Exogenous testosterone metabolism, T/E ratio screening, carbon isotope confirmation, and the Athlete Biological Passport — anatomy of the anti-doping pipeline
Anabolic-androgenic steroids (AAS) — testosterone esters, nandrolone, stanozolol, trenbolone and dozens of designer analogues — are taken to accelerate muscle protein synthesis and recovery. Once administered, they are extensively metabolized by hepatic CYP450 enzymes and peripheral 5α/5β-reductases, glucuronidated for renal excretion, and leave a chemical trail in urine that anti-doping laboratories are built to read.
Anabolic-androgenic steroids fall into structural families with distinct metabolic signatures:
• Testosterone esters (cypionate, enanthate, propionate): hydrolyzed by plasma/tissue esterases to free testosterone, then hepatically reduced/oxidized to androsterone and etiocholanolone, and glucuronidated by UGT2B17 (and UGT2B7) for urinary excretion as testosterone glucuronide (TG) and epitestosterone glucuronide (EG).
• 19-nor steroids (nandrolone decanoate): metabolized to 19-norandrosterone (19-NA) and 19-noretiocholanolone — markers essentially absent in untreated men, giving nandrolone abuse a very low natural background (WADA reporting threshold 2.5 ng/mL for men, 2.5 ng/mL unified since 2021 revision).
• 17α-alkylated orals (stanozolol, methandienone): resist first-pass hepatic clearance (hence oral bioavailability) but are hepatotoxic; stanozolol metabolites (3′-hydroxystanozolol) remain detectable for weeks to months due to fat-tissue sequestration and slow release.
• Designer steroids: unapproved, unregulated. Detected only after reference standards are synthesized — famously how BALCO's tetrahydrogestrinone (THG) evaded testing for years until a used syringe was submitted anonymously to USADA in 2003.
Urine remains the primary anti-doping matrix because steroid glucuronide/sulfate conjugates concentrate there at analytically convenient levels, and collection is non-invasive enough for out-of-competition testing. A chain-of-custody sample is split into A and B bottles at collection; only the A sample is opened for initial screening, preserving the B sample, sealed under athlete-witnessed conditions, for confirmatory re-analysis if the A sample triggers an Adverse Analytical Finding.
Samples are frozen and can be stored for up to 10 years (WADA Code statute of limitations) — enabling retrospective reanalysis with newer, more sensitive methods. This retroactive testing has produced a wave of stripped medals years after competition, most visibly the 2008 and 2012 Olympic re-tests using improved IRMS and mass spectrometry protocols unavailable at the time of the original Games.
Long-term sample storage means an athlete doping today can still be caught a decade later as detection technology improves — a substantial deterrent baked directly into the WADA Code's 10-year retention and re-testing rule.
The T/E ratio is the oldest and cheapest indirect steroid screen still in routine use: measure urinary testosterone glucuronide against epitestosterone glucuronide by GC-MS. Epitestosterone is an inactive stereoisomer of testosterone that the body produces at a roughly fixed rate independent of exogenous testosterone dosing, so administering testosterone selectively pushes the ratio upward while epitestosterone stays flat — an easily computed red flag.
Using a ratio rather than an absolute testosterone concentration cancels out hydration status, urine dilution, and inter-individual variance in baseline steroid production — all of which vary sample-to-sample for the same athlete. Because epitestosterone production is essentially unaffected by exogenous testosterone administration (it derives from a different, largely independent enzymatic branch point in steroidogenesis — 17α-hydroxylation rather than 17β), any exogenous T dose raises the numerator while the denominator stays put, amplifying the signal.
The original WADA/IOC threshold was set at T/E > 6 in the 1980s, then lowered to > 4 in 2005 as assay sensitivity and population data improved, tightening the net while still leaving margin against the small fraction of the population with naturally elevated ratios.
A substantial minority of the population — especially of East Asian ancestry — carries a homozygous deletion of the UGT2B17 gene, the primary enzyme that glucuronidates testosterone for urinary excretion. These individuals excrete far less testosterone glucuronide at baseline and, critically, their T/E ratio rises much less after a testosterone dose that would flag a UGT2B17-positive athlete — the same dose can leave a null-genotype athlete's T/E comfortably under 4.
This genetic confound is precisely why T/E screening alone is now treated only as a triage step, never a standalone violation. A T/E > 4 result is not itself an Adverse Analytical Finding — it is a mandatory trigger for confirmatory IRMS testing (Stage 3) to determine whether the elevated ratio reflects exogenous administration or natural physiology/genotype.
Because UGT2B17-null athletes can dope with testosterone while staying under T/E = 4, WADA labs now perform IRMS confirmation on any sample with T/E > 4 — and, since 2014, run steroidal-module Bayesian passport screening regardless of any single T/E value, closing the genotype loophole.
Gas Chromatography–Combustion–Isotope Ratio Mass Spectrometry is the definitive confirmatory test for exogenous testosterone. It exploits a subtle but robust biochemical fact: pharmaceutical testosterone is synthesized from plant sterols (soy sitosterol, wild yam diosgenin) via industrial C3-photosynthesis-pathway precursors, which are measurably depleted in the heavy carbon isotope ¹³C relative to the human body's own steroid pool.
Carbon exists naturally as two stable isotopes: ¹²C (~98.9%) and ¹³C (~1.1%). Plants using the C3 photosynthetic pathway (soy, yam) discriminate against the heavier ¹³C isotope during CO₂ fixation more strongly than the human body does when synthesizing cholesterol-derived steroids from dietary precursors — leaving industrially-derived pharmaceutical testosterone measurably more ¹³C-depleted (more negative δ¹³C) than an athlete's own endogenous testosterone.
δ¹³C is reported in per-mil (‰) deviation from the Vienna Pee Dee Belemnite (VPDB) international carbon isotope standard:
δ¹³C = [(¹³C/¹²C)sample / (¹³C/¹²C)VPDB − 1] × 1000
Endogenous human urinary steroid metabolites typically cluster around δ¹³C ≈ −20 to −24‰; pharmaceutical testosterone and nandrolone products cluster around −29 to −32‰ because of their plant-sterol synthetic origin. This gap of several per-mil is small in absolute terms but is trivially resolved by IRMS instrumentation, which routinely achieves precision better than 0.3‰.
A single δ¹³C measurement of testosterone alone is not sufficient — individual diet (proportion of C3 vs. C4 plant intake — e.g., corn and sugarcane are C4 plants with a very different, less-depleted baseline δ¹³C) shifts everyone's baseline steroid isotope signature. The critical comparison is therefore differential, not absolute:
Δδ¹³C = δ¹³C(target metabolite: testosterone, androsterone, etiocholanolone) − δ¹³C(endogenous reference compound, ERC — typically pregnanediol or 11-ketoetiocholanolone, steroids unaffected by testosterone doping and sharing the same dietary/metabolic carbon background)
Because the ERC is drawn from the same urine sample and reflects the same dietary carbon background, subtracting it out cancels the diet confound. A Δδ¹³C exceeding the WADA technical document (TD2021IRMS) threshold of approximately 3‰ between a target androgen metabolite and the ERC is interpreted as compelling evidence of an exogenous source — chemically indistinguishable from the body's own testosterone by immunoassay, but isotopically distinct at the atomic level.
GC-C-IRMS turns an undetectable "designer" problem — synthetic testosterone is molecularly identical to endogenous testosterone — into a solvable isotopic one, because industrial steroid synthesis cannot cheaply replicate the human body's own carbon-fixation history.
Rather than testing each sample against one fixed population cutoff, the Athlete Biological Passport (ABP, introduced by WADA in 2009 for blood/hematological parameters and extended to steroids in 2014) builds an individualized longitudinal profile for every elite athlete. A Bayesian adaptive model learns each athlete's own physiological baseline across repeated tests, so it can flag deviations from that individual's normal range even when every single sample would look "clean" against a population-wide threshold.
Traditional threshold testing (T/E > 4) has a fundamental weakness: population variance is large, so a threshold loose enough to avoid false-positives in unusual-but-clean athletes is also loose enough to miss micro-dosing or genetically buffered doping (Stage 2's UGT2B17 problem). The ABP instead asks: is this athlete's current result consistent with THIS athlete's own history?
Each athlete accumulates a longitudinal profile of blood (hematological module: hemoglobin mass, reticulocytes, OFF-score — targets EPO/blood doping) and urine steroid (steroidal module) values over a competitive career, collected via unannounced out-of-competition testing. A Bayesian model — starting from population priors and updated with each new athlete-specific measurement — computes an individual's expected range for each marker, narrowing (increasing statistical power) as more data points accumulate.
The ABP's statistical engine (ADAMS software with the ABP Bayesian module, developed initially by Sottas & Saugy) models each biomarker as following a physiologically bounded distribution, updated sequentially:
P(x_new | history) computed from a prior (population distribution, adjusted for age/sex/genotype where relevant) combined with the athlete's own prior test results via Bayes' theorem, yielding a 99% or 99.9% specificity interval unique to that athlete.
• A new result falling outside this interval — but not yet reviewed by an expert panel — becomes an Atypical Passport Finding (APF), triggering targeted re-testing and expert evaluation. • If a panel of three independent experts unanimously concludes doping is the most likely explanation (beyond a reasonable doubt standard, considering all confounds: illness, dehydration, altitude, genetic variants), it becomes an Adverse Passport Finding (APF confirmed) — sanctionable under Article 2.2 (Use) even without ever recording a single Adverse Analytical Finding on any individual sample.
This is the single biggest conceptual shift in modern anti-doping: proof no longer requires catching one "hot" sample — a pattern across many clean-looking samples can itself constitute the evidence.
Lance Armstrong-era cycling cases and numerous track-and-field bans since 2015 have relied substantially on ABP longitudinal patterns rather than a single positive test — closing a detection gap that direct-analyte testing structurally cannot cover on its own.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Testosterone/Epitestosterone (T/E) | Primary composite marker | Individual reference range vs. population T/E > 4 | Sensitive to exogenous T, resistant to single-sample dilution |
| Androsterone (A) / Etiocholanolone (Etio) | 5α/5β-reduced androgen metabolites | Absolute concentration + ratios tracked over time | Flags DHEA and androstenedione precursor doping |
| 5α-Androstanediol / 5β-Androstanediol | Minor urinary metabolites | Ratio sensitive to 5α-reductase activity shifts | Detects some designer steroids and dihydrotestosterone use |
| ABPI (Atypical Passport Index) | Composite abnormality score | Combined Bayesian likelihood across all 7 markers | Single interpretable score for expert panel review |
A flagged result — whether an Adverse Analytical Finding from IRMS confirmation or an Adverse Passport Finding from ABP longitudinal review — moves into WADA Code results-management: a formal legal process combining the scientific evidence with expert panel review, athlete right-of-reply (opening the B sample), and, if upheld, a defined sanction under the World Anti-Doping Code.
A positive A-sample screening result does not itself constitute a doping violation. The WADA Code results-management pathway requires:
1. Confirmation procedure: the A sample is re-analyzed by the laboratory using confirmatory methods (GC-C-IRMS for steroids, LC-MS/MS for peptide hormones) to rule out screening-assay artifacts. 2. Notification: the athlete's National Anti-Doping Organization or International Federation is notified; the athlete is informed and has the right to request the B sample be opened and independently analyzed, often with their own expert witness present. 3. B-sample analysis: performed by the same lab (different analyst) on the sealed B bottle. Confirmation of the A-sample finding is required before a formal Adverse Analytical Finding is declared. 4. Provisional suspension: commonly imposed immediately upon a confirmed AAF for non-Specified Substances, pending full hearing. 5. Results management hearing: an independent panel (e.g., Court of Arbitration for Sport, CAS, or national anti-doping tribunal) reviews the full scientific record, chain of custody, and any explanation offered by the athlete (therapeutic use exemption, contamination, ABP confound).
WADA Code Article 2.2 defines an anti-doping rule violation as "Use or Attempted Use by an Athlete of a Prohibited Substance or a Prohibited Method" — critically, this violation category does not require a positive sample at all. Evidence can include: ABP Adverse Passport Findings, admissions, witness testimony, non-analytical evidence (e.g., seized coach records, biological data from an athlete's own wearable devices), or any reliable means consistent with Article 3.2 evidentiary rules.
The standard of proof is "comfortable satisfaction of the hearing panel" — a standard higher than civil balance-of-probabilities but lower than criminal beyond-reasonable-doubt. For ABP-based cases, an independent panel of three experts must unanimously conclude that the most likely explanation for the abnormal profile is doping, having excluded plausible pathological or physiological explanations.
Sanctions under the 2021 WADA Code: • Standard: 4 years for intentional use of a non-Specified Substance (testosterone/AAS are non-Specified — Category S1). • Reduced: down to 2 years if the athlete establishes the violation was not intentional; down further (as low as a reprimand) for a Substance of Abuse used out-of-competition with no performance-enhancing intent. • Aggravated: up to lifetime ban for participation in a doping conspiracy or being a repeat offender.
Because ABP evidence and Article 2.2 do not require ever catching a single positive sample, an athlete who successfully times exogenous testosterone around competition windows to always test "clean" on absolute thresholds can still be sanctioned purely from the shape of their longitudinal biomarker trajectory — the central design goal of the passport system.