When “Normal” Testosterone Misses the Diagnosis: A Functional Case Study in Female Androgen Deficiency

The Patient

A 47-year-old perimenopausal woman presented with a two-year history of low libido, persistent fatigue despite reported “adequate sleep”, declining exercise tolerance, flat affect, and difficulty recovering between workouts. She had been told her hormone levels were “normal” on two prior evaluations and was offered an SSRI for mood. She declined and sought a functional evaluation.

Relevant history: regular menstrual cycles that had become increasingly irregular over the prior 12 months, moderate chronic stress, a desk-based job with low daily movement, and a history of oral contraceptive use for twelve years in her twenties and thirties. No prior hormone replacement therapy. No significant gut complaints when asked, though she reported occasional bloating and loose stools.

Initial Lab Evaluation: What the Numbers Actually Show

The ordering clinician had previously checked total testosterone, which came back at 24 ng/dL — within the conventional female reference range of 8 to 48 ng/dL. The result was reported as normal, and the conversation ended there.

A full functional hormone assessment told a different story:

The central finding was not the total testosterone. It was the free testosterone of 0.5 pg/mL in the context of an SHBG of 68 nmol/L. While 68 nmol/L sits within the conventional reference range, it is in the upper portion of that range. When total testosterone is already low-normal, a SHBG at the higher end of normal is sufficient to sequester the majority of available testosterone, leaving a free fraction that falls below functional threshold. A reminder that SHBG does not need to be frankly elevated to be clinically relevant. Its impact is always interpreted relative to the total testosterone it is binding.

The insulin finding tells a separate but compounding story. A fasting insulin of 17 mIU/L reflects early insulin resistance and notably, doesn’t seem to be driving the SHBG elevation. Insulin typically has a suppressive effect on hepatic SHBG production; hyperinsulinemia is associated with lower SHBG, not higher. In this patient, the insulin resistance was contributing to the androgen picture through a different mechanism: likely driving increased aromatization of androgen precursors to estrogens in peripheral adipose tissue, diverting substrate away from testosterone synthesis, and compounding the low DHEA-S picture through its effects on adrenal and metabolic function. These are concurrent findings that compound each other through distinct pathways, not a single driver, but a converging set of moderate disruptions, very typical of what we see in practice 

Understanding the Compounding Drivers

Several factors were contributing simultaneously to the patient’s functional androgen insufficiency, each through a different mechanism:

SHBG in the upper-normal range relative to low-normal total testosterone — the arithmetic of binding capacity matters here. With total testosterone at 26 ng/dL and SHBG at 68 nmol/L, the proportion of testosterone available in the unbound, biologically active fraction was insufficient to meet tissue demand. This relationship is why ordering total testosterone alone — without SHBG and free testosterone — is clinically incomplete in women. A total testosterone of 26 ng/dL means something very different in a woman with SHBG of 30 nmol/L versus one with SHBG of 68 nmol/L.

Early insulin resistance as an independent metabolic disruptor — the fasting insulin of 17 mIU/L was corroborated by CGM data showing a characteristic dysfunctional pattern. Glucose levels were running flat and low through the morning hours, consistent with prolonged overnight fasting, then rising progressively through the afternoon and evening as the patient consumed the majority of her caloric load. Post-dinner glucose excursions regularly exceeded 140 mg/dL and remained elevated for two or more hours before bedtime. This back-loaded eating pattern — combined with fasted morning cardio — was generating repeated late-day insulin surges at a time when insulin sensitivity is naturally declining, producing cumulative hyperinsulinemia despite an acceptable fasting glucose. Metabolically, elevated insulin in this context was contributing to increased peripheral aromatization and placing additional substrate competition pressure on an already depleted DHEA-S pool.

Low DHEA-S — at 68 mcg/dL, the adrenal androgen precursor pool was meaningfully reduced for a 49-year-old. DHEA-S serves as the upstream substrate for peripheral testosterone synthesis. Reduced precursor availability can impact how much free testosterone can be generated regardless of binding dynamics. The low DHEA-S in the setting of preserved cortisol output suggests a relative shift in adrenal steroidogenesis favoring glucocorticoid over androgen production.

Low-grade inflammation — the hsCRP of 1.4 mg/L, while not dramatically elevated, adds to the clinical picture. Inflammatory signaling upregulates hepatic acute-phase proteins and can contribute to SHBG production independent of other drivers. In the context of insufficient Vitamin D, early insulin resistance, and HPA dysregulation, even modest inflammation represents a modifiable variable worth addressing.

This pattern — upper-normal SHBG compressing an already low-normal total testosterone, concurrent insulin resistance disrupting precursor metabolism and aromatization, depleted DHEA-S limiting upstream substrate, and low-grade inflammation sustaining the environment — is a common presentation in symptomatic perimenopausal women who are told their labs are normal. No single finding is dramatic. The clinical significance emerges from reading them together.

DUTCH Testing: Completing the Picture

A DUTCH Plus test was ordered to assess cortisol pattern, estrogen metabolite distribution, and androgenic metabolite excretion.

The cortisol awakening pattern showed a blunted awakening response with a flattened diurnal slope. Flattened diurnal cortisol rhythms have been associated in observational studies with fatigue, depressive symptoms, and reduced well-being, though a direct causal relationship remains difficult to establish. The concurrent finding of low DHEA-S is consistent with age-related zona reticularis decline, which may be further influenced by chronic HPA axis activation. Together, these findings suggest a relative shift in adrenal steroidogenesis favoring glucocorticoid over androgen output, which could contribute to the patient’s symptom burden.

5-alpha reductase activity, as reflected by the androsterone marker, was low. This enzyme converts testosterone to its more potent derivative dihydrotestosterone in peripheral tissues. Low androsterone excretion indicated impaired downstream androgen activation — meaning even when free testosterone was available at the tissue level, its local conversion to the more biologically potent DHT was attenuated.

Estrogen metabolite distribution showed preferential production of 16-OH estrone relative to 2-OH estrone — a pattern associated with compromised Phase II methylation capacity. This finding was important context for any future hormone therapy decisions and pointed toward the need for methylation support independent of the testosterone question.

The Gut as a Variable

Given the estrogen metabolite pattern on DUTCH and the patient’s history of occasional GI symptoms, stool testing was ordered. Results revealed low Lactobacillus and Bifidobacterium populations, elevated beta-glucuronidase activity, and insufficiency dysbiosis. Elevated beta-glucuronidase deconjugates estrogen metabolites that have been packaged for fecal excretion, returning them to circulation and increasing the total estrogen burden and compounding the already-compromised Phase II metabolism picture. The combination of impaired methylation on DUTCH and elevated beta-glucuronidase on stool testing meant that estrogen clearance was dysfunctional at two separate checkpoints. Before initiating any hormone therapy, the gut environment required attention.

Clinical Reasoning: What This Case Teaches

Total testosterone is an unreliable primary metric in women. The reference range is wide and was derived from population data that does not account for SHBG-mediated binding differences. Free testosterone, calculated or direct, is a clinically actionable number and should be ordered in addition to SHBG and total testosterone to interpret meaningfully.

SHBG does not need to be frankly elevated to be clinically significant. In a woman with low-normal total testosterone, an SHBG in the upper portion of the normal range could produce the same functional outcome as markedly elevated SHBG would in a woman with higher total testosterone. The clinical question is always: what is the binding capacity relative to the available substrate?

Insulin resistance and hyperinsulinemia affect androgen metabolism through several well-characterized mechanisms, though their net effect is typically androgen excess rather than insufficiency. Hyperinsulinemia suppresses hepatic SHBG production, increasing free androgen bioavailability. Simultaneously, insulin acts synergistically with LH to stimulate ovarian androgen synthesis and, via adrenal insulin receptor signaling, can increase adrenal androgen output. In the setting of obesity, increased adipose aromatase activity may convert androgens to estrogens, though the clinical significance of this pathway for androgen levels in women remains less well defined. These mechanisms have direct implications for how metabolic intervention is framed — improving insulin sensitivity may modulate androgen levels, but the expected direction is typically a reduction in androgen excess rather than correction of androgen insufficiency.

Crosstalk between the HPA and HPG axes is well established. Chronic stress suppresses reproductive function primarily through CRH-mediated inhibition of the GnRH pulse generator and direct glucocorticoid suppression of pituitary gonadotropin secretion and gonadal steroidogenesis. Separately, chronic stress is associated with diminished DHEA-S output, likely reflecting age-related zona reticularis decline that may be accelerated by sustained HPA activation. When functional contributors to androgen insufficiency are present, addressing reversible underlying factors — including metabolic, nutritional, and stress-related drivers — alongside or before initiating testosterone therapy is consistent with guideline recommendations for managing functional hypogonadism

Meal timing and carbohydrate distribution are underutilized clinical levers. This patient was not eating too many carbohydrates — she was eating them at the wrong time and in the wrong pattern. Prolonged morning fasting combined with back-loaded evening carbohydrate intake generates repeated late-day insulin surges and suppresses morning cortisol recovery. For women with both HPA dysregulation and early insulin resistance, this pattern creates compounding metabolic and adrenal dysfunction that is entirely addressable without caloric restriction.

Gut health and estrogen metabolism are part of the androgen picture. Elevated beta-glucuronidase and poor methylation capacity affect the broader steroid hormone milieu. In a patient preparing for any hormone therapy, the gut and detoxification pathways are clinical prerequisites, not afterthoughts.

Treatment Approach and Sequencing

The treatment plan was built in deliberate sequence rather than initiating testosterone replacement as a first step.

Phase one addressed the foundational variables: a gut protocol targeting dysbiosis and beta-glucuronidase normalization, activated B vitamins and methylation support for the compromised Phase II picture, vitamin D repletion targeting 50 to 70 ng/mL, and a targeted metabolic intervention.

The metabolic intervention warrants specific clinical detail because it illustrates a principle that applies broadly in this population: the goal was not carbohydrate restriction but carbohydrate redistribution. The patient had been habitually skipping breakfast, performing fasted cardio in the morning, and consuming the majority of her carbohydrate load in the evening before bed. This pattern was modified in several deliberate ways.

Morning fasting was discontinued. A protein-forward meal was introduced within 60 to 90 minutes of waking to support the cortisol awakening response and attenuate the HPA-driven catabolic state that prolonged overnight fasting was sustaining. Carbohydrate intake was redistributed toward midday, with a meaningful portion consumed around the afternoon resistance training session, where it could support performance and glucose disposal rather than late-evening insulin surges. Evening meals were restructured around protein and non-starchy vegetables, directly addressing the post-dinner glucose excursions that CGM had confirmed were regularly exceeding 140 mg/dL and allowing for a 3 hour window between dinner and bedtime.

Fasted morning cardio was replaced with short, low-intensity morning walks — providing movement stimulus without the cortisol and catabolic cost of sustained fasted aerobic output. Resistance training three days per week was introduced and scheduled in the afternoon, when cortisol is naturally declining and insulin sensitivity is at its daily peak. This timing supports both glucose disposal and anabolic signaling and is a clinically underutilized strategy in perimenopausal women managing both HPA dysfunction and body composition concerns simultaneously.

Phase two addressed HPA axis dysregulation directly: adaptogenic support, cortisol awakening response improvement strategies, and sleep hygiene scaffolding. DHEA-S was monitored with intent to supplement if foundational work did not produce adequate recovery of the adrenal precursor pool.

Eight-Week Reassessment: Lab and Symptom Response

At the eight-week mark, repeat labs reflected meaningful improvement:

Symptomatically, the patient reported meaningful improvement in morning energy, more stable mood across the day, and improved sleep quality. Libido and exercise recovery, however, remained below her functional baseline. This is the expected outcome of well-executed foundational work: it resolves the modifiable drivers, improves the hormonal environment, and clarifies which symptoms have a residual hormonal component requiring direct support. Partial symptomatic response after foundational correction is not a failure — it is a clinical signal that the sequencing is working and that the next layer of intervention is now appropriately indicated.

Hormonal sequencing at reassessment proceeded as follows:

Progesterone was introduced first. With the Pg/E2 ratio remaining suboptimal and the 16-OH estrone pattern confirmed on DUTCH, 100 mg of bioidentical oral micronized progesterone was initiated at bedtime. Progesterone addresses the ratio directly, supports allopregnanolone production with downstream implications for sleep architecture, mood stability, and GABAergic tone, and provides a well-tolerated first hormonal intervention with a favorable safety profile in this population.

DIM was introduced concurrently to support Phase II estrogen metabolism, promoting the 2-OH pathway and reducing 16-OH estrone dominance. This addressed the methylation and detoxification picture identified on DUTCH without requiring more aggressive intervention at this stage. Together, progesterone and DIM targeted the estrogen metabolism and Pg/E2 findings directly, creating a more favorable hormonal environment before testosterone was introduced.

Following two to three months on progesterone and DIM, with repeat DUTCH and symptom reassessment planned, the patient’s interest in trialing testosterone was revisited. She had expressed this preference consistently throughout the evaluation, and after foundational, gut, and estrogen-pathway work was in place, this was a reasonable and well-sequenced clinical decision.

Prescribing Testosterone in Women: Clinical Considerations

When testosterone is indicated in women after appropriate foundational sequencing, the prescribing approach differs substantially from male TRT. Compounded testosterone in women is typically prepared as a topical cream, allowing precise low-dose titration that is not achievable with commercially available preparations. Concentrations are generally in the range of 1 to 2 mg per 0.1 mL, applied to thin-skinned areas such as the inner wrist, inner arm, or upper thigh. Starting doses in women are substantially lower than in men — commonly 0.5 to 1 mg per day — with upward titration guided by symptom response and free testosterone levels rather than total testosterone.

Application site rotation matters clinically. Repeated application to the same site can cause local accumulation and unintended androgenic effects at doses that would otherwise be well-tolerated. The monitoring interval post-initiation is typically six to eight weeks, with free testosterone, SHBG, and structured symptom tracking as the primary response parameters. Estradiol is monitored for aromatization, and virilization symptoms — acne, hair thinning, voice change, clitoral sensitivity changes — require explicit review and documentation at each visit.

The full treatment sequence — gut and methylation support, carbohydrate redistribution and resistance training, HPA support, progesterone and DIM, followed by low-dose systemic testosterone, representing a layered approach in which each intervention has a defined rationale, a monitoring parameter, and a clear position in the overall sequence.

Monitoring Parameters

Key monitoring parameters in this patient included: free testosterone at six to eight weeks post-initiation, SHBG to assess ongoing response to metabolic intervention, hematocrit at baseline and periodically (erythrocytosis is uncommon in women at physiologic doses but warrants monitoring), estradiol for aromatization assessment, and structured symptom tracking across libido, energy, affect, and exercise recovery. Virilization symptoms were reviewed at each visit with documentation of any changes.

A Note on the Evidence Landscape

Testosterone therapy in women occupies an unusual position in the evidence hierarchy. The most robust RCT data supports its use for hypoactive sexual desire disorder in postmenopausal women, with a 2019 global consensus statement from the International Society for the Study of Women’s Sexual Health providing the clearest clinical guidance to date. Evidence for broader symptom domains — fatigue, mood, cognitive function, body composition — is largely observational and mechanistically grounded rather than RCT-confirmed in female populations. Practitioners working in this space should communicate this distinction to patients while recognizing that the absence of large-scale trial data in women reflects a research funding and design problem as much as an evidence problem.

The clinical skills required to evaluate androgen deficiency in women accurately — ordering and interpreting free testosterone and SHBG, understanding how upper-normal SHBG compresses free testosterone in the context of low-normal total levels, reading DUTCH results for adrenal and androgenic output, sequencing treatment across gut, metabolic, HPA, and hormone layers, and prescribing compounded testosterone at physiologic doses — are not taught in conventional training. They are core content in the Functional Hormone Mastery program. If you are seeing women whose symptoms are not explained by their standard hormone panels, this training provides the framework to find what the panels are missing.