published cardiac safety dataset. The reversibility literature (HAARLEM) says a single short cycle in a young athlete recovers on most markers. The persistence literature (Baggish, Abdullah, Rasmussen) says cumulative years of exposure lay down coronary plaque and microvascular dysfunction that do not.

Practical minimum: baseline, mid-cycle and end-of-cycle bloodwork on the panel above, plus home BP monitoring. Echo after 5 accumulated years of blast-and-cruise, or earlier if home BP is running above 135/85 during blocks. See the on-cycle support protocol article for the supplement side of harm reduction, and trenbolone bloodwork for the harshest of the harsh compounds.

Key takeaways

• TRAVERSE (Lincoff et al., NEJM 2023) cleared replacement-dose testosterone for major cardiac events in 5,246 hypogonadal men. MACE HR 0.96. This is real, holds up, and rehabilitates TRT for men who need it.
• The trial titrated to 350-750 ng/dL and achieved a median of about 375 ng/dL on treatment. The FDA dropped the cardiovascular boxed warning in February 2025 based on this data. A blood pressure warning was added at the same time.
• Windfeld-Mathiasen et al. in Circulation 2025 followed 1,189 sanctioned AAS users for a median 11 years and found 3x MI risk, 9x cardiomyopathy risk, and 2.4x VTE risk versus general-population controls. Both papers are true. They apply to different populations at different doses.
• The line between "TRT data applies to you" and "TRT data does not apply to you" is your on-treatment total testosterone, not milligrams per week. Above roughly 916 ng/dL on a harmonised assay, TRAVERSE says nothing.
• Milligrams is a bad proxy because body mass, SHBG, injection frequency and ester change the trough total T by hundreds of ng/dL at the same dose.
• Mechanistically, supraphysiological androgen use damages the heart through erythrocytosis (dose-linear, Coviello 2008), LVH (HAARLEM Smit 2021), HDL crash driven by oral 17-alpha-alkylated compounds (Thompson 1989), ApoB drift, and blood pressure elevation of about 6.87 mmHg systolic during cycles (HAARLEM Smit 2022).
• HAARLEM shows short-cycle amateur users normalise blood pressure, HDL, haematocrit and LV mass within 8-12 months of cessation. Baggish 2017, Abdullah 2024 and Bulut 2024 show long-term users accumulate coronary plaque and microvascular dysfunction that persist after stopping.
• Bloodwork panel for anyone above the reference range: home BP first, haematocrit second, ApoB third, hs-CRP fourth, Lp(a) once. hs-troponin and NT-proBNP only for symptoms or abnormal echo.

---

References

1. Lincoff, A. M., Bhasin, S., Flevaris, P., et al. (2023). Cardiovascular safety of testosterone-replacement therapy. New England Journal of Medicine, 389(2), 107-117. PubMed
2. Bhasin, S., Travison, T. G., Pencina, K. M., et al. (2023). Prostate safety events during testosterone replacement therapy in men with hypogonadism: A randomized clinical trial. JAMA Network Open, 6(12), e2348692. PubMed
3. Vigen, R., O'Donnell, C. I., Barón, A. E., et al. (2013). Association of testosterone therapy with mortality, myocardial infarction, and stroke in men with low testosterone levels. JAMA, 310(17), 1829-1836. PubMed (Correction: JAMA. 2014;311(9):967)
4. Finkle, W. D., Greenland, S., Ridgeway, G. K., et al. (2014). Increased risk of non-fatal myocardial infarction following testosterone therapy prescription in men. PLoS ONE, 9(1), e85805. PubMed
5. Corona, G., Rastrelli, G., Di Pasquale, G., et al. (2018). Testosterone and cardiovascular risk: Meta-analysis of interventional studies. Journal of Sexual Medicine, 15(6), 820-838. PubMed
6. Corona, G., Rastrelli, G., Sparano, C., et al. (2024). Cardiovascular safety of testosterone replacement therapy in men: An updated systematic review and meta-analysis. Expert Opinion on Drug Safety, 23(5), 565-579. PubMed
7. Windfeld-Mathiasen, J., Heerfordt, I. M., Dalhoff, K. P., et al. (2025). Cardiovascular disease in anabolic androgenic steroid users. Circulation, 151(12), 828-834. PubMed
8. Baggish, A. L., Weiner, R. B., Kanayama, G., et al. (2017). Cardiovascular toxicity of illicit anabolic-androgenic steroid use. Circulation, 135(21), 1991-2002. PubMed
9. Smit, D. L., Buijs, M. M., de Hon, O., den Heijer, M., & de Ronde, W. (2022). Prospective study on blood pressure, lipid metabolism and erythrocytosis during and after androgen abuse. Andrologia, 54(4), e14372. PubMed
10. Smit, D. L., Voogel, A. J., den Heijer, M., & de Ronde, W. (2021). Anabolic androgenic steroids induce reversible left ventricular hypertrophy and cardiac dysfunction. Echocardiography results of the HAARLEM study. Frontiers in Reproductive Health, 3, 732318. PubMed
11. Coviello, A. D., Kaplan, B., Lakshman, K. M., et al. (2008). Effects of graded doses of testosterone on erythropoiesis in healthy young and older men. Journal of Clinical Endocrinology & Metabolism, 93(3), 914-919. PubMed
12. Thompson, P. D., Cullinane, E. M., Sady, S. P., et al. (1989). Contrasting effects of testosterone and stanozolol on serum lipoprotein levels. JAMA, 261(8), 1165-1168. PubMed
13. Ory, J., Nackeeran, S., Balaji, N. C., et al. (2022). Secondary polycythemia in men receiving testosterone therapy increases risk of major adverse cardiovascular events and venous thromboembolism in the first year of therapy. Journal of Urology, 207(6), 1295-1301. PubMed
14. Bhasin, S., Woodhouse, L., Casaburi, R., et al. (2001). Testosterone dose-response relationships in healthy young men. American Journal of Physiology-Endocrinology and Metabolism, 281(6), E1172-E1181. PubMed
15. Bhasin, S., Brito, J. P., Cunningham, G. R., et al. (2018). Testosterone therapy in men with hypogonadism: An Endocrine Society clinical practice guideline. Journal of Clinical Endocrinology & Metabolism, 103(5), 1715-1744. PubMed
16. Travison, T. G., Vesper, H. W., Orwoll, E., et al. (2017). Harmonized reference ranges for circulating testosterone levels in men of four cohort studies in the United States and Europe. Journal of Clinical Endocrinology & Metabolism, 102(4), 1161-1173. PubMed
17. Abdullah, R., Bjørnebekk, A., Hauger, L. E., et al. (2024). Severe biventricular cardiomyopathy in both current and former long-term users of anabolic-androgenic steroids. European Journal of Preventive Cardiology, 31(5), 599-608. PubMed
18. Bulut, Y., Rasmussen, J. J., Brandt-Jacobsen, N., et al. (2024). Coronary microvascular dysfunction years after cessation of anabolic androgenic steroid use. JAMA Network Open, 7(12), e2451013. PubMed
19. Greenberg, D. R., Kohn, T. P., Asanad, K., Brannigan, R. E., & Halpern, J. A. (2024). Association of testosterone replacement therapy with atrial fibrillation and acute kidney injury. Journal of Sexual Medicine, 21(12), 1201-1203. PubMed
20. Gagliano-Jucá, T., Pencina, K. M., Shang, Y. V., et al. (2025). Association of testosterone-induced increase in neutrophil and monocyte counts with thromboembolic events: The TRAVERSE trial. American Heart Journal, 288, 77-88. PubMed
21. Grace, F. M., & Davies, B. (2004). Raised concentrations of C reactive protein in anabolic steroid using bodybuilders. British Journal of Sports Medicine, 38(1), 97-98. PubMed
22. U.S. Food and Drug Administration. (2025, February 28). FDA issues class-wide labeling changes for testosterone products. FDA