Urinary progesterone metabolites as non-invasive biomarkers: pregnanediol and pregnanetriol

Project: L7-5
Stage: REPORT_EN
Date: 2026-08-09
Classification: progesterone and neurosteroid metabolomics
Inference population: non-pregnant women aged 35–55 years, classified by STRAW+10; woman and cycle as units
Status: complete scientific report for QA; not a clinical or prescriptive guide
Overall maturity: H1 for biochemical links and the narrow temporal validity of PdG; H0 for the integrated chain, organ function, healthspan, and longevity

1. Executive scientific abstract

Urinary pregnanediol—often measured as pregnanediol-3-glucuronide, PdG/P3G—and pregnanetriol—PT; 5β-pregnane-3α,17α,20α-triol—are not equivalent readings of one amount of “progesterone.” They are effluents from distinct pathways that converge in liver processing, conjugation, and the kidney. PdG derives mainly from 5β metabolism of progesterone (P4); PT derives from 5β metabolism of the 17-hydroxyprogesterone (17OHP) pool. In both cases, the observed urinary value convolves endocrine input, metabolic partition fraction, conjugation, lag, clearance, urine volume, and analytical identity.

The strongest human conclusion is narrow: in regular cycles with ovulation independently determined by ultrasound, a serial rise in PdG after ovulation reports recent luteinization. By itself it does not demonstrate normal follicular rupture, luteal sufficiency, integrated ovarian production, PGR signaling, brain exposure, or health benefit. For PT, human evidence shows responsiveness to the 17OHP pool in 21-hydroxylase deficiency and after ACTH, together with historical variation by cycle phase; it does not quantify how much derives from adrenal, ovarian, or peripheral conversion during natural cycles.

The central mechanism separates input, partition, and function. P4 can signal through PGR, enter the 5α branch to 5α-DHP/allopregnanolone, or enter the AKR1D1→AKR1C 5β branch ending in pregnanediol/PdG. 17OHP can proceed through CYP21A2→CYP11B1 toward cortisol or enter the 5β shunt ending in PT. High excretion can therefore reflect more precursor, more terminal clearance, or both; it is not monotonic with hormonal action. The 5β branch also includes pregnanolone, a neurosteroid capable of modulating GABA-A, so it cannot be modeled as a functionally inert sink.

The computational layer simulated 384 scenarios and 52,800 replicates. Under a common endocrine cause, daily observational series and urine fractionation frequently generated false biphasicity. Only the randomized ACTH–saline contrast aggregated over 24 hours and repeated across two cycles passed the gates under a simulated material effect and at least 24 women. This is not an empirical sample size: it depends on the generator and assumed effect. Adversarial review further showed that ACTH is not a clean anatomical instrument because it can change LH/FSH and the 17OHP response depends on phase. A positive challenge demonstrates ACTH perturbability of the system, not an adrenal percentage.

There is no direct human evidence that PdG, PT, or their combination predicts healthspan, mortality, or longevity. For bone, the adjusted analysis of 643 SWAN women was null for uPdG; for sleep duration, a study of 106 women was also null for PdG. The longevity chain breaks before the effector tissue. The decisive next program separates three questions: M0 validates identity and urinary mass; E1 tests ACTH perturbability of PT; D0 tests 5β flux control by AKR1D1 using tracers and rescue. None validates health or longevity.

Scientific delta of this report. The minimum identifiable unit is not the raw PdG+PT panel, but a three-component vector measured separately: (1) precursor integral—P4 AUC or 17OHP AUC—(2) chemically resolved 24-hour urinary product mass, and (3) product response to a controlled perturbation. Precursor informs input; the molar product/precursor ratio informs partition only after mass balance; and the ACTH–saline difference informs perturbability. Collapsing these three estimands is the principal source of overinterpretation.

2. Scientific question and relevance

Question

What mechanisms connect urinary pregnanediol/PdG and pregnanetriol with women’s health and longevity, what can they measure non-invasively, and which new hypotheses can be falsified?

Short answer

Their demonstrated connection is to metabolized and excreted endocrine dynamics, not longevity. PdG can report recent luteinization; PT can report part of the 17OHP pool and respond to ACTH. Both are separated from the effector tissue by hepatic partitioning, conjugation, clearance, and parallel functional branches. Their possible future value would be to decompose steroid physiology or pharmacology if they add information beyond serum precursors. There is currently no basis for calling them biomarkers of systemic health or longevity.

Why it matters

Urine supports dense series with less invasiveness than blood, but this temporal advantage tempts investigators to turn a stable curve into a causal measure. Effluent stability does not remove source ambiguity. During the menopause transition, ovulation, P4, E2, 17OHP, HPA activity, hepatic metabolism, and renal function may move together. Without decomposition, an apparently informative panel may summarize phase or matrix only.

3. Scope, population, and life stage

The primary population is non-pregnant women aged 35–55 years with ovaries present, classified by STRAW+10 as late reproductive (−3b/−3a), early transition (−2), late transition (−1), or early postmenopause (+1a/+1b). The priority contrast is longitudinal and within-woman. Woman and cycle—not each sample—are the units of inference.

The following remain separate:

• cycles with independently confirmed ovulation, confirmed non-ovulatory cycles, and postmenopausal states;
• spontaneous cycles and exposure to exogenous progesterone, progestins, estrogens, or glucocorticoids;
• physiology, pregnancy, puerperium, CAH/21OHD, IVF, oophorectomy, and adrenalectomy;
• reproductive stage and chronological age;
• intact PdG, hydrolyzed total pregnanediol, free/total PT, and intact PT-3G.

Pregnancy, CAH, and source-negative states may provide pathway controls, but not effect sizes, thresholds, or performance transportable to spontaneous transition. Late postmenopause and age >55 years are reserved for later transportability.

If health outcomes are ever opened, they must be restricted to one prespecified proximal domain. Sleep, neurophysiology, bone, CVD, cancer, fertility, mortality, and longevity cannot be fused into a score.

4. Background and mechanism map (mapa de mecanismo)

4.1 Chemical identity

Reference audit identified an inherited error: the correct DOI for the Vogg et al. joint urinary pregnanediol/PT method is 10.1016/j.jmsacl.2022.07.006 (PMID 35910411). DOI 10.1016/j.jmsacl.2022.06.001 belongs to a remdesivir method and does not support steroid metrology.

4.2 Causal mechanism map

RESERVE / HPO                                      HPA / ACTH
AMH–FSH–LH → ovulation → corpus luteum             MC2R–MRAP–cAMP/PKA
                  │                                      │
                  ▼                                      ▼
 cholesterol–StAR–CYP11A1–HSD3B2 → P4 ──CYP17A1──→ 17OHP
                  │                                      │
       ┌──────────┼───────────────┐                      ├─CYP21A2→11-deoxycortisol→CYP11B1→cortisol
       │          │               │                      └─AKR1D1→3α/20α reduction→PT
       ▼          ▼               ▼                                               │
      PGR      SRD5A→5α-DHP   AKR1D1→5β-DHP                                      ▼
 target tissue     │               │                                         PT/PT-3G
                   ▼               ├→ pregnanolone → GABA-A                      │
            allopregnanolone       └→ pregnanediol→PdG                           ▼
                   │                               │                         excretion
                 GABA-A                            ▼
                                            renal excretion

observed = input × partition fraction × conjugation × renal kernel
           ÷ volume + identity/method error

4.3 Three levels that must not collapse

1. Input: P4 and 17OHP derive from ovulatory-luteal architecture, HPO, HPA, and producing tissues. A point concentration is not an AUC.
2. Partition: AKR1D1 commits Δ4 steroids to the 5β route; AKR1C1–4, conjugation, and transport determine downstream products. The physiological in-vivo fraction is not quantified.
3. Function: PGR, allopregnanolone, pregnanolone, GABA-A, and cortisol are parallel branches or functional outputs. A terminal urinary metabolite does not measure them directly.

4.4 Conceptual equations

U_PdG(t) = ∫ P4(s) × f5β(s) × f3α20α(s) × fconj(s) × Krenal(t−s) ds
           ---------------------------------------------------------------- + error
                                urine volume(t)

U_PT(t)  = ∫ [17OHP_adrenal(s)+17OHP_ovarian(s)+17OHP_peripheral(s)]
             × f5β→PT(s) × fconj(s) × Krenal(t−s) ds
           ---------------------------------------------------------------- + error
                                urine volume(t)

The same urinary value can therefore arise from different biological states. High PdG may reflect high P4 or a high 5β fraction. High PT may coexist with reduced functional output toward cortisol, as 21OHD illustrates. Excreted molecules retain no anatomical label.

5. Evidence method

The scoping, evidence-map, verification, mechanistic-synthesis, compute decision/results, hypothesis-generation, adversarial-review, and experiment-design artifacts were integrated completely. The search covered human, biochemical, and metrological literature through August 9, 2026. PubMed-indexed primary studies, full text where available, and DOI/PMID metadata were prioritized. Recent or uncertain claims were revalidated in PubMed/NCBI and scientific search; Crossref DOI auditing produced the Vogg et al. correction.

For each decisive source, design, population, stage, species/system, matrix, sample size, comparator, outcome, effect, and limitations were distinguished. Evidence was classified as:

• direct human: physiology or intervention in women with relevant analytes;
• human pathology/perturbation: demonstrates pathway capacity, with limited transport;
• in-vitro/recombinant: demonstrates catalysis or receptor pharmacology, not in-vivo contribution;
• computational: evaluates identifiability under assumptions, not biological existence;
• inferred: a compatible chain not yet observed end to end.

Null and contradictory findings were sought. Using PdG as both predictor and reference standard, transporting internal fertility thresholds to perimenopause, converting between-person correlation into within-woman coupling, and treating a ratio as a substitute for its components were rejected.

6. Evidence map

6.1 Decisive evidence

6.2 What is established and what is not

Established with moderate confidence: serial PdG reports a postovulatory transition in regular cycles; PT can follow 17OHP in a pathological range and respond to ACTH; AKR1D1/AKR1C can form 5β products; identity and method alter the construct.

Plausible but unidentified: the natural organ fraction of PT; whether product/precursor yield reflects physiological clearance variation; whether intact and deconjugated species preserve ordering; whether urine adds information beyond serum precursors.

Not demonstrated: PdG/PT as measures of PGR, GABA-A, cortisol, luteal sufficiency, bone function, sleep, systemic health, healthspan, mortality, or longevity.

7. Contradictory evidence and null findings

1. PdG is temporal but not sufficient. Its rise after ultrasound ovulation is reproducible, but modern thresholds were selected within cohorts and may appear with luteinization without normal follicular rupture. Temporal accuracy does not demonstrate functional sufficiency.
2. Menopause-transition decline but circular classification. O’Connor observed lower PdG across transition; the anovulation algorithm uses PdG and cannot validate it independently.
3. PT is connected to 17OHP but not specific. The association is strong in 21OHD; PT overlaps in neonates with and without 21OHD and is not exclusive to CYP21A2 or adrenal origin.
4. ACTH perturbability but pleiotropy. ACTH can raise 17OHP/PT, gonadotropins, and other steroids; response depends on phase. The contrast identifies total challenge effect, not anatomy.
5. Terminal 5β branch but not an inert branch. Pregnanediol is terminal; upstream pregnanolone modulates GABA-A. More 5β cannot be assumed to mean less neuroactivity.
6. Direct nulls for PdG and health. Adjusted uPdG was null for BMD in SWAN; PdG was null for sleep duration. An association of ovulatory frequency with bone loss did not measure PdG/PT and is less causally proximal.
7. Joint method exists; joint physiology does not. Vogg et al. validated deconjugated pregnanediol/PT with precision <10.1%, but in adrenal tumors; Leoni measured intact PdG, not the complete physiological panel.

These tensions favor a common HPO–HPA cause plus matrix/clearance as the working explanation. They do not prove that every signal is noise; they require orthogonalization.

8. Multiscale mechanistic synthesis

Molecular and enzymatic mechanism

P4 and 17OHP are Δ4 steroids. AKR1D1 catalyzes 5β reduction and changes A-ring geometry; AKR1C completes 3α/20α reductions. Conjugation increases solubility and renal excretion generates the observable analytes. Human recombinant structure and kinetics support this route. Whether it controls flux in female human hepatocytes or in vivo remains unproven.

Cellular and tissue mechanism

The corpus luteum produces P4; adrenal, ovary, and peripheral conversion can feed 17OHP. The liver processes precursors regardless of origin. AKR1D1 also participates in bile-acid metabolism and GR/PXR/FXR signaling; cellular perturbation can reprogram hepatocytes. Catalytic loss–rescue, early timing, and bile-acid controls are therefore essential.

Systemic mechanism

HPO controls ovulation, luteal mass/duration, and P4. HPA/ACTH alters adrenal steroidogenesis but is coupled to HPO. Renal function, urine volume, time, liver function, and phase affect urine. Observed dynamics contain no natural intervention that separates all pathways.

Effector-organ mechanism

The plausible neural bridge is P4→5α-DHP/Allo and P4→5β-DHP/pregnanolone toward GABA-A, not PdG→GABA-A. The plausible bone bridge requires functional human PGR and control for E2/FSH/HPO; uPdG was not independent in SWAN. Any PT bridge to stress, immunity, or metabolism must directly measure cortisol/androgens because high PT may reflect diversion away from cortisol.

Life course

The required chain is:

validated source/input → quantified partition → tissue exposure
→ proximal engagement/function → persistence in the same domain
→ years free of deterioration with competing risks

Evidence stops before tissue exposure. Longevity is not yet an interpretable outcome.

9. Computational layer

Objective and data

Computation tested whether two inputs into PT could be identified without manufacturing biphasicity. QUIDEL (kml3d 2.5.0) was used only to calibrate missingness and variation: the real object contained 80 women × 30 days and 5.17% missingness, not the documented 107 × 49. It contains no PT, 17OHP, or ACTH. No private or current Lua data were used.

Simulation

There were 384 scenarios and 52,800 replicates spanning common cause or two inputs, HPA–HPO correlation, analytical CV, one/two cycles, MCAR/MNAR, correct or long-tail kernel, and intact/discordant bridge species. The gate required low false biphasicity, coverage, low bias, and recovery above a 0.15 standardized log margin.

Result

• Daily observation: 34.3–99.0% false biphasicity under confirmatory common-cause scenarios.
• Fractionated challenge: 7.7–90.7%; sensitive to kernel and method.
• 24 h aggregated ACTH–saline: 0–2.67%; go only with 24 or 32 women, two cycles, and simulated effect 0.30.

More density or larger n sometimes increased certainty around a confounded model. Integrating 24 h was robust to the simulated kernel because it preserved mass, but incomplete collection and all ACTH-induced pleiotropic pathways were not modeled.

Correct interpretation

The result eliminates daily observational series and fractionation as primary tests of dual source. It prioritizes 0–24 h mass and uses fractions to learn carryover/kernel. It does not validate biology, set n=24, or locate anatomy. Adversarial review requires sample-size re-estimation by precision after the M0/E1 pilot.

10. Primary hypothesis

L7-5-AR-H1A v2 — PT has a reproducible ACTH-perturbable response

Falsifiable statement. In women with regular spontaneous cycles and early follicular phase, low-dose ACTH versus saline will materially and reproducibly increase PT mass over 0–24 h, provided the challenge produces a valid serum 17OHP/cortisol response and collection/identity pass M0.

Mechanism. ACTH activates MC2R/MRAP and expands the steroid pool; a fraction of 17OHP enters the 5β/3α/20α route and is excreted as PT. The saline contrast estimates total perturbable effect, not an adrenal fraction, because ACTH can move HPO, other steroids, and clearance.

Predictions. (1) ΔPT_ACTH−saline > δ_mat; (2) concordant direction in two periods/cycles; (3) effect follows the 17OHP/cortisol response; (4) it is not explained only by volume, recovery, or proportional change in every steroid; (5) ordering is preserved across analytically bridged species.

Evidence for. Pinzani observed PT-3G/17OHP after ACTH; Itonaga showed PT–17OHP coupling in 21OHD; the randomized 24 h design had less false biphasicity in simulation.

Evidence against. ACTH–PT evidence is historical, small, and immunoassay-based; ACTH alters multiple steroids and gonadotropins; modern physiological effect size is unknown; phase modifies 17OHP.

Kill criteria. Valid serum challenge + valid mass + CI90% within ±δ_mat; failure to replicate or sign reversal; effect attributable to collection/method; or no gain of PT over the parallel serum/urinary panel.

Status and maturity. proposed_narrowed; H1 for parent links, H0 for the integrated chain; confidence 0.38.

Auxiliary mechanism to test separately. L7-5-AR-H3 v3 proposes that AKR1D1 redistributes tracer among 5β products, 5α products, and precursor. It is necessary to interpret partition, but a positive D0 does not rescue a negative E1 or determine neurofunctional direction.

11. Competing hypothesis

L7-5-AR-H2 v2 — common endocrine cause, matrix, and incremental null

Falsifiable statement. P4 AUC, 17OHP/cortisol, HPO/HPA, phase, volume, liver/renal function, and analytical identity will explain the PdG/PT pattern; PT will provide no specific, reproducible, or incremental response after those components are measured.

Mechanism. Reserve/stage, phase, and systemic state move precursors, gonadotropins, and metabolism simultaneously. Liver, kidney, and matrix add covariation. A curve or ratio can appear source-specific while arising from an omitted common cause.

Predictions. (1) The serum-reference plus urinary-mass model equals a model that adds PT; (2) product|precursor residuals have low between-cycle ICC; (3) gain concentrates between people, in pathological range, or within one method; (4) ACTH effect appears proportionally across several 5β steroids; (5) PdG:PT is unstable and denominator-dependent.

Evidence for. No joint validation; uPdG–BMD and PdG–sleep nulls; computational false biphasicity; direct ACTH–HPO coupling; route, hydrolysis, and lag dependence.

Evidence against. PdG preserves postovulatory timing across methods; PT may respond to ACTH and couple to 17OHP. A replicated PT-specific response would refute the sufficient version of common cause.

Kill criteria. Second laboratory + replicated E1 + material PT-specific response + out-of-sample error/calibration improvement over references + between-cycle stability. Significance alone does not kill H2.

Status and maturity. strengthened_and_testable; H1 for metrological risk and H0 for causal equivalence; confidence 0.72. This is the current leading explanation.

12. Translational hypothesis

L7-5-AR-HT1 v3 — a PdG residual adds GABA-A function beyond direct neurosteroids

Falsifiable statement. Only after M0, D0, and between-cycle reliability are validated, the residual PdG mass | P4 AUC will add out-of-sample prediction of one prespecified GABA-A function beyond directly measured P4, E2, ovulation, stage, allopregnanolone, and pregnanolone; PT residual will serve as a shared-clearance control.

Mechanism. A conditioned urinary yield might contain systemic partition information absent from point P4. Allo and pregnanolone—not PdG—would be plausible mediators. The residual has value only if it outperforms direct measurements.

Predictions. Adequate ICC/CCC; stable direction across cycles; residual improves error/calibration over direct neurosteroids; effect attenuates after mediators are included; PT does not reproduce the neural pattern if it controls clearance only.

Evidence for. Distinct 5α/5β branches and the possibility that dynamics add information. Evidence against: no direct human evidence; urine does not represent brain; the 5β branch includes pregnanolone; PdG was null for sleep duration; the residual may be error-dominated.

Kill criteria. Failure of M0 or D0, ICC <0.60, sign reversal, no external improvement over Allo/pregnanolone, or explanation by phase/E2/HPA/baseline.

Status and maturity. parked_translational; H0; confidence 0.02; HUMAN_QA_REQUIRED. Even if supported, it would validate a proximal function, not healthspan or longevity.

13. Falsifiable predictions and kill criteria

14. Discriminating experiment

14.1 M0 — identity and mass

Question: Can declared species be measured and a 24 h mass closed?
Assay: LC-MS/MS/HRMS with authentic/isotopic standards; intact PdG and PT-3G; free and total pregnanediol/PT as separate constructs; pregnanetriolone; hydrolysis, matrix, recovery, stability, and carryover controls.
Samples: 40 independent matrices, six QC pools; 12 volunteers aged 35–45 for two repeated 24 h collections; 20 blinded aliquots in a second laboratory.
Gates: CV ≤10% low and ≤7.5% medium/high; recovery 85–115%; ≥90% detectable; completeness ≥90%; fractionated mass balance 90–110%; second-laboratory CCC ≥0.80. If no adequate PT-3G standard exists, it cannot be primary.

14.2 E1 — follicular ACTH–saline

Question: Is PT ACTH-perturbable?
Design: randomized blinded AB/BA crossover, research ACTH 1 μg IV versus saline on days 2–5; two periods/cycles; mandatory ethics review and clinical supervision.
Blood: 0/30/60/90 min for cortisol/17OHP; P4, E2, LH, FSH, androstenedione, and DHEA.
Urine: baseline −24–0 h and post 0–24 h; fractions for kernel/carryover.
Estimand: paired ACTH minus saline difference in log(post/baseline mass).
Size: blinded n=12 pilot for SD and completeness; δ_mat=max[log(1.20),2×CV_log M0]; recalculate for CI95% half-width ≤0.15 log; 20–36 analyzable, not automatic n=24. If >36 are required, declare low value/infeasibility.
Kill: CI90% within ±δ_mat with valid challenge/collection or failure to reproduce.

14.3 D0 — 5β partition

Question: Does AKR1D1 control labeled yield?
Sequence: recombinant protein (D0a), isogenic loss–rescue (D0b), and replication in primary female hepatocytes (D0c, six discovery and up to 12).
Inputs: isotopic P4 and 17OHP separately.
Panel: precursor, 5β-DHP, 17OH-pregnanolone, pregnanolone, pregnanediol, PT, conjugates; 5α-DHP, allopregnanolone/epimers; ATP, NADPH, viability, hepatocyte function, AKR1C/SRD5A, and bile acids/GR–PXR–FXR.
Gate: δ_flux=max(20%,3×technical CV); two concordant perturbations; competent rescue restores ≥50% of lost flux; inactive rescue fails; ≥80% tracer recovery; effect in ≥9/12 donors or interval outside equivalence.
Kill: target engagement without flux change, no rescue, broken balance, or late effect explained by pleiotropy.

14.4 Sequence

M0 → {E1, D0} → external replication → E2 phase modification
   → E3 source-negative only if feasible → T0 proximal function
   → organ-specific persistence → healthspan

E1 and D0 answer independent questions. A positive result in one does not rescue the other. Dose is not escalated and post-hoc subgroups are not searched if E1 is equivalent to zero.

15. Biomarkers and stratification

Candidate status

Mandatory stratification

• STRAW+10, age, time since FMP, and regularity;
• independent ovulation, phase, cycle length, and architecture;
• P4 AUC, E2, LH/FSH, 17OHP, ACTH/cortisol;
• time, volume, completeness, creatinine, and specific gravity;
• eGFR and liver function; adiposity/steatosis;
• drugs and exogenous hormones;
• method, conjugation/hydrolysis, laboratory, batch, and stability;
• ancestry and population, without default transport of European/young thresholds to LATAM.

Absolute quantities precede ratios. Creatinine corrects concentration, not a missed void or muscle mass. A spot sample does not represent a cycle.

16. Individual variability

Variability has at least six sources:

1. Reproductive architecture: ovulation, corpus-luteum mass/duration/perfusion, and LH support.
2. HPA–HPO coupling: ACTH, cortisol, gonadotropins, and phase-dependent response.
3. Hepatic metabolism: AKR1D1/AKR1C, conjugation, redox state, bile acids, steatosis, and drugs.
4. Renal/matrix clearance: eGFR, volume, creatinine, specific gravity, and completeness.
5. Metrology: intact/deconjugated species, hydrolysis, standard, coelution, batch, and laboratory.
6. Effector tissue: receptor, local metabolism, and neurosteroids that may decouple from blood/urine.

Between-woman stability does not substitute for within-woman repeatability. A longitudinal biomarker needs ICC/repeatability by cycle and stage, not only high cross-sectional correlation.

17. Pharma relevance and maturity

The nearest opportunity is a research-use-only assay for steroid systems pharmacology, not a diagnosis, companion diagnostic, or validated therapeutic target.

There is no validated context of use, clinical utility, net benefit, external replication, or partnering readiness. Every Pharma advance requires human QA. No compound, therapeutic dose, or efficacy trial is proposed.

18. Limitations

1. Physiological PT literature is old, small, and immunoassay-dependent; there is no complete modern serial panel.
2. The strongest PdG evidence comes from young women with regular cycles and fertility technologies; thresholds do not transport to perimenopause.
3. 21OHD demonstrates pathway capacity in an extreme range, not physiological sensitivity.
4. No in-vivo human isotope flux quantifies the 5β fraction of P4/17OHP.
5. AKR1D1 is pleiotropic and the hepatocyte does not reproduce kidney, perfusion, microbiota, or target tissue.
6. ACTH identifies a total effect, not anatomical source; phase is not an ovarian label.
7. Computation depends on its generator and simulated effects; its rates are not physiological properties.
8. A 24 h collection improves mass estimation but can fail through omission/carryover; not all such errors were simulated.
9. Isotopic standards were not located for every conjugate; a species without an adequate standard is an informative no-go.
10. BMD and sleep nulls constrain those bridges but do not prove universal absence in every organ.
11. There is no direct evidence on cognition, cardiometabolism, frailty, mortality, or longevity.
12. This is a deep adversarial review, not a registered systematic review; “not located” does not mean nonexistent.
13. No current Lua data, private data, PHI/PII, or private genomes were used.

19. Conclusions

PdG and PT are scientifically useful only when urine is treated as the end of a chain, not a transparent window into ovary, adrenal, brain, or longevity. PdG retains evidence as a delayed reporter of recent luteinization; PT retains evidence as a 17OHP effluent perturbable by ACTH in limited contexts. Neither measures hormonal action monotonically.

The most defensible inference decomposes integrated precursor, product mass, and perturbational response. A point value or ratio cannot separate input, partition, and clearance. ACTH–saline can test perturbability, not anatomy. AKR1D1 may control 5β catalysis, but only tracer, rescue, and mass balance can demonstrate flux control. Pregnanolone requires the neurofunctional direction to remain open.

The leading explanation is a common HPO–HPA cause plus matrix/clearance, with higher confidence than the incremental-signal hypothesis. The program remains valuable precisely because it can refute that explanation: M0 can kill the measurement, E1 can kill perturbability, and D0 can kill partition. Until those gates and a tissue bridge are passed, the correct term is candidate endocrine/metrological reporter, not a health or longevity biomarker.

Claims added by the report

• L7-5-REPORT-EN-C01: The panel’s minimum identifiable unit is precursor AUC + 24 h product mass + perturbational contrast; none substitutes for the others.
• L7-5-REPORT-EN-C02: PdG/PT are effluents of input, partition, and clearance; their concentration has no guaranteed monotonic relationship with PGR, GABA-A, or cortisol.
• L7-5-REPORT-EN-C03: ACTH–saline identifies system perturbability only; anatomical source remains unlocalized because of HPA–HPO coupling and phase dependence.
• L7-5-REPORT-EN-C04: The chain to longevity is broken before tissue exposure; PdG–BMD and PdG–sleep nulls preclude a pan-systemic narrative.
• L7-5-REPORT-EN-C05: The correct DOI for Vogg et al. 2022 is 10.1016/j.jmsacl.2022.07.006; the inherited incorrect DOI is rejected.

20. References

1. Roos J, Johnson S, Weddell S, et al. Monitoring the menstrual cycle: comparison of urinary and serum reproductive hormones referenced to true ovulation. Eur J Contracept Reprod Health Care. 2015;20:438–450. DOI: 10.3109/13625187.2015.1048331. PMID: 26018113.
2. Bouchard TP, et al. Validating ovulation prediction and confirmation with the Mira monitor: blinded ultrasound and serum hormone comparison. Reprod Biomed Online. 2026;53:105881. DOI: 10.1016/j.rbmo.2026.105881.
3. Handelsman DJ, et al. Direct measurement of pregnanediol 3-glucuronide (PDG) in dried urine spots by liquid chromatography-mass spectrometry to detect ovulation. J Steroid Biochem Mol Biol. 2021;212:105900. DOI: 10.1016/j.jsbmb.2021.105900. PMID: 33872762.
4. Gifford RM, Reynolds RM, Ghaffari P, et al. Confirmation of ovulation from urinary progesterone analysis: assessment of two automated assay platforms. Sci Rep. 2018;8:17621. DOI: 10.1038/s41598-018-36051-6. PMID: 30514870.
5. Kassam A, Overstreet JW, Snow-Harter C, et al. Identification of anovulation and transient luteal function using a urinary pregnanediol-3-glucuronide ratio algorithm. Environ Health Perspect. 1996;104:408–413. DOI: 10.1289/ehp.96104408. PMID: 8732951.
6. O’Connor KA, Ferrell R, Brindle E, et al. Progesterone and ovulation across stages of the transition to menopause. Menopause. 2009;16:1178–1187. DOI: 10.1097/gme.0b013e3181aa192d. PMID: 19568209.
7. O’Connor KA, Brindle E, Miller RC, et al. Ovulation detection methods for urinary hormones: precision, daily and intermittent sampling and a combined hierarchical method. Hum Reprod. 2006. DOI: 10.1093/humrep/dei497.
8. Abdullah S, et al. Distinct urinary progesterone metabolite profiles during the luteal phase. Horm Mol Biol Clin Investig. 2023;44. DOI: 10.1515/hmbci-2022-0065. PMID: 36578213.
9. Stanczyk FZ, et al. Urinary progesterone and pregnanediol: use for monitoring progesterone treatment. J Reprod Med. 1997;42. PMID: 9131494.
10. Newman M, Pratt SM, Curran DA, Stanczyk FZ. Evaluating urinary estrogen and progesterone metabolites using dried filter paper samples and GC-MS/MS. BMC Chem. 2019;13:20. DOI: 10.1186/s13065-019-0539-1. PMID: 31384769.
11. Leoni L, et al. Rapid UHPLC-MS/MS measurement of pregnanediol 3-glucuronide in spot urine samples for detecting ovulation. Biomed Chromatogr. 2024;38:e5982. DOI: 10.1002/bmc.5982. PMID: 39149929.
12. Bedolla-Tovar N, et al. Prediction and detection of ovulation by measurement of urinary pregnanetriol-3α-glucuronide: a multicentre study. J Steroid Biochem. 1986;24. DOI: 10.1016/0022-4731(86)90455-3. PMID: 3702465.
13. Pinzani P, Messeri G, Serio M. A chemiluminescent method for measurement of pregnanetriol-3α-glucuronide in diluted human urine. J Biolumin Chemilumin. 1989;4. DOI: 10.1002/bio.1170040176.
14. Itonaga T, Hasegawa Y, Sano S, et al. First morning pregnanetriol and 17-hydroxyprogesterone correlated significantly in 21-hydroxylase deficiency. Front Endocrinol. 2022;12:808254. DOI: 10.3389/fendo.2021.808254. PMID: 35140686.
15. Hughes IA, Winter JSD. Relationships between serum 17OH-progesterone and other serum and urinary steroids in congenital adrenal hyperplasia. J Clin Endocrinol Metab. 1978;46:98–104. DOI: 10.1210/jcem-46-1-98. PMID: 752027.
16. Homma K, Hasegawa T, Nagai T, et al. Elevated urine pregnanetriolone definitively establishes classical 21-hydroxylase deficiency in term and preterm neonates. J Clin Endocrinol Metab. 2004;89. DOI: 10.1210/jc.2004-0473. PMID: 15579762.
17. Ackermann D, Groessl M, Pruijm M, et al. Reference intervals for the urinary steroid metabolome: impact of sex, age, day and night time on adult steroidogenesis. PLoS One. 2019;14:e0214549; correction 2021. DOI: 10.1371/journal.pone.0214549; correction 10.1371/journal.pone.0253678. PMID: 30925175.
18. Vogg N, Mueller T, Floren A, et al. Targeted metabolic profiling of urinary steroids with a focus on analytical accuracy and sample stability. J Mass Spectrom Adv Clin Lab. 2022;25:44–52. DOI: 10.1016/j.jmsacl.2022.07.006. PMID: 35910411.
19. Di Costanzo L, Drury JE, Penning TM, Christianson DW. Crystal structure of human liver Δ4-3-ketosteroid 5β-reductase (AKR1D1) and implications for substrate binding and catalysis. J Biol Chem. 2008;283:16830–16839. DOI: 10.1074/jbc.M801778200. PMID: 18407998.
20. Jin Y, Mesaros AC, Blair IA, Penning TM. Stereospecific reduction of 5β-reduced steroids by human AKR1C1–AKR1C4 in testosterone and progesterone metabolism. Biochem J. 2011;437:53–61. DOI: 10.1042/BJ20101804. PMID: 21521174.
21. Schiffer L, Barnard L, Baranowski ES, et al. Human steroid biosynthesis, metabolism and excretion are differentially reflected by serum and urine steroid metabolomes: a comprehensive review. J Steroid Biochem Mol Biol. 2019;194:105439. DOI: 10.1016/j.jsbmb.2019.105439. PMID: 31362062.
22. Aleknaviciute J, Tulen JHM, De Rijke YB, et al. Adrenocorticotropic hormone elicits gonadotropin secretion in premenopausal women. Hum Reprod. 2016;31. DOI: 10.1093/humrep/dew190. PMID: 27591239.
23. Kruyt N, Rolland R. Cortisol, 17α-OH-progesterone and androgen responses to standardized ACTH stimulation in different stages of the normal menstrual cycle. Acta Endocrinol. 1982;100:427–433. DOI: 10.1530/acta.0.1000427. PMID: 6287785.
24. Pierce SR, Bracamontes JR, Steinbach JH, Akk G. Inhibitory actions of potentiating neuroactive steroids in the human α1β3γ2L GABA-A receptor. Mol Pharmacol. 2024;106. DOI: 10.1124/molpharm.124.000960. PMID: 39214710.
25. Nikolaou N, Gathercole LL, Kirkwood L, et al. Glucocorticoids regulate AKR1D1 activity in human liver in vitro and in vivo. J Endocrinol. 2020;245. DOI: 10.1530/JOE-19-0473. PMID: 32106090.
26. Nikolaou N, Gathercole LL, Marchand L, et al. AKR1D1 is a novel regulator of metabolic phenotype in human hepatocytes and is dysregulated in non-alcoholic fatty liver disease. Metabolism. 2019;99:67–80. DOI: 10.1016/j.metabol.2019.153947. PMID: 31330134.
27. Grewal J, Sowers MR, Randolph JF Jr, et al. Low bone mineral density in the early menopausal transition: role for ovulatory function. J Clin Endocrinol Metab. 2006;91:3780–3785. DOI: 10.1210/jc.2006-0544. PMID: 16835283.
28. Touzet S, Rabilloud M, Boehringer H, et al. Relationship between sleep and secretion of gonadotropin and ovarian hormones in women with normal cycles. Fertil Steril. 2002;77:738–744. DOI: 10.1016/S0015-0282(01)03254-X. PMID: 11937126.
29. Direito A, Bailly S, Mariani A, Ecochard R. Relationships between the luteinizing hormone surge and other characteristics of the menstrual cycle in normally ovulating women. Fertil Steril. 2013;99:279–285.e3. DOI: 10.1016/j.fertnstert.2012.08.047. PMID: 22999798.