Hormones, Explained: What Testosterone, Estrogen and Thyroid Actually Do to Performance

Why hormones became the most oversold category in health

Few areas of health have travelled so quickly from clinical obscurity to consumer commodity. Direct-to-consumer testosterone platforms, menopause telehealth, thyroid 'optimization' protocols and adrenal fatigue programmes now form a multibillion-dollar market built on a genuinely persuasive premise: that fatigue, poor recovery, low motivation, weight gain and diminished libido might share a single, correctable cause.

Sometimes they do. Untreated hypogonadism, symptomatic menopause and overt thyroid disease all produce exactly that constellation, and all respond to properly indicated treatment. The problem is that the same symptom cluster is produced far more often by insufficient sleep, chronic energy deficit or surplus, untreated sleep apnoea, alcohol, overtraining, depression and unmanaged stress — none of which is corrected by a prescription, and several of which suppress hormone levels themselves, producing a lab value that appears to confirm the wrong diagnosis.

Testosterone: what the trials actually showed

Testosterone regulates muscle protein synthesis, bone density, erythropoiesis, libido, and — less appreciated — mood and motivational drive. Levels decline gradually with age, roughly 1–2% per year after the third decade in most longitudinal cohorts, but the decline attributable to ageing alone is smaller than the decline attributable to accumulating obesity, illness and medication use.

The Testosterone Trials, published in the New England Journal of Medicine, remain the most informative evidence in older men. In 790 men aged 65 and older with unequivocally low testosterone and symptoms, one year of gel therapy produced a clear improvement in sexual function, a modest improvement in mood and depressive symptoms, and correction of unexplained anaemia. It did not meaningfully improve vitality or walking distance. A companion substudy found increased coronary artery non-calcified plaque volume on CT — a finding of uncertain clinical meaning that nonetheless tempered enthusiasm.

The safety question was addressed directly by TRAVERSE, a randomised placebo-controlled trial of over 5,200 men aged 45–80 with hypogonadism and either established cardiovascular disease or high risk. Testosterone was non-inferior to placebo for major adverse cardiac events. It did, however, show higher rates of atrial fibrillation, acute kidney injury and pulmonary embolism — a reminder that non-inferiority is not the same as harmlessness.

The Endocrine Society clinical practice guideline is correspondingly narrow: diagnose hypogonadism only in men with consistent symptoms and unequivocally low morning total testosterone confirmed on at least two separate occasions, and only after evaluating for reversible and secondary causes. It explicitly recommends against treating men with normal levels or with low levels in the absence of symptoms.

That last row deserves emphasis, because it is the most common omission in direct-to-consumer prescribing. Exogenous testosterone suppresses the hypothalamic-pituitary-gonadal axis and reliably reduces sperm production. Any man who may want children should have that conversation before the first dose, not after.

The reversible causes that are skipped

Before therapy is a reasonable consideration, a serious evaluation looks for the causes of a low reading that respond to something other than a prescription.

• Obesity and visceral adiposity — adipose aromatase activity lowers testosterone, and weight loss reliably raises it; meta-analyses of both diet-induced and surgical weight loss show substantial increases.
• Obstructive sleep apnoea — independently associated with suppressed testosterone and a driver of the same fatigue attributed to hormones.
• Sleep restriction — one week of five-hour nights reduced daytime testosterone by 10–15% in healthy young men in a controlled study.
• Energy deficiency and overtraining — relative energy deficiency in sport suppresses the reproductive axis in both men and women.
• Medications — opioids, glucocorticoids and some antidepressants suppress the axis directly.
• Alcohol, uncontrolled diabetes, thyroid disease, and hyperprolactinaemia — each with its own workup.

Treating a suppressed hormone level without asking what suppressed it is not optimization. It is masking a diagnosis.

Estrogen and the menopause transition

The women's side of this field is still recovering from a communication failure. The 2002 Women's Health Initiative results were reported in a way that led to a rapid, near-global withdrawal from hormone therapy — including among women for whom the benefit-risk balance was clearly favourable.

Subsequent age-stratified analysis changed the picture materially. In the WHI's own 2013 follow-up across both trials, women aged 50–59 at initiation showed a more favourable risk profile than older initiators, and the estrogen-alone arm in hysterectomised women showed no increase — and in some analyses a reduction — in breast cancer incidence. The prevailing clinical framing became the timing hypothesis: benefit-risk depends heavily on age at initiation and time since menopause.

The Menopause Society's current position statement is explicit that for symptomatic women under 60 or within ten years of menopause onset, and without contraindications, the benefits of hormone therapy generally outweigh the risks for vasomotor symptoms, genitourinary syndrome of menopause and fracture prevention. Risk is not zero — combined therapy carries a small absolute increase in breast cancer risk with duration, and oral formulations carry venous thromboembolic risk that transdermal routes largely avoid.

Two clarifications matter for anyone evaluating a programme. First, compounded 'bioidentical' hormone preparations are not FDA-approved, are not subject to the same potency and purity oversight, and were reviewed by the National Academies in 2020 with a recommendation to restrict their use to patients who cannot use approved products. Second, salivary hormone testing to titrate therapy has no validated basis and is not endorsed by endocrine professional societies.

Testosterone in women is a narrower story. The 2019 Global Consensus Position Statement concluded that the only evidence-based indication is hypoactive sexual desire disorder in postmenopausal women, using doses that produce premenopausal physiological concentrations. Evidence for benefit in cognition, mood, bone or body composition was judged insufficient.

Thyroid: the most over-treated number in the category

Thyroid hormone sets metabolic rate, thermogenesis, cardiac output and cognitive pace. Overt hypothyroidism is unambiguous and treatment is unambiguously beneficial. The controversy sits entirely in the subclinical zone — mildly elevated TSH with normal free T4 — which is where most 'thyroid optimization' occurs.

The TRUST trial randomised 737 adults aged 65 and over with persistent subclinical hypothyroidism to levothyroxine or placebo. Thyroid levels normalised. Hypothyroid symptoms and tiredness did not improve. A subsequent pooled analysis in JAMA across 21 trials and more than 2,100 adults found no clinically relevant benefit in quality of life or symptoms from treating subclinical hypothyroidism.

There are exceptions clinicians take seriously — markedly elevated TSH, positive thyroid peroxidase antibodies, pregnancy or fertility planning, and younger symptomatic patients — but they are exceptions, and they are defined clinically rather than by a laboratory flag. Reverse T3 testing, meanwhile, is not recommended by the American Thyroid Association for the evaluation of hypothyroidism, despite its popularity in wellness panels.

Cortisol and the 'adrenal fatigue' problem

Cortisol is the most misrepresented hormone in consumer health. It follows a steep diurnal rhythm, spikes with acute stress and exercise, and is chronically dysregulated by poor sleep — all of which is real and consequential.

What is not real is 'adrenal fatigue' as a diagnosis. A systematic review of 58 studies and more than 10,000 participants, published in BMC Endocrine Disorders, found no substantiation for adrenal fatigue as a medical condition, and noted the heterogeneity and inadequate validation of the tests used to diagnose it. Adrenal insufficiency is a genuine and serious endocrine disorder with specific diagnostic criteria. It is not the same thing, and conflating the two delays real diagnoses.

How a defensible hormone evaluation is actually conducted

1. 01Start with symptoms and history, not a panel. Fatigue, low libido, poor recovery, mood change and cognitive fog are non-specific; the pattern, timeline and context determine which axis is plausible.
2. 02Test correctly. Total testosterone on two separate early-morning, fasted samples; free testosterone by equilibrium dialysis or calculated with SHBG when binding is likely altered; TSH with free T4; and, where indicated, LH, FSH, prolactin and estradiol.
3. 03Interpret against the clinical picture, not against a marketing target. Assay methods and reference ranges vary between laboratories, and a single borderline value is not a diagnosis.
4. 04Rule out reversible drivers — sleep apnoea, sleep restriction, obesity, energy deficiency, alcohol, medications, uncontrolled metabolic disease, depression.
5. 05Correct the foundation first and re-measure. Sleep, training load, nutrition and body composition change hormone levels measurably, and a re-test after eight to twelve weeks of real behaviour change often resolves the question.
6. 06If therapy is indicated, it is prescribed, titrated and monitored by a licensed clinician — with baseline and follow-up haematocrit, PSA where appropriate, and a documented plan for fertility, cardiovascular risk and duration.