Free Testosterone: Drugs That Distort This Test

What Free Testosterone Actually Measures

Free testosterone is the fraction of circulating testosterone that is unbound to carrier proteins and available to enter cells and activate androgen receptors. About 65 to 68% of testosterone binds tightly to SHBG, 30 to 33% binds loosely to albumin, and only 1 to 3% circulates truly free [1]. The Endocrine Society's 2018 clinical practice guideline recommends measuring free T "when total testosterone concentrations are close to the lower limit of the normal range or when SHBG concentrations are likely to be abnormal" [2].

This matters because any drug that raises or lowers SHBG will change the free fraction without necessarily changing total testosterone production. A patient on oral estrogen therapy might have a total T of 450 ng/dL but a free T well below the reference interval because her SHBG doubled. Clinicians who rely on total T alone will miss this shift entirely.

Equilibrium dialysis remains the gold-standard measurement method, though liquid chromatography-tandem mass spectrometry (LC-MS/MS) for total T combined with a calculated free T (using the Vermeulen equation) is widely accepted [3]. Direct analog immunoassays for free T are considered unreliable by the Endocrine Society and should be avoided when accuracy is critical [2].

Opioids: The Most Underrecognized Suppressor

Chronic opioid use is one of the most common and most overlooked causes of misleadingly low free testosterone. A 2010 systematic review in The Journal of Clinical Endocrinology & Metabolism found that opioid-induced androgen deficiency (OPIAD) occurs in 21 to 86% of men on long-term opioid therapy, depending on the opioid type and dose [4]. The mechanism is straightforward: opioids suppress gonadotropin-releasing hormone (GnRH) pulsatility at the hypothalamus, leading to reduced LH and FSH secretion and, consequently, reduced testicular testosterone synthesis.

Methadone appears to be the worst offender. A study of 54 men on methadone maintenance found that 65% had total testosterone below 300 ng/dL [5]. Sustained-release morphine, oxycodone, and hydrocodone all produce similar effects at higher doses.

The Endocrine Society's 2018 guideline states: "We suggest against routinely prescribing testosterone therapy to all men with opioid-induced androgen deficiency but recommend discussing potential benefits and risks with individual patients" [2]. The practical takeaway: if a patient on chronic opioids returns a low free T, the opioid itself may be the cause. Reducing the opioid dose or rotating to buprenorphine (which has a weaker suppressive effect) should be considered before initiating TRT [4].

Oral Estrogens and SHBG Manipulation

Oral estrogens, including combined oral contraceptive pills (COCs) and oral menopausal hormone therapy, produce a pronounced first-pass hepatic effect that stimulates SHBG synthesis. A 2004 study in Fertility and Sterility demonstrated that COCs increased SHBG by 200 to 400% in premenopausal women, dropping free testosterone by 40 to 60% [6]. This SHBG elevation can persist for months after discontinuation.

This has direct diagnostic implications. A woman being evaluated for polycystic ovary syndrome (PCOS) who recently stopped COCs may show a transiently low free T that masks underlying hyperandrogenism. The 2023 international evidence-based PCOS guideline recommends waiting at least three months after discontinuing hormonal contraception before measuring androgens for PCOS diagnosis [7].

Transdermal estrogen does not produce the same SHBG spike because it bypasses first-pass hepatic metabolism. A comparative trial published in The New England Journal of Medicine found that transdermal estradiol had minimal impact on SHBG, while oral conjugated equine estrogens increased SHBG by roughly 100% at standard doses [8]. For patients who need estrogen therapy and accurate androgen assessment, the transdermal route preserves test validity.

Glucocorticoids: Central Suppression of the HPG Axis

Glucocorticoids suppress free testosterone through hypothalamic-pituitary-gonadal (HPG) axis inhibition. Both exogenous steroids (prednisone, dexamethasone, hydrocortisone) and endogenous hypercortisolism reduce GnRH pulse frequency and amplitude. A 2019 study published in The Journal of the Endocrine Society found that men receiving prednisone 10 mg/day for 12 weeks had a mean total testosterone decline of 32%, with proportional free T reductions [9].

Even inhaled corticosteroids at high doses can measurably suppress testosterone. A cross-sectional analysis of 236 men with asthma using high-dose fluticasone (>1 to 000 mcg/day) showed significantly lower morning testosterone than matched controls [10]. The effect is dose-dependent and typically reversible within weeks of discontinuation.

For patients on chronic glucocorticoids who return low free T values, the clinical question is whether testosterone replacement provides benefit beyond addressing the glucocorticoid. A small RCT (N=61) in men on chronic prednisone found that testosterone supplementation improved lean mass and bone mineral density at the lumbar spine over 12 months [11].

Anticonvulsants: The SHBG-Raising Enzyme Inducers

Hepatic enzyme-inducing anticonvulsants, including phenytoin, carbamazepine, and phenobarbital, increase SHBG production by 20 to 40%. A classic 1990 study in Epilepsia measured SHBG in 60 men on carbamazepine monotherapy and found levels 39% higher than untreated controls, with corresponding free T reductions of 22% [12]. The mechanism mirrors the oral estrogen effect: increased hepatic protein synthesis.

Valproate works in the opposite direction. It inhibits hepatic SHBG production and has been associated with elevated free testosterone and hyperandrogenic features in women with epilepsy. A 2001 study in The New England Journal of Medicine found that women taking valproate had significantly higher free testosterone and a 40% prevalence of polycystic ovaries on ultrasound compared to 20% in those on lamotrigine [13].

For men on enzyme-inducing anticonvulsants, free T measurement is especially important because total T may appear normal while free T is clinically low. The Endocrine Society recommends free T in exactly this scenario: when total T is borderline and SHBG is expected to be altered by medication [2].

Exogenous Testosterone and Anabolic Steroids

Exogenous testosterone obviously raises free T. That is its purpose in TRT. But timing and formulation create interpretive pitfalls. Testosterone cypionate injections produce a peak 24 to 48 hours post-injection and a trough at 7 to 14 days. A free T drawn at the peak might show supraphysiologic levels (above 25 pg/mL in men) while a trough draw might appear low-normal, even on the same dose [14].

The Endocrine Society recommends drawing testosterone levels at the midpoint between injections for cypionate and enanthate formulations, and at any time for daily gels or patches [2]. Dr. Shalender Bhasin, lead author of the 2018 guideline, has noted: "The timing of blood sampling relative to testosterone injection is one of the most common sources of error in TRT monitoring" [2].

Anabolic-androgenic steroid (AAS) use from non-medical sources creates a different problem. Supraphysiologic doses of nandrolone, trenbolone, or stanozolol will suppress endogenous LH/FSH to near-zero and may not cross-react with testosterone immunoassays, resulting in paradoxically low measured testosterone despite extreme androgenic exposure [15]. LC-MS/MS specific to testosterone will correctly show low testosterone in these patients, but the clinical picture (acne, erythrocytosis, suppressed gonadotropins) will point toward exogenous androgen use.

Biotin, Supplements, and Assay Interference

Biotin (vitamin B7) deserves special attention because it interferes with the streptavidin-biotin immunoassay platform used by many commercial labs. The FDA issued a safety communication in 2017 warning that biotin supplementation at doses of 5 to 10 mg/day (common in hair and nail supplements) can cause "falsely high or falsely low results" depending on the assay architecture [16].

For competitive immunoassays (the format used by many free T analog tests), biotin causes falsely elevated results. For sandwich immunoassays, it causes falsely low results. A 2019 study in Clinical Chemistry showed that ingestion of 10 mg biotin produced clinically significant interference in testosterone immunoassays within 2 hours, with effects lasting up to 24 hours [17].

The practical fix is simple: stop biotin supplements 48 to 72 hours before any hormone blood draw. Labs using the Roche Elecsys platform are particularly susceptible. LC-MS/MS methods are not affected by biotin.

Other Drugs That Move Free Testosterone

Several additional medication classes alter free T results in clinically meaningful ways.

5-alpha reductase inhibitors (finasteride, dutasteride) do not change free testosterone directly. They block conversion of testosterone to dihydrotestosterone (DHT), and free T may actually rise 10 to 15% as the upstream substrate accumulates [18]. A clinician unaware of this effect might interpret a mildly elevated free T as a sign to lower a TRT dose, which would be the wrong move.

Ketoconazole at systemic doses (200 to 400 mg/day) directly inhibits testicular and adrenal steroidogenesis. This antifungal can reduce testosterone production by 30 to 50% within hours of administration [19]. Topical ketoconazole (shampoo) does not produce meaningful systemic absorption.

Spironolactone blocks the androgen receptor and weakly inhibits testosterone synthesis. In women treated for acne or hirsutism, doses of 100 to 200 mg/day reduce free testosterone by approximately 30 to 40% [20].

GnRH agonists and antagonists (leuprolide, degarelix) used in prostate cancer therapy suppress testosterone to castrate levels (<50 ng/dL total T), with free T falling proportionally [21]. A patient on these agents will have near-undetectable free T, which is the therapeutic goal, not a lab error.

Statins have a modest effect. A meta-analysis of 11 RCTs (N=1,352) found that statin therapy reduced total testosterone by a mean of 0.66 nmol/L but did not significantly change free testosterone [22]. This is generally not clinically significant but worth noting for patients with borderline values.

How to Get an Accurate Free Testosterone Result

Getting a reliable free T measurement requires controlling for medication effects and assay selection. The Endocrine Society recommends morning blood draws (before 10 AM) because testosterone follows a circadian rhythm with 20 to 30% higher levels in the early morning [2]. Dr. Bradley Anawalt, a co-author of the Endocrine Society guideline, has stated: "Two morning measurements showing low testosterone, using a reliable assay, are required before diagnosing androgen deficiency" [23].

A medication inventory is the first step. Patients should report all prescription drugs, OTC supplements (especially biotin), and any history of anabolic steroid use. For drugs that raise SHBG (oral estrogens, anticonvulsants), calculated free T using the Vermeulen equation or equilibrium dialysis will reflect the true bioavailable fraction better than total T alone.

For patients on TRT, timing relative to the last injection determines the result. Drawing blood at the midpoint for cypionate/enanthate injections, or at least 2 hours post-application for gels, provides the most representative value [2]. A free T above 25 pg/mL at the midpoint suggests the dose may need reduction; a free T below 5 pg/mL at the midpoint suggests inadequate dosing or poor absorption.

If a result seems discordant with clinical presentation, the most productive next step is to repeat the test using a different assay methodology. A free T by equilibrium dialysis will not be affected by SHBG fluctuations or biotin interference the way immunoassays can be.