Low-Dose Testosterone and Cognitive Function in Women: What the Evidence Actually Shows

Why Women Are Asking About Testosterone and the Brain

Brain fog is one of the most reported and least treated symptoms of perimenopause and menopause. In the Study of Women's Health Across the Nation (SWAN), over 60% of perimenopausal women reported new-onset memory or concentration difficulties, yet most clinical visits address hot flushes long before cognition. Testosterone, long dismissed as a "male hormone," is now understood to act directly on the female brain through androgen receptors expressed throughout the hippocampus, prefrontal cortex, and amygdala. That basic neuroscience has driven a sharp rise in off-label prescribing for women who feel like they have lost mental sharpness they cannot get back.

This article examines what the clinical trials actually show, where the evidence is solid, where it is extrapolated, and how to think about low-dose testosterone for cognition across different life stages.

Testosterone Is Not a "Male" Hormone

In premenopausal women, serum testosterone concentrations range from roughly 15 to 70 ng/dL, and free testosterone accounts for about 1-2% of that total. Testosterone peaks in the mid-to-late 20s and declines by approximately 50% between the ages of 25 and 45, before menopause even begins. After natural menopause, ovarian testosterone production falls further, though the adrenal contribution persists. After bilateral oophorectomy, the drop is abrupt: ovarian androgen output stops essentially overnight. This distinction between natural and surgical menopause matters enormously for the cognition data, as you will see below.

How Testosterone Acts on the Female Brain

Testosterone crosses the blood-brain barrier and acts through two routes. It binds androgen receptors directly, and it aromatizes locally to estradiol, which then acts on estrogen receptors in the same hippocampal and cortical circuits that govern verbal memory, working memory, and processing speed. Animal models show that testosterone replacement in ovariectomized rodents improves spatial memory and reduces amyloid-beta burden, though rodent-to-human translation in neuroscience is notoriously unreliable. The dual-pathway mechanism does create a plausible biological rationale, but plausibility is not efficacy.

The 2019 Global Consensus: What It Said About Cognition

The Global Consensus Position Statement on the Use of Testosterone Therapy for Women (2019), endorsed by 11 international societies including The Menopause Society (formerly NAMS), is the closest thing to an authoritative guideline that exists. It concluded:

• Testosterone therapy improves sexual function in postmenopausal women. That conclusion is supported by Level I evidence from multiple randomized controlled trials.
• The evidence for cognitive benefit was reviewed and found to be insufficient to make a recommendation for or against testosterone for cognition as a primary indication.
• The Consensus authors explicitly noted that "there is a signal, particularly in surgically menopausal women, that warrants further study" but cautioned against extrapolating that signal to the general postmenopausal or perimenopausal population.

The Consensus document also established the reference range for "physiological" female testosterone as the premenopausal range, and specified that doses should keep free testosterone within that range. A total testosterone above 150 ng/dL (by liquid chromatography-mass spectrometry, or LC-MS/MS) is considered supraphysiologic and associated with increased androgenic side-effect risk.

What the RCTs on Cognition Show

The most direct evidence comes from three lines of work.

The Sherwin trials (1988-2003). Barbara Sherwin's landmark studies at McGill examined women who underwent surgical menopause and received either estrogen alone, testosterone plus estrogen, or placebo. Women in the testosterone-plus-estrogen arm showed better performance on verbal memory tasks at 3 months compared with placebo, a finding she replicated in subsequent work. These were small trials (n = 20-53), but they were methodologically careful and the effect sizes were clinically meaningful, not just statistically significant.

The APHRODITE trial (Davis et al., 2008). This 52-week RCT of a 300 mcg/day testosterone patch in naturally postmenopausal women (n = 814) was powered for sexual endpoints, not cognition. A secondary cognitive battery showed no significant improvement in the testosterone arm versus placebo. Davis SR et al., NEJM 2008. The trial was not designed to detect cognitive effects, and the null result should be interpreted with that limitation in mind.

The TEAM trial (Wahlin-Jacobsen et al., 2019). A Danish RCT of testosterone gel (300 mcg/day) in pre- and postmenopausal women with HSDD found no significant cognitive benefit at 12 weeks on a composite battery, though premenopausal subgroup data were underpowered to draw any conclusion. Wahlin-Jacobsen S et al., J Sex Med 2019.

Taken together: the strongest positive signal remains in surgically menopausal women (Sherwin), the most adequately powered general postmenopausal trial was null for cognition (APHRODITE), and perimenopause and premenopause are essentially unstudied.

Life-Stage Breakdown: Who Might Benefit, and Who Probably Won't

Cognition is not a single thing. Verbal memory, working memory, processing speed, and executive function are distinct domains that may respond differently to hormonal interventions. Here is how the evidence maps onto life stage.

Reproductive Years (Ages 18-40)

There is no RCT evidence supporting low-dose testosterone for cognitive complaints in premenopausal women. PCOS is the major androgen-related condition in this group, and women with PCOS frequently report fatigue and brain fog. Interestingly, those symptoms in PCOS appear more strongly correlated with insulin resistance and sleep disruption than with androgen level per se. Testosterone therapy in women of reproductive age carries meaningful risks, including virilization, polycythemia, and fetal harm, and is not recommended for cognitive symptoms in this life stage.

Perimenopause (Typically Ages 42-52)

This is where most women seeking help are, and paradoxically where the evidence is thinnest. Testosterone levels are still within or near the premenopausal range for many perimenopausal women; what changes dramatically is estradiol variability. The brain fog of perimenopause is largely attributed to estradiol fluctuation and sleep disruption from vasomotor symptoms, not androgen deficiency. Adding testosterone to address a problem that may be driven by estrogen is mechanistically uncertain. Some clinicians add low-dose testosterone to an HRT regimen in perimenopause when sexual function complaints coexist, but prescribing it specifically for cognition in this stage is extrapolation, and patients deserve to know that.

Natural Postmenopause

This is the population studied in most RCTs. The picture is mixed: the largest trial (APHRODITE) was negative for cognition, while smaller trials and a 2020 systematic review by Islam et al. In Maturitas found a trend toward improved verbal memory and processing speed in postmenopausal women on testosterone, particularly in those not already on estrogen therapy. The review covered 8 RCTs and noted significant heterogeneity in doses, formulations, and cognitive assessment tools, making pooled conclusions fragile.

Surgical Menopause (Bilateral Oophorectomy)

The Sherwin trials, the strongest positive data, were all in this population. The abrupt loss of both estrogen and testosterone following oophorectomy may create a deficit that is meaningfully different from natural menopause. Women who undergo oophorectomy before age 45 have a significantly elevated risk of cognitive decline and dementia later in life, a finding from the Rochester Epidemiology Project that raises the stakes for hormonal replacement in this group. Whether testosterone contributes independently to protection, or whether the signal is driven entirely by the aromatization to estradiol, is still unresolved.

Dosing, Formulations, and Monitoring for Women

No testosterone product is FDA-approved for women in the United States. In clinical practice, compounded transdermal preparations (gels, creams, lotions) are the most common route, though a 1% testosterone cream at a dose of 10 mg/day (delivering roughly 300 mcg systemically) is the benchmark used in the majority of trials. In the UK and Australia, a 1% testosterone gel (Androfeme, available in Australia) is formulated specifically for women at 5 mg/day application.

Starting Dose and Titration

The Global Consensus recommends starting at the lowest dose that keeps free testosterone within the premenopausal reference range as measured by LC-MS/MS. Immunoassay testosterone measurements are unreliable at female concentrations; if your provider is using a standard immunoassay, the measurement is essentially meaningless.

A practical starting approach used in many menopause specialty practices:

• Transdermal gel or cream: 1-2 mg testosterone per day applied to inner thigh or lower abdomen, avoiding breast tissue
• Recheck free testosterone by LC-MS/MS at 6 weeks and 3 months
• Dose adjustments guided by serum level, not symptom alone
• Maximum dose should not push free testosterone above the upper premenopausal reference limit (roughly 8-10 pg/mL free testosterone by LC-MS/MS)

Monitoring Parameters

| Parameter | Frequency | Threshold for Action | |---|---|---| | Free testosterone (LC-MS/MS) | 6 weeks, 3 months, then every 6 months | Dose reduce if supraphysiologic | | Hematocrit | Every 6 months | Hold if >50% | | Lipid panel | Every 6 months | Review cardiovascular risk | | Signs of virilization | Each visit | Dose reduce or stop | | Liver enzymes | At baseline, annually | Elevated: evaluate formulation |

Oral methyltestosterone is not recommended in women because of hepatotoxicity risk and unfavorable HDL effects. Pellet implants are also not endorsed by the Global Consensus because they produce uncontrolled, often supraphysiologic blood levels with no ability to adjust the dose once implanted.

Pregnancy, Lactation, and Contraception

Testosterone is contraindicated in pregnancy. This is not a nuanced statement. Testosterone is a Category X teratogen: fetal exposure causes virilization of a female fetus, including clitoromegaly and labioscrotal fusion. Any woman who has not reached menopause (defined as 12 consecutive months without a period) and who is prescribed testosterone must use reliable contraception throughout treatment and for at least 3 months after stopping, given the time required for serum levels to clear.

IUDs (hormonal or copper) or barrier methods are preferred over combined hormonal contraceptives in this context, because combined OCs suppress SHBG and further lower free testosterone, potentially negating the treatment effect.

Lactation: Testosterone transfers into breast milk. There are no adequate safety studies in nursing infants. Testosterone should not be used during breastfeeding. The risk to an infant from androgen exposure during the neonatal period includes theoretical effects on developing reproductive tissue.

If you are trying to conceive: Stop testosterone before attempting conception and allow serum levels to normalize. Testosterone suppresses LH pulsatility and can impair ovulation. Women with PCOS who are already hyperandrogenic should not be prescribed additional testosterone under any circumstances.

Who This Is Right For, and Who It Is Not

May be a reasonable option if you are:

• Postmenopausal (naturally or surgically) with documented low-normal free testosterone by LC-MS/MS
• Experiencing both HSDD and cognitive symptoms, since the sexual function evidence is stronger and cognition is a plausible secondary benefit
• Already on or considering systemic HRT, since testosterone is generally added to rather than substituted for estrogen therapy in postmenopause
• Informed that the cognitive evidence is preliminary and that a treatment trial should be time-limited (3-6 months) with objective reassessment

Probably not right for you if you are:

• Perimenopausal with intact ovarian function and brain fog as the sole complaint: the evidence does not support testosterone as a first-line or second-line option here
• Pregnant, trying to conceive, or breastfeeding
• Living with a history of breast cancer or any androgen-sensitive cancer: the Global Consensus states insufficient safety data exist and counsels against use
• Already showing signs of androgen excess (hirsutism, acne, alopecia): adding testosterone will worsen these
• Relying on a pellet implant with no ability to monitor or adjust dose

The Evidence Gap: What Women Deserve to Know

Women were systematically excluded from early testosterone trials, which were designed around male hypogonadism. The result is that the pharmacokinetics, optimal dosing, long-term safety (cardiovascular, breast, endometrial), and dose-response for cognitive outcomes have never been fully characterized in female populations. The 2019 Global Consensus explicitly acknowledged this gap and called for large, long-term RCTs in women. As of this writing, those trials have not been completed.

What clinicians currently work with is this: a biologically plausible mechanism, a handful of small positive trials in surgical menopause, one large null trial in natural menopause, no long-term safety data beyond 2 years, and a patient population that is genuinely suffering from brain fog and finding few other options. That is an honest summary. Providers who tell you this is established therapy for cognition are overstating the evidence; providers who dismiss it entirely are ignoring a real signal.

A 2023 position statement from The Menopause Society on testosterone therapy reiterated that testosterone should not be recommended for cognitive symptoms outside of a clinical trial setting, while acknowledging that individual clinical judgment may differ for postmenopausal women with concurrent HSDD and cognitive complaints.

Practical Clinical Considerations for the Prescribing Conversation

If you are considering bringing up testosterone with your provider, come prepared with the following:

1. Your menopausal status and how it was determined (last period date, FSH if tested)
2. Whether you have had a bilateral oophorectomy and at what age
3. A free testosterone level drawn by LC-MS/MS (not immunoassay) if available
4. A list of current medications, particularly those affecting SHBG (combined OCs, thyroid hormone, liver enzyme inducers)
5. Your contraception status if premenopausal or perimenopausal

The conversation should include a documented discussion of the off-label nature of the prescription, the formulation being used (compounded vs. Commercially available), the monitoring plan, and a defined reassessment point. ACOG Committee Opinion 532 supports shared decision-making for off-label hormonal prescribing when the clinical rationale is documented and the patient is informed.

A 3-month trial with objective cognitive assessment before and after (even a validated self-report tool such as the Perceived Deficits Questionnaire-Menopause, PDQ-M) gives both patient and provider something concrete to evaluate. If there is no discernible improvement in 3-6 months, continuing for cognitive reasons is not justified by current evidence.