Testosterone

Limits on what this shows

• Null/negative findings: No human RCT measuring mitochondrial structure or function directly in men on testosterone therapy was identified in this review.
• At-risk subgroups: None identified; the available data lean protective/anabolic rather than depleting.

Regulatory status

Not addressed in FDA labeling.

Population evidence

Usui et al. 2014, mouse model (FEBS Lett, PMID 24726723): testosterone administration associated with elevated skeletal-muscle mitochondrial biogenesis alongside testosterone-induced body-weight loss. Broader mechanistic review of androgen receptor effects on mitochondrial biogenesis, fusion/fission, mitophagy, and bioenergetics (Am J Physiol Cell Physiol 2022, PMID 35704694), which states the precise molecular mechanisms are not yet established despite the described associations.

Mechanism

Named but incompletely characterized: androgen receptor signaling implicated in mitochondrial biogenesis and quality-control pathways.

Counter-evidence held on record

The evidence base is animal/cell only for this question; this is a Tier B ceiling regardless of direction. Animal and cell data are not a basis for human mitochondrial-support claims; this entry is not evidence of harm nor of confirmed human-level benefit.

Limits on what this shows

• Null/negative findings: Single trial (n=38); no independent replication identified in this review.
• At-risk subgroups: None identified; this is a protective-direction signal in older hypogonadal men.

Regulatory status

Not addressed in FDA labeling.

Population evidence

Christensen et al. 2024, double-blind RCT, testosterone gel vs placebo x24 weeks, n=38 men aged 60–78 (Andrology, PMID 37177884, DOI 10.1111/andr.13458): whole-body oxidative stress (urinary nucleic-acid oxidation markers) decreased with testosterone versus placebo; the treatment group had greater lean-mass gain and fat loss, and the oxidative-stress change was reported as mediated by the change in body composition.

Mechanism

Not a direct antioxidant mechanism; proposed as indirect, via reduced fat mass/adiposity-linked oxidative burden.

Counter-evidence held on record

The mechanism is indirect (mediated by body composition, not a direct redox effect of testosterone itself), so this should not be read as evidence that testosterone is a direct antioxidant. Reverse-direction and mechanistically indirect; must not be marketed as testosterone "boosting antioxidant status" beyond what this body-composition-mediated evidence supports, and does not license an oxidative-stress-based supplement recommendation.

Limits on what this shows

• Null/negative findings: In the iron-deficiency-anemia subgroup (already iron-depleted at baseline), Artz 2020 found no significant hepcidin or ferritin change. Transferrin itself did not change significantly in either subgroup.
• At-risk subgroups: Men with borderline iron stores at baseline; higher or injectable dosing; older age (effect more pronounced).

Regulatory status

Not addressed as a distinct iron-status risk in FDA labeling; labeling addresses only the downstream hematocrit effect.

Population evidence

Artz et al. 2020 (JCEM, PMID 32785689, n=95, RCT substudy of the Testosterone Trials): in the unexplained-anemia subgroup, hepcidin fell −8.2 ng/mL (95% CI −13.7 to −2.7, P=0.004) and ferritin fell −19.6 µg/L (95% CI −32.8 to −6.3, P=0.004) over the first 3 months; hepcidin decline correlated with hemoglobin rise (r=−0.35, P=0.01). Bachman et al. 2014 (PMID 24158761) found the same ferritin/hepcidin decline pattern at 1 and 3 months. Bachman et al. 2010 (PMID 20660052, n=109) showed dose-dependent hepcidin suppression within 1 week of testosterone enanthate dosing.

Mechanism

Named: testosterone suppresses hepatic hepcidin transcription, increasing ferroportin-mediated iron export from stores and intestinal absorption, mobilizing ferritin-bound iron toward erythropoiesis.

Counter-evidence held on record

Effect size is modest (ferritin change largely within normal range in most participants) and partially reverses by 6 months despite continued dosing (Bachman 2014), arguing against a progressive, unbounded depletion. This reflects iron mobilization/utilization, not a dietary deficiency. Iron supplementation should not be started empirically on the basis of this mechanism alone. Ferritin/iron studies are reasonable baseline and follow-up labs when anemia or unexplained fatigue is present, but treatment decisions belong to the prescriber based on confirmed iron status.

Limits on what this shows

• Why no grade: Searched specifically against testosterone/testosterone-therapy across PubMed (via cached search-result parsing, since PubMed's abstract pages render via JavaScript and returned only page metadata to automated fetch), multiple WebSearch phrasings ("testosterone copper serum," "testosterone replacement therapy copper," "anabolic steroid copper deficiency ceruloplasmin"), and an attempted Europe PMC REST API query that returned HTTP 429 during this review. No drug-specific evidence at any tier was found, and no circulating consumer claim pairing testosterone therapy itself with copper depletion was found. (The only copper-adjacent consumer content located concerns zinc supplementation causing copper depletion, a supplement-supplement interaction, not a testosterone-copper claim; it is noted in the Monitoring panel, not graded here.) Both dual conditions for UNGRADED are met: no Tier B/C evidence, and no circulating claim to rebut as white.

Tooling note

Europe PMC REST API access was rate-limited partway through this review. A later revision should re-query Europe PMC and Semantic Scholar directly for "testosterone AND copper" before treating this as a durable conclusion rather than a possible access gap.

Limits on what this shows

• Null/negative findings: No significant alpha-diversity change; no species-level signal tied to treatment duration in the mSphere pilot.
• At-risk subgroups: None clearly identified from this small pilot.

Regulatory status

Not addressed in FDA labeling.

Population evidence

Harris et al. 2024, prospective pilot, n=9 transgender individuals (ages 14–20) starting testosterone, mostly weekly subcutaneous injections, 26 stool samples over up to 8 months (mSphere, DOI 10.1128/msphere.00557-24): alpha diversity trended down but was not statistically significant; beta diversity changed significantly (R²=0.035, P=0.009) though each individual's microbiome remained most similar to their own baseline; no species reached significance for association with treatment duration; genes for glutamate generation increased and glutamate-consuming genes decreased. Separately, Mohr et al. 2026, human trial in a severe energy-deficit (military) context (Am J Physiol Endocrinol Metab, DOI 10.1152/ajpendo.00291.2025): testosterone partially preserved short-chain-fatty-acid-producing bacteria and microbial energy-metabolism pathways during caloric restriction, a protective-direction signal in that specific stress context.

Mechanism

Proposed: testosterone may increase host intestinal glutamate absorption, altering substrate availability for glutamate-metabolizing gut microbes. Precise mechanism not established.

Counter-evidence held on record

n=9 is too small to generalize; the Mohr 2026 finding runs in a protective direction in a specific energy-deficit context, not a general depletion direction. This is preliminary, small-sample human evidence with mixed direction; it does not support a microbiome-depletion or probiotic-need claim for testosterone therapy generally.

Limits on what this shows

• Null/negative findings: In the iron-deficiency-anemia subgroup of Artz 2020, the hemoglobin response was smaller and not statistically significant (+0.38 g/dL, 95% CI −0.19 to 0.95, P=0.19), and hepcidin/ferritin did not change significantly — the erythropoietic drive is blunted when iron stores are already depleted.
• At-risk subgroups: Older men; higher/injectable doses producing supraphysiologic peaks; obstructive sleep apnea; smokers; baseline hematocrit near the upper-normal range.

Regulatory status

Current FDA labeling for systemic testosterone products requires periodic hemoglobin/hematocrit monitoring for polycythemia; a hematocrit above approximately 54% is a labeled reason to interrupt or reduce dose.

Population evidence

Bachman et al. 2014, RCT, testosterone gel vs placebo x6 months, older men with mobility limitation: 7–10% rise in hemoglobin/hematocrit; erythropoietin up, ferritin and hepcidin down at 1 and 3 months (PMID 24158761). Bachman et al. 2010, RCT with GnRH-agonist suppression plus graded testosterone enanthate doses (25–600 mg/week) x20 weeks, n=109 (53 younger, 56 older men): dose-dependent hepcidin suppression within 1 week, more pronounced in older men, predictive of later hematocrit rise (PMID 20660052). Artz et al. 2020, RCT substudy of the Testosterone Trials, n=95 hypogonadal men ≥65y with anemia: hemoglobin +0.58 g/dL (95% CI 0.31–0.85, P<0.001) in the unexplained-anemia subgroup (PMID 32785689).

Mechanism

Named: testosterone stimulates renal/marrow erythropoietin production and suppresses hepatic hepcidin transcription, increasing iron availability for erythropoiesis and expanding red-cell mass.

Counter-evidence held on record

Bachman 2014 found erythropoietin and hepcidin returned toward baseline by 6 months despite continued dosing, consistent with a new physiological set-point rather than unbounded escalation; this bounds but does not negate the finding. This is an excess/retention finding, not a deficiency. No supplement or dietary intervention is indicated to "support" oxygen delivery on this drug. The clinically appropriate response to a rising hematocrit is dose adjustment, interruption, or therapeutic phlebotomy under a prescriber, never iron or erythropoiesis-supporting supplementation.

Limits on what this shows

• Null/negative findings: None identified; this reflects an evidence gap rather than a negative trial.
• At-risk subgroups: None identified.

Regulatory status

Not addressed in FDA labeling.

Population evidence

No direct ATP-flux study identified. Adjacent evidence: the same AJP Cell Physiol 2022 review (PMID 35704694) covering androgen effects on mitochondrial bioenergetics generally, and the Usui 2014 mouse biogenesis study (PMID 24726723). A 1996 human study (Brodsky, Balagopal, Nair; PMID 8855787) found testosterone replacement increased muscle protein synthesis in hypogonadal men, an anabolic finding adjacent to, but distinct from, ATP flux.

Mechanism

Not established at the human level for this specific endpoint.

Counter-evidence held on record

None specific; the entry is limited by absence of direct data, disclosed here rather than overstated. Do not present this as evidence testosterone "boosts ATP" or cellular energy production; human-level evidence for this specific endpoint does not exist.