Testosterone in women boosts energy, mood, and cognition — a 332-patient study maps the timeline.

The 332-patient study — what it measured and what it found

The study that has reignited clinical interest in testosterone therapy for women was published in the Journal of Personalized Medicine in May 2026.^[1] The design was retrospective and observational — not a randomized controlled trial — but the scale and scope were notable: 332 women aged 27 to 78 (mean 45.7 years), treated through a telehealth-based platform using individualized, biomarker-guided testosterone replacement.

The primary outcome measure was patient-reported symptom burden. The results were directionally strong: 84% of participants reported increased energy, and approximately 90% reported improved overall quality of life. The five tracked biomarkers all shifted in the expected direction: total testosterone rose 151.8%, free testosterone rose 216.7%, SHBG (sex hormone-binding globulin — the protein that binds and inactivates testosterone) fell 13.3%, hemoglobin rose 5.5%, and triglycerides fell 12.6%.

What's worth noting in those numbers: the triglycerides result. Women with elevated androgens are sometimes counseled away from testosterone therapy out of concern for lipid profiles — this study observed the opposite, consistent with broader data on transdermal (non-oral) testosterone showing neutral or favorable lipid effects at physiologic doses. The drop in SHBG is also clinically meaningful: lower SHBG means more free, bioavailable testosterone reaching target tissues.

More interesting than the aggregate numbers was the symptom hierarchy. Energy and fatigue was the top self-reported benefit, cited by 64.2% of participants. Mood improvement followed at 49.7%. Sexual desire came third at 41.3%. That ordering matters — it runs counter to the dominant clinical framing of female testosterone therapy as a treatment for hypoactive sexual desire disorder (HSDD — clinically defined as persistently low or absent sexual desire causing personal distress). Most women in this cohort were seeking help with energy and mood first. Their libido was a secondary, not primary, complaint.

Who benefits most: the subgroup signal

The most reliable evidence base — drawn from the 2019 Lancet meta-analysis by Islam et al., which pooled data from 36 randomized controlled trials^[3] — supports testosterone therapy most strongly for postmenopausal women with HSDD. The meta-analysis found transdermal testosterone significantly improved satisfying sexual episodes, sexual desire, pleasure, and responsiveness compared to placebo. The effect size was consistent across trials.

The Global Consensus Position Statement, simultaneously published in 2019 across eight major medical society journals,^[4] formalized HSDD as the only evidence-based indication. But "evidence-based indication" is a regulatory and guideline concept — it means this is where the RCT data is concentrated, not that benefit is absent elsewhere. The Bikman 2026 cohort and a 2025 UK menopause clinic study of 510 women^[2] suggest the energy, mood, and cognitive improvement signals are real; they simply haven't been the subject of large prospective RCTs because those trials were never designed to measure them.

Beyond postmenopausal women, the evidence is thinner but directionally consistent for:

• Perimenopausal women on hormone replacement therapy (HRT) with persistent symptoms. The UK cohort of 510 women — all on existing HRT — showed significant improvement across all nine cognitive and mood symptom domains despite adequate estrogen and progesterone levels. Adding testosterone resolved symptoms that estrogen alone did not.
• Women with high SHBG. Women on oral contraceptives or oral estrogen formulations often have elevated SHBG — which binds circulating testosterone and reduces bioavailable levels below what a total testosterone measurement reflects. Total testosterone in the normal range does not rule out functional androgen deficiency if SHBG is high.
• Women with chronic stress-driven HPA activation. Cortisol and prolactin both suppress the hypothalamic-pituitary-gonadal (HPG) axis — the hormonal cascade that governs testosterone production. This creates a feedback loop: chronic stress suppresses testosterone, low testosterone reduces stress resilience, which sustains the stress response. Addressing the HPA axis and testosterone together produces more durable results than either alone.

The weakest evidence supports testosterone for bone density protection, cardiovascular benefit, or cognitive dementia prevention as standalone indications. The biological plausibility is there; the RCT data is not. We can reason from mechanism without overstating what trials have confirmed.

The response timeline — energy before cognition

One of the most clinically actionable findings across both the Bikman 2026 study and the UK menopause clinic 2025 cohort^[2] is the temporal mismatch between symptom domains. Energy and mood respond within the first 4–8 weeks of therapy. Cognitive improvements — memory, concentration, mental clarity — lag by a substantial margin, with meaningful gains emerging at 4–6 months.

The ISSWSH clinical practice guideline (International Society for the Study of Women's Sexual Health, 2021) formalizes this in its titration protocol:^[5] first testosterone assessment 3–6 weeks after initiation, with dose adjustment if needed; re-assess at 6 weeks after any dose change; meaningful improvement in sexual motivation at 4–8 weeks; if no meaningful response by 6 months, discontinue and reassess. That 6-month window is the key number for cognitive outcomes — women who discontinue at 8 weeks because their brain fog hasn't shifted are stopping before the cognitive signal has had time to develop.

The underlying mechanism for this delay likely reflects different tissue sensitivities and receptor expression patterns. Testosterone acts on androgen receptors (ARs) expressed throughout the brain — the hippocampus, prefrontal cortex, and limbic system all have significant AR density. But receptor upregulation and neuroplastic changes occur on a slower timescale than the mood and energy effects, which are driven in part by testosterone's conversion to estradiol via aromatase activity in peripheral and central tissues.

For women whose primary complaint is brain fog or cognitive decline rather than low energy or mood, the clinical implication is direct: a 6-month committed trial is the minimum meaningful assessment window. Evaluating cognitive response at 4–8 weeks is premature and will produce apparent non-responders who would have responded at month five.

Cortisol and prolactin: the hidden signal suppressors

Testosterone therapy does not exist in an isolated hormonal vacuum. Two axes upstream of testosterone production have outsized influence on both endogenous production and response to exogenous administration: the HPA (hypothalamic-pituitary-adrenal) axis, which governs cortisol, and prolactin regulation at the pituitary level.

Wdowiak et al. (2020) documented these interactions in a cohort of infertile and fertile cycling women, tracking cortisol, prolactin, LH, FSH, and sex hormones across three cycle points.^[9] The researchers found that elevated cortisol correlates with reduced LH surges and suppressed estradiol — a pattern consistent with published background evidence that chronically high cortisol induces pituitary insensitivity to gonadotropin-releasing hormone (GnRH — the hypothalamic signal that initiates testosterone production). This suppresses luteinizing hormone (LH) and follicle-stimulating hormone (FSH) output, which in turn suppresses both estrogen and testosterone production. Chronic cortisol elevation doesn't just raise stress hormones — it down-regulates the hormonal machinery that produces testosterone.

Prolactin works by a related but distinct mechanism. Hyperprolactinemia — elevated prolactin, whether from a prolactinoma, medications, or sustained stress — suppresses the pulsatile GnRH secretion that drives HPG axis function. It is one of the primary functional causes of low endogenous testosterone in women under chronic stress. The mechanistic irony: the stress response that depletes testosterone also elevates prolactin, compounding the suppression from both directions simultaneously.

There is also an upstream SHBG effect: chronic cortisol elevation raises SHBG production in the liver. Higher SHBG means more of whatever testosterone is present gets bound and inactivated, reducing free (bioavailable) levels further. A woman under sustained chronic stress may show normal total testosterone on a standard lab panel while her free testosterone — the fraction that acts on receptors — is functionally low.

The clinical implication is layered. Women with high cortisol or elevated prolactin will likely show attenuated response to testosterone therapy until the HPA axis is addressed. Testosterone may partially restore HPG axis signaling, but the effect is partial and the human data on this cross-axis interaction remains limited. Addressing chronic stress, sleep, and if appropriate prolactin levels, creates the biological substrate in which testosterone therapy works most reliably.

Cycle phase and the testosterone–cortisol ratio

For premenopausal women, the interaction between testosterone and cortisol is not static — it varies systematically across the menstrual cycle in a pattern that has direct relevance to symptom severity and treatment response.

Cook, Fourie, and Crewther (2021) measured testosterone and cortisol responses to both physical and psychological stressors across three menstrual cycle points in 30 athletic women.^[8] The finding: the periovulatory phase (around day 14) produced the highest testosterone stress response (+13.7%) and the lowest cortisol reactivity. The mid-luteal phase (day 21) showed the opposite pattern — attenuated testosterone response (+7.0%) with higher cortisol reactivity (+12.0%). The early follicular phase (day 7) sat in between.

What this means practically: the testosterone-to-cortisol (T:C) ratio — a compound measure of anabolic versus catabolic signal balance — is most favorable periovulatorily. Women in the luteal phase, particularly premenopausal women with luteal-phase mood symptoms (historically labeled premenstrual dysphoric disorder, or PMDD), experience the worst T:C ratio conditions of their cycle.

A separate meta-analysis by Hamidovic et al. (2020) pooling 35 studies found that circulating cortisol is measurably higher in the follicular phase than in the luteal phase — with the luteal phase producing a cortisol-buffering effect via progesterone.^[10] Perimenopausal women who are losing progesterone lose this buffering, which partially explains why perimenopause amplifies stress sensitivity at the same time that testosterone begins to decline. The two hormonal losses compound each other.

For clinicians titrating testosterone in cycling women: symptoms are likely to be worst in the mid-to-late luteal phase not solely because of progesterone withdrawal but because of unfavorable T:C ratio conditions. Assessing symptom response at one fixed calendar point each cycle may miss this variation and lead to inaccurate conclusions about therapy efficacy.

Delivery, dosing, and the off-label reality

No testosterone formulation is currently approved for women in most jurisdictions, including the United States and Canada. This is not a reflection of the evidence quality — the Lancet meta-analysis and the Global Consensus represent substantial evidence bases — but of regulatory priority. The Global Consensus explicitly calls on regulators to address this as an unmet clinical need.^[4]

In practice, clinicians use male formulations (testosterone gels, creams, or injectable cypionate) at approximately 1/10th to 1/20th the male dose, or compounded formulations designed for female dosing parameters. The standard starting dose for transdermal therapy is 5 mg/day, with titration to 10 mg/day based on symptom response and measured free testosterone levels. The target range is within the premenopausal physiological range — not above it.

Pellet implants deserve specific mention: a 2025 review in Cureus examined their pharmacokinetic profile^[6] and identified a consistent problem — pellets produce supraphysiologic peak levels of 100–250 ng/dL early post-insertion, well above the premenopausal physiological ceiling recommended by consensus guidelines. There is high interindividual variability in absorption, and no reliable dose-adjustment mechanism once a pellet is inserted. Current guidelines from the major medical societies explicitly do not recommend pellets on pharmacokinetic grounds.

A critical and underappreciated issue is measurement accuracy. Lara et al. (2024) note that there is no established cutoff for androgen deficiency in women and that testosterone measurement is not indicated for diagnosis — reflecting the lack of analytic standardization for female ranges.^[7] A broader analytical limitation compounds this: standard immunoassay platforms are calibrated primarily for male testosterone concentrations, which run 10–20 times higher than female levels. Published accuracy evaluations show immunoassay bias of 30–89% at female-range concentrations versus LC-MS/MS reference measurements. Liquid chromatography–mass spectrometry (LC-MS/MS) is the analytically superior approach for female testosterone measurement but is rarely available in routine clinical settings. This matters: a normal-reading immunoassay result in a symptomatic woman does not rule out functional androgen deficiency. The measurement tool itself may be inadequate.

Safety: what the RCT data shows

The Islam et al. Lancet meta-analysis is the most complete and rigorous safety dataset available.^[3] At physiologic transdermal doses, the confirmed androgenic side effects are acne and increased body or facial hair — statistically significant versus placebo across multiple trials, categorized as non-serious adverse events. Voice changes and clitoromegaly are documented primarily at supraphysiologic doses, not at guideline-recommended physiologic doses.

Cardiovascular effects are formulation-dependent. Oral testosterone formulations — essentially never used in current practice for women — reduce HDL cholesterol. Transdermal formulations show neutral lipid profiles. The Bikman 2026 cohort, using primarily transdermal delivery, saw triglycerides fall 12.6% — directionally consistent with the RCT data on metabolic neutrality of transdermal administration.

Breast cancer remains an evidence gap. Current data from observational studies does not show increased breast cancer risk at physiologic doses, but long-term RCT data beyond 24 months is absent. Every major guideline acknowledges this as a limitation. For women with hormone-sensitive breast cancer history, this is a conversation with a clinician, not a self-directed decision.

The overarching safety picture at physiologic transdermal doses is: minor androgenic side effects in some women, neutral cardiovascular and metabolic profile, and an unresolved (but currently unfavorable) long-term data picture on cancer. The signal for energy, mood, and sexual function benefit at those doses is real. The risk-benefit calculation is individual, and it requires a clinician who understands female hormonal physiology — not one defaulting to "testosterone is a male hormone and isn't appropriate for women."

For context on what drives testosterone decline in both sexes from the population level, see the testosterone decline article covering the four major longitudinal cohorts. For women considering testosterone as part of a broader hormone protocol, the 2026 FDA libido indication update is relevant background on where the regulatory landscape stands. On the question of combining testosterone with other hormonal interventions to preserve HPG axis function, the enclomiphene and HCG article covers the related mechanisms in men — the HPG axis architecture is analogous.

References

1. Bikman BJ, Reynolds PR, et al. Testosterone Replacement Therapy in Women Is Associated with Improved Symptom Burden and Favorable Biomarker Changes: A Retrospective Observational Study. J Pers Med. 2026;16(5):231. doi:10.3390/jpm16050231
2. Glynne S, Kamal A, Kamel AM, Reisel D, Newson L. Effect of transdermal testosterone therapy on mood and cognitive symptoms in peri- and postmenopausal women: a pilot study. Arch Womens Ment Health. 2025. PMC12092509. doi:10.1007/s00737-024-01513-6
3. Islam RM, Bell RJ, Green S, Page MJ, Davis SR. Safety and efficacy of testosterone for women: a systematic review and meta-analysis of randomised controlled trial data. Lancet Diabetes Endocrinol. 2019;7(10):754–766. PMID 31353194. doi:10.1016/S2213-8587(19)30189-5
4. Davis SR, Baber R, Panay N, et al. Global Consensus Position Statement on the Use of Testosterone Therapy for Women. J Clin Endocrinol Metab. 2019;104(10):4660–4666. PMID 31488288. doi:10.1210/clinem/dgz111
5. Parish SJ, Simon JA, Davis SR, et al. International Society for the Study of Women's Sexual Health Clinical Practice Guideline for the Use of Systemic Testosterone for Hypoactive Sexual Desire Disorder in Women. J Sex Med. 2021;18(5):849–867. PMID 33792440. PMC8064950
6. Viana DPC, Jacobsen L, Gabriel LH, et al. Testosterone Pellets in Women: Revisiting Safety and Clinical Outcomes. Cureus. 2025. PMID 41089583. PMC12516641
7. Lara LAS, et al. Challenges of prescribing testosterone for sexual dysfunction in women. Rev Bras Ginecol Obstet. 2024. PMID 39176198. PMC11341187
8. Cook CJ, Fourie P, Crewther BT. Menstrual variation in the acute testosterone and cortisol response to laboratory stressors correlate with baseline testosterone fluctuations at a within- and between-person level. Stress. 2021;24(4):458–467. PMID 33287617. doi:10.1080/10253890.2020.1860937
9. Wdowiak A, et al. Interactions of Cortisol and Prolactin with Other Selected Menstrual Cycle Hormones Affecting the Chances of Conception in Infertile Women. Int J Environ Res Public Health. 2020;17(20):7537. PMID 33081268. PMC7588978
10. Hamidovic A, et al. Higher Circulating Cortisol in the Follicular vs. Luteal Phase of the Menstrual Cycle: A Meta-Analysis. Front Endocrinol. 2020;11:311. PMC7280552. doi:10.3389/fendo.2020.00311