What are the long‑term side effects of testosterone replacement therapy when dosed to maintain serum testosterone within the normal physiological range?

Long-Term Side Effects of Testosterone Replacement Therapy at Physiological Doses

When testosterone replacement therapy is dosed to maintain serum testosterone within the normal physiological range (approximately 450–600 ng/dL), the long-term side effects are generally limited and manageable with appropriate monitoring, though certain risks—particularly erythrocytosis—require vigilant surveillance.

Cardiovascular Safety

• The most recent high-quality evidence from the 2023 TRAVERSE trial (5,246 men, mean follow-up 21.7 months) demonstrated no significant increase in major adverse cardiovascular events or stroke risk when testosterone therapy maintained physiological levels in men with confirmed hypogonadism and pre-existing or high cardiovascular risk. 1, 2
• Multiple recent meta-analyses and systematic reviews confirm that testosterone therapy at physiological replacement doses shows a neutral or possibly beneficial effect on cardiovascular outcomes, contradicting earlier concerns. 1, 3, 2
• The 2025 European Expert Panel position statement concluded that testosterone therapy is safe from a cardiovascular standpoint when prescribed to appropriately selected patients, with benefits outweighing risks when used responsibly. 2
• Injectable testosterone formulations may carry slightly higher cardiovascular risk than transdermal preparations due to supraphysiologic peak levels alternating with subtherapeutic troughs, though this remains controversial. 4

Hematologic Effects: The Primary Concern

• Erythrocytosis (elevated hematocrit) is the most common and clinically significant long-term side effect, with incidence varying dramatically by formulation:
  • Injectable testosterone: up to 44% of users develop hematocrit >52% 1, 4
  • Transdermal patches: approximately 15.4% incidence 1, 4
  • Transdermal gel (dose-dependent): 2.8% at low dose (5 mg/day) to 17.9% at high dose (100 mg/day) 1
• The most pronounced hematocrit rise occurs within the first 3 months of therapy, with values increasing from subnormal to mid-normal ranges. 1
• Hematocrit >54% is an absolute contraindication to continuing therapy; treatment must be withheld and therapeutic phlebotomy considered in high-risk patients. 1, 5
• Elevated blood viscosity from erythrocytosis can exacerbate coronary, cerebrovascular, and peripheral vascular disease, particularly in older adults. 1

Prostate-Related Effects

• The 2023 TRAVERSE trial confirmed no difference in prostate cancer incidence between testosterone and placebo groups at mean follow-up of 33 months. 1, 6
• Current evidence suggests that testosterone therapy does not increase prostate-specific antigen (PSA) levels or prostate cancer risk when dosed physiologically, though PSA may rise modestly and typically remains within clinically acceptable ranges. 1, 7
• Benign prostatic hyperplasia symptoms rarely worsen with physiological testosterone replacement and may actually improve in some men, except in those with severe baseline symptoms. 1, 6
• For prostate cancer survivors, safety data remain limited, though meta-analyses suggest no increased recurrence risk—however, evidence quality is poor and testosterone should be avoided in advanced disease. 1

Metabolic and Lipid Effects

• Most studies show no adverse change in lipid profiles with physiological replacement doses; testosterone therapy may actually improve triglycerides and HDL cholesterol. 1, 6
• Testosterone therapy can improve insulin resistance, glycemic control, and HbA1c (reduction of approximately 0.37%) in hypogonadal men with type 2 diabetes. 4, 8
• Modest improvements in metabolic syndrome markers including fasting glucose and insulin sensitivity are consistently observed. 5, 8

Fertility and Reproductive Effects

• Testicular atrophy and infertility are common, especially in young men, and are usually reversible with cessation of treatment but may require prolonged recovery (potentially 6–18 months or longer). 1, 6
• Exogenous testosterone suppresses spermatogenesis through feedback inhibition of pituitary LH and FSH, causing azoospermia in most men. 1, 5
• This effect is dose-dependent and duration-dependent; longer treatment courses may result in more prolonged suppression after discontinuation. 9

Fluid Retention and Edema

• Testosterone promotes sodium and water retention, which is rarely of clinical significance in healthy men but may precipitate or worsen congestive heart failure in susceptible patients. 1, 4
• This effect is more pronounced with injectable formulations due to supraphysiologic peak levels. 1

Sleep Apnea

• Obstructive sleep apnea may initially worsen in some men, particularly those with obesity or chronic lung disease, though this is typically not a longstanding effect. 1, 7
• Recent evidence suggests that testosterone therapy does not worsen obstructive sleep apnea in most patients, and combined CPAP plus testosterone gel showed better outcomes than CPAP alone. 1, 6

Dermatologic Effects

• Skin reactions are common with patch formulations (up to 66%) but rare with gel (5%) and injections. 1
• Acne or oily skin occurs infrequently and is typically mild. 1, 6

Gynecomastia

• Gynecomastia is rare and usually reversible with dose adjustment or discontinuation. 1, 6
• This occurs due to peripheral aromatization of testosterone to estradiol, particularly in men with obesity. 5

Hepatotoxicity

• Hepatotoxicity is limited to oral alkylated testosterone agents, which are infrequently used in the United States and not recommended. 1
• Injectable and transdermal formulations at physiological doses do not cause clinically significant liver toxicity. 1

Bone Density Effects (Beneficial)

• Testosterone therapy produces modest improvements in bone mineral density (3.2% increase at lumbar spine, 1.4% at femoral neck), which is beneficial for men with osteopenia or osteoporosis. 8

Quality of Life and Symptom Improvements

• Small but statistically significant improvements in sexual function and libido (standardized mean difference ≈0.35) are consistently observed. 1, 5, 4
• Little to no effect on physical functioning, energy, vitality, depressive symptoms, or cognition even with confirmed hypogonadism. 1, 5, 4
• Modest improvements in vitality and reduced fatigue may occur, though effect sizes are small (standardized mean difference ≈0.17). 6, 4

Critical Monitoring Requirements to Minimize Long-Term Risks

• Hematocrit/hemoglobin at every visit (baseline, 2–3 months, then every 3–6 months first year, then annually); withhold if >54%. 1, 5
• PSA levels in men >40 years (baseline, then every 3–6 months first year, then annually); refer to urology if PSA rises >1.0 ng/mL in first 6 months or >0.4 ng/mL per year thereafter. 1, 5
• Testosterone levels at 2–3 months after initiation or dose change, then every 6–12 months once stable, targeting mid-normal range (450–600 ng/dL). 1, 5
• Digital rectal examination annually in men >40 years to assess for prostate abnormalities. 5
• Lipid profile and fasting glucose/HbA1c at baseline and annually. 5

Formulation-Specific Risk Profiles

• Transdermal testosterone gel (1.62%, ~40 mg daily) is first-line due to stable serum levels and lower erythrocytosis risk (15.4% vs 44% with injectables). 1, 5
• Injectable testosterone (cypionate/enanthate 100–200 mg every 2 weeks) is more cost-effective (~$156/year vs ~$2,135/year for gel) but carries higher erythrocytosis risk. 5, 4
• Weekly dosing of injectables reduces peak-trough fluctuations compared to bi-weekly dosing, potentially lowering adverse event risk. 1

Common Pitfalls to Avoid

• Do not ignore mild erythrocytosis (hematocrit 50–52%) in elderly patients or those with cardiovascular disease, as even modest elevations increase thrombotic risk. 1
• Do not continue full-dose testosterone when hematocrit exceeds 54%—this is an absolute indication to withhold therapy. 1, 5
• Do not initiate testosterone without confirming the patient does not desire fertility, as suppression of spermatogenesis can be prolonged and potentially irreversible. 1, 5
• Nearly 50% of men on testosterone therapy never have their levels re-checked, representing a dangerous practice pattern; regular monitoring is mandatory. 5
• Approximately 20–30% of men receiving testosterone do not meet diagnostic criteria for hypogonadism, highlighting the need for strict adherence to diagnostic protocols (two morning testosterone measurements <300 ng/dL plus specific symptoms). 4

Long-Term Safety Data Limitations

• Long-term safety data beyond 2–3 years are limited, as few randomized trials extend beyond one year of follow-up. 4, 3
• The TRAVERSE trial provides the most robust medium-term safety data (mean 21.7 months), but decade-long prospective studies are still lacking. 1, 2
• No large-scale studies have definitively assessed mortality outcomes, as event rates remain too low for meaningful analysis. 1, 3

Bottom Line for Clinical Practice

When testosterone replacement therapy is appropriately prescribed (confirmed biochemical hypogonadism with two morning testosterone measurements <300 ng/dL plus specific symptoms, primarily sexual dysfunction), dosed to maintain physiological levels (450–600 ng/dL), and monitored according to guideline recommendations, the long-term side effect profile is favorable. 1, 2 The primary risk requiring vigilance is erythrocytosis, which occurs in 15–44% of patients depending on formulation and necessitates regular hematocrit monitoring with treatment interruption if levels exceed 54%. 1 Cardiovascular safety concerns have been largely refuted by recent high-quality evidence, prostate cancer risk is not increased, and most other side effects (testicular atrophy, mild fluid retention, skin reactions) are manageable or reversible. 1, 6, 2 The key to safe long-term use is strict patient selection, appropriate dosing, regular monitoring, and avoidance of therapy in men seeking fertility preservation. 1, 5