Libido. Muscle. Energy.
Different claims.
Different evidence.

Testosterone has a demonstrated, modest average benefit for distressing low sexual desire in selected postmenopausal women. That finding does not establish a general treatment for fatigue, brain fog or difficulty building muscle. Human experiments show that raising androgen exposure can change lean mass and some performance measures, but those experiments often ask a different question from restoring usual female physiological exposure. The useful answer depends on the outcome, the population and the concentration achieved. ^[1][2][4][5]

Start by separating a hormone's normal biological importance from the effects of adding more of it. A substance can participate in muscle and brain physiology without supplementation improving every function of those tissues. Receptor activation is a mechanism. Relief of a symptom, a change on a scan and better performance on a task are clinical outcomes that require separate measurements.

The androgen receptor is a hormone-responsive regulator of gene transcription. Its effects depend on the tissue and cellular context, and androgen signaling also includes faster cellular responses. Testosterone can additionally be converted to dihydrotestosterone or estradiol. This makes a simple 'more testosterone equals more of every benefit' model unreliable: one circulating molecule participates in several pathways, and a blood concentration does not describe all of them. ^[6][13]

An instructive human-and-cell study illustrates the distinction. Alexander and colleagues found associations between bioavailable testosterone and training-related hypertrophy in premenopausal women, while total testosterone was not related to muscle mass or strength. Cultured muscle cells exposed to supraphysiological testosterone enlarged and showed androgen-receptor nuclear localization without the proposed mTOR signaling response. The human component was not a randomized testosterone treatment experiment; the cell component cannot establish clinical efficacy. ^[12]

For readers comparing claims online, the first question should therefore be concrete: what was actually improved? A libido questionnaire cannot establish muscle growth. Lean mass cannot establish better memory. A more favorable subjective experience may be important, but it should not be retroactively relabeled as proof of every mechanism proposed for the hormone. That discipline allows genuine benefits to remain visible without expanding them beyond the study.

Total testosterone combines hormone circulating in different binding states. Much is protein-bound, including to sex hormone-binding globulin, or SHBG, and albumin. Changes in binding proteins can alter how total and free measurements relate. Tissue uptake, metabolism and receptor context add further complexity. This is why two identical total concentrations need not summarize the same endocrine situation, and why a calculated free value is not a direct measurement of what every tissue experiences. ^[14]

Measurement is especially challenging at the relatively low concentrations typical of women. The Global Consensus statement favors mass-spectrometry methods for accurate total-testosterone measurement and cautions about direct assays in the female range. It also states that no circulating androgen cutoff distinguishes women with and without sexual dysfunction. A laboratory result can inform an assessment; it cannot independently diagnose hypoactive sexual desire disorder, or HSDD. ^[6]

HSDD concerns low desire associated with distress, assessed in its broader context. The ISSWSH guideline emphasizes a biopsychosocial assessment, including contributors such as relationship difficulties, mental-health conditions and medications. Its discussion of SHBG does not support raising testosterone beyond the physiological range simply to overcome a high SHBG result. These are clinical reasoning boundaries, not a universal target or a self-treatment formula. ^[7]

Consider how a misleading claim can develop. A person reports fatigue, a test returns a value toward the lower end of a reference interval, and a treatment is followed by feeling better. That sequence may be sincerely reported. On its own, it cannot distinguish a hormone-specific effect from concurrent changes, expectation, natural fluctuation or another cause of the symptom. Controlled studies help answer that causal question; dismissing the experience and treating it as proof are both avoidable mistakes.

Reference intervals also serve a different purpose from outcome thresholds. A range describes a reference population using a particular method. It does not identify a concentration at which every person's desire, strength or energy becomes optimal. Importing a male threshold would add another unsupported step. In this article, reported concentrations describe experimental exposure, not a desirable level for the reader.

The APHRODITE trial randomized 814 postmenopausal women with HSDD who were not taking estrogen to placebo or patches delivering 150 or 300 micrograms daily. These are study exposures, not instructions. At 24 weeks, satisfying sexual episodes per four weeks increased by 2.1 in the 300-microgram group and 0.7 with placebo. The 150-microgram group's increase was 1.2 and did not significantly beat placebo on that primary endpoint. Desire and distress measures improved with both active exposures. ^[1]

The broader 2019 synthesis included 36 randomized trials and 8,480 participants. For satisfactory sexual-event frequency in postmenopausal women, the pooled difference was 0.85 events per month, with a 95% confidence interval from 0.52 to 1.18. This is an average additional benefit over the comparator, not the total number of events and not a promise for an individual. The review also found improvement across several sexual-function measures. ^[5]

An average of roughly one additional event monthly can matter to someone experiencing distress, particularly if desire and comfort with sexual life also improve. It would nevertheless be misleading to describe that result as a dramatic, universal restoration of libido. Event counts capture only part of sexual wellbeing. The person's concern, degree of distress and preferred outcome matter alongside the scale used in the trial.

The evidence population is equally important. The 2019 consensus identifies postmenopausal HSDD as the established indication; ISSWSH discusses limited evidence in late reproductive-age women as well. The 2026 VA/DoD menopause guideline's indexed evidence discussion continues to describe modest benefits and important gaps, rather than endorsing testosterone for every menopausal symptom. Guidance should be read as an interpretation of evidence, not as another randomized experiment. ^[6][7][11]

Finally, pooled reviews and individual trials are not independent votes that can simply be added together. APHRODITE contributes to the wider evidence base. Counting its participants again as though the meta-analysis were a new cohort exaggerates the amount of independent information. The value of synthesis is that it examines consistency, magnitude and uncertainty across studies, while retaining differences in participants and formulations.

Huang and colleagues randomized 71 hysterectomized postmenopausal women, with or without ovary removal, after an estradiol run-in. The 24-week experiment compared placebo with weekly testosterone-enanthate injections of 3, 6.25, 12.5 or 25 mg. Sexual function was primary; body composition and performance were secondary. Sixty-two women had analyzable primary-outcome data. Mean on-treatment total testosterone reached 210 ng/dL in the highest group. Significant placebo-separated lean-mass and selected power gains occurred there, not in the lower groups. ^[2]

This does not establish that moving a low result into an ordinary female physiological range will produce the same gains. Nor is it a clean test of testosterone without estrogen co-treatment. The exposure and clinical history are integral parts of the experiment. A concentration-response relationship within this study cannot supply a general dosing algorithm, predict an individual response or establish the balance of benefits and harms over years.

A separate trial randomized 48 healthy, active women aged 18 to 35 to testosterone cream, 10 mg applied daily, or placebo for ten weeks. Mean testosterone rose from 0.9 to 4.3 nmol/L in the active group. Running time to exhaustion increased by 21.17 seconds within that group, while the between-group difference was 15.5 seconds (P=0.045). Lean-mass changes were 923 versus 135 grams. Jump and knee-extension measures did not significantly improve; the Wingate comparison was also nonsignificant. ^[4]

The applied cream amount must not be confused with absorbed testosterone or with the delivery rate of a patch. Similarly, the running result is not a percentage improvement in race speed. Time to exhaustion under a laboratory protocol measures one aspect of performance. It does not guarantee an equivalent change in competition results, maximal strength or long-term training adaptation. This is a positive controlled finding with a specific meaning.

Lean mass adds another interpretive layer. A body-composition estimate is not a direct count of new contractile muscle protein, and strength or power can change differently from mass. The most useful reading preserves each measured outcome instead of replacing them with a single 'muscle-building' label. The physiological-exposure literature has not established a general musculoskeletal benefit, and its smaller outcome samples leave uncertainty. Positive higher-exposure experiments do not erase that distinction. ^[5][6]

The cognitive report from the Huang parent trial included 46 women with both baseline and end-of-treatment data. Across 24 weeks and a wide exposure range, the investigators did not find significant improvements versus placebo in spatial ability, verbal fluency, verbal memory or executive function. This is a secondary report from the same trial discussed above, not an independent replication. Its smaller analytic sample also limits what a null result can exclude. ^[3]

A 2026 systematic review adds a timely check. Faulkner and colleagues identified five eligible studies; three, totaling 244 postmenopausal participants, entered the meta-analysis. Pooled effects across the evaluated cognitive domains were not statistically significant and were mostly close to zero. No included study addressed subjective cognition or perimenopause. The authors explicitly frame the evidence as insufficient to establish an effect, rather than definitive proof that no effect is possible. ^[8]

That distinction is particularly relevant to brain fog. A person can experience substantial difficulty concentrating even when a short standardized task captures little change. Conversely, an improved test score does not necessarily mean daily functioning has improved. A convincing study should decide in advance whether it is testing subjective complaints, objective performance or both, then use measures suited to that question.

General energy is broader still. It can refer to wakefulness, motivation, exercise tolerance, recovery or relief from distress. These are not interchangeable. The consensus literature does not establish an overall wellbeing benefit, and the positive libido studies cannot fill that gap by implication. Someone feeling more engaged after an improvement in sexual wellbeing has not demonstrated a direct energy-producing action of testosterone. ^[6]

A future trial could make these claims more testable by enrolling participants with a defined complaint, accounting for sleep and other concurrent changes, and selecting a meaningful primary endpoint before treatment starts. Adequate follow-up and transparent reporting of unchanged outcomes would be essential. Until then, 'may influence brain biology' and 'improves brain fog' should remain separate statements. This is an evidence gap worth investigating, not a reason to dismiss symptoms.

In APHRODITE, androgenic adverse events, mainly unwanted hair growth, occurred in 30.0% of the higher-patch group versus 23.1% with placebo. Four breast cancers were diagnosed in testosterone recipients and none in placebo recipients; timing included one early diagnosis and one woman with symptoms before randomization. Those small numbers do not establish causation, but they also do not resolve long-term breast safety. ^[1]

Across trials, acne and increased hair growth are clearer signals than rare long-term outcomes. The 2019 meta-analysis found unfavorable lipid changes with oral testosterone that were not seen with nonoral administration. Absence of a serious-harm signal in relatively short studies is not proof of safety in every population, particularly when higher-risk patients are underrepresented. Formulation and exposure matter to risk as well as benefit. ^[5]

Voice deserves specific attention. A 2026 retrospective report compared 30 testosterone-exposed women presenting to a voice clinic with 20 controls. Acoustic and aerodynamic measures differed; most patients had used pellets or injections. Because the exposed group was selected for voice complaints, the study cannot tell us how often voice problems occur among all users, or prove one route is riskier. It provides a clinically relevant signal, not an incidence estimate. ^[9]

ISSWSH identifies acne, excess hair, scalp hair loss and voice deepening as concerns with supraphysiological exposure. Pregnancy or potential pregnancy changes the risk context because androgen exposure can masculinize a female fetus. Postmenopausal trial results should not be generalized to pregnancy, fertility goals or younger women without considering those differences. A short study with few androgenic events is particularly poorly suited to rule out uncommon or persistent changes. ^[7]

As checked November 4, 2026, FDA testosterone information describes products approved for male indications, not a female HSDD indication. The FDA's September 2026 workshop on menopausal testosterone explicitly highlighted unresolved efficacy, measurement and long-term safety questions. A workshop or off-label clinical use is not equivalent to regulatory approval. Male cardiovascular trial findings also cannot establish long-term safety in women. ^[10][15]

The conclusion is specific: low-desire treatment has supportive human evidence in selected women; muscle and performance effects depend on substantially different experimental settings; cognition and general-energy claims remain unestablished. Reading the actual outcome, exposure and population is more informative than deciding whether testosterone is simply 'good' or 'bad' for women. This article was prepared with AI assistance and source checking, without independent clinician review. It is evidence education, not individualized treatment advice.