Lua Labs Report — Subclinical Hypothyroidism and the Luteal Phase: Reduced Progesterone, Anovulation, and Recurrent Miscarriage

Date: 2026-05-29 Researcher: Lua Labs Classification: Neuroendocrine / Emerging therapies Line: L3 — Thyroid axis and reproductive function Subtopic: 3.2 — Subclinical hypothyroidism and luteal phase: reduced progesterone, anovulation, recurrent miscarriage

External sources

1. Anon (2024). "Role of thyroid stimulating hormone in the maintenance and functioning of the human corpus luteum." Journal of Endocrinological Investigation. DOI: 10.1007/s40618-023-02269-z. PMID: 38190029. [Experimental paper in human hGL — TSHR confirmed in corpus luteum, TSH causes dose-dependent apoptosis, does NOT stimulate P4/StAR/HSD3β]

2. ASRM Practice Committee (2024). "Subclinical hypothyroidism in the infertile female population: a guideline." Fertility and Sterility. DOI: 10.1016/j.fertnstert.2023.09.003. S0015-0282(23)02109-X. [Definitive 2024 clinical guideline: TSH>4.12 treat; TSH 2.5-4.12 + anti-TPO → consider]

3. Korevaar TIM et al. (2022). "Levothyroxine in euthyroid thyroid peroxidase antibody positive women with recurrent pregnancy loss (T4LIFE trial)." Lancet Diabetes & Endocrinology 10(5):322–329. PMID: 35298917. [Phase 3 RCT n=368 — LT4 in euthyroid anti-TPO+ women does NOT improve live birth rate]

4. Wang Y et al. (2025). "Association of thyroid autoimmunity and pregnancy outcomes in unexplained recurrent pregnancy loss women: a prospective cohort study." Frontiers in Endocrinology. DOI: 10.3389/fendo.2025.1711369. PMC: 12685602. [Prospective cohort n=576 — TAI in 17.5% of unexplained RPL, documented Th17/NK mechanism]

5. Carrillo-Lozano E et al. (2021). "Effect of the cut-off level for thyroid-stimulating hormone on the prevalence of subclinical hypothyroidism among infertile Mexican women." Diagnostics 11(3):426. PMC: 8001256. [n=1,496 infertile Mexican women — 40.7% vs 14.7% SCH with threshold 2.5 vs 4.1 mIU/L]

6. Velázquez-García JA et al. (2025). "Progression of subclinical hypothyroidism in a Mexican public hospital population: a retrospective cohort study." Cureus. DOI: 10.7759/cureus.82923. PMC: 12021256. [Retrospective cohort n=222, public hospital in Tamaulipas — natural history of SCH in LATAM]

7. Rashid MH et al. (2023). "The association of Treg and Th17 cells development factors and anti-TPO autoantibodies in patients with recurrent pregnancy loss." PMC: 10619307. [Foxp3↓/TGFβ↓ + RORγT↑/IL-17↑ in anti-TPO+ RPL]

Baseline knowledge (what I know before searching)

The corpus luteum as a T3-dependent organ

Progesterone production in the corpus luteum is not an autonomous process: it critically depends on external endocrine inputs, and among them the active thyroid hormone T3 (triiodothyronine) occupies a structural role that clinical endocrinology often overlooks. The corpus luteum expresses thyroid hormone receptors (TRα1, TRα2, TRβ1) — nuclear transcription factors from the nuclear receptor superfamily that, when bound to T3, regulate the expression of multiple steroidogenic genes. The three critical steps of progesterone synthesis have documented thyroid influence.

The complete chain:

• Cholesterol uptake (LDL/HDL): the LDL-R receptor has thyroid response elements (TRE) in its promoter. T3 upregulates LDL-R → more substrate available.
• Intramitochondrial transport: StAR (Steroidogenic Acute Regulatory Protein) carries cholesterol from the outer to the inner mitochondrial membrane — this is the true rate-limiting step. Its promoter contains functional TREs (documented in rat and bovine, probable in human). T3 upregulates StAR directly or indirectly.
• CYP11A1 (p450scc): converts cholesterol into pregnenolone at the inner membrane. Modulated by StAR availability.
• HSD3B2 (3β-HSD type 2): converts pregnenolone into progesterone in the endoplasmic reticulum. Also apparently modulated by T3.

A woman with free T3 in the lower quartile of the "normal" range has each of these steps operating below its potential. The result is a corpus luteum that produces progesterone suboptimally, with peak luteal P4 20-35% below the expected value — too subtle to be detected in a single spot test, but sufficient to shorten the luteal phase and impair endometrial receptivity.

The TRH→TSH→PRL axis as an upstream mechanism of anovulation or LPD

In hypothyroidism — including subclinical hypothyroidism — the hypothalamus responds to reduced thyroid tone by raising TRH. Here is the point reproductive endocrinology often omits: TRH has pleiotropic action on the pituitary. In addition to stimulating thyrotrophs (the "official" effect), it directly stimulates lactotrophs to secrete prolactin (PRL). In overt hypothyroidism, the resulting hyperprolactinemia can cause amenorrhea and complete anovulation. In SCH, the effect is more gradual: PRL may be only 115-130% of normal — within range on standard testing, upper quartile — and produces:

1. Partial inhibition of kisspeptin neurons in ARC → reduced amplitude of the preovulatory LH surge → follicle that ovulates "halfway" → low-quality corpus luteum from origin (follicular-upstream LPD)
2. In more pronounced SCH (TSH 3.5-10): PRL sufficiently elevated to suppress GnRH pulsatility → normogonadotropic anovulation (LH and FSH "normal" on a single test but without adequate pulses)

This mechanism explains a confusing clinical phenomenon: the woman with "normal" LH on day 3, "normal" FSH, who does not ovulate or has a short luteal phase. The system is not broken in its static parameters — it is broken in its dynamics.

Anti-TPO antibodies: the immunological layer independent of TSH

Positive anti-TPO is not only a marker of thyroid inflammation — it is evidence that the immune system has systemic Th1-dominant activation with reproductive consequences independent of TSH level:

1. Hypercytokinetic uterine NK activation: women with autoimmune thyroiditis have peripheral NK cells with greater cytotoxic activity. Uterine NK cells (uNK) reflect the systemic immune environment. Trophoblast invading in the presence of hyperactivated uNK → higher probability of rejection in weeks 4-8 (before complete placentation).

2. Th17/Treg imbalance: thyroid autoimmunity coexists with expansion of Th17 (IL-17, pro-inflammatory) and reduction of tolerogenic Treg (Foxp3+, TGFβ+). A receptive endometrium requires a Treg-dominant window to avoid rejecting the semi-allogeneic embryo. When Treg falls because of thyroid autoimmunity, this window narrows.

3. Complement activation: high-titer anti-TPO can activate the classical complement pathway in the decidua → C3a/C5a → local inflammation → failed implantation. Same mechanism as antiphospholipid antibodies in APLA.

The most disputed threshold in reproductive endocrinology

Guidelines have oscillated dramatically: in 2010-2015 many societies recommended TSH < 2.5 in pregnant/trying-to-conceive women; in 2019-2024 there was retreat based on RCTs that did not show benefit from treating the gray range 2.5-4.0. However, these studies did not stratify by DIO2 Thr92Ala, cortisol phenotype, selenium nutritional status, or inflammatory burden. The gray range 2.5-4.0 remains the most underdiagnosed territory in reproductive endocrinology — especially in LATAM, where prevalence with that threshold can reach 40%.

Findings from recent papers

JEI 2023 (PMID 38190029): the luteal TSHR is not what we thought

The paper published in Journal of Endocrinological Investigation (DOI: 10.1007/s40618-023-02269-z) provides the most counterintuitive finding of this session: although TSHR is confirmed by RT-PCR and immunohistochemistry in human granulosa-lutein cells (hGL) and in the human corpus luteum, direct TSH stimulation in hGL culture does NOT modulate progesterone production or the expression of StAR, p450scc (CYP11A1), or HSD3β1/2.

What elevated TSH does do in those cells: induces dose-dependent cell death.

This result reverses the intuitive paradigm. TSHR in the corpus luteum is not a steroidogenesis-promoting trophic receptor (like the LH receptor). It is a mediator of apoptosis when TSH is elevated. The implications:

• The SCH → LPD pathway is not "TSH does not stimulate P4" (absent positive function)
• The real pathway is "elevated TSH kills corpus luteum cells" (active cytotoxicity)
• This explains why the SCH-LPD relationship is more robust than TSH→P4 pharmacology would predict

What does stimulate P4 is free T3 in luteal cells. T3 stimulation of P4 was documented (Rao et al. 1998, PMID 9846161) and is blocked by cycloheximide — it requires synthesis of a mediator protein that has not been definitively identified, whose candidate is StAR. In SCH, available T3 is reduced proportionally to the reduction in T4→T3 conversion by DIO2.

ASRM 2024 and the "LATAM gray zone": 25.9% of invisible women

The ASRM 2024 guideline updates thresholds for infertile women:

• TSH > 4.12 mIU/L: treat
• TSH 2.5–4.12 mIU/L + anti-TPO+: consider treatment
• TSH 2.5–4.12 mIU/L without anti-TPO: insufficient evidence

The LATAM argument is compelling: Carrillo-Lozano 2021 (n=1,496 infertile Mexican women) showed 40.7% SCH with threshold 2.5 vs. 14.7% with threshold 4.1. The difference (25.9%) represents women who remain invisible to diagnosis unless they have anti-TPO+. Again, Velázquez-García 2025 (Cureus, n=222, public hospital in Tamaulipas, TSH 4.2-10.0) documents the natural history in Mexico: progression rates and predictive factors — the first temporal follow-up data on SCH in LATAM, published in 2025.

T4LIFE (Lancet 2022): the RCT that clarifies but does not close the debate

T4LIFE (Korevaar et al., n=368, 13 hospitals NL/BE/DK) is the largest RCT on LT4 and recurrent pregnancy loss. Primary result: LT4 in euthyroid anti-TPO+ women did NOT improve live birth rate vs. placebo.

Correct reading in three points:

1. The women were euthyroid — TSH within the normal range. Not SCH.
2. The anti-TPO → RPL mechanism in euthyroid women runs through immunology, not through TSH. LT4 does not correct immunology.
3. For women with SCH (TSH > threshold), the pre-T4LIFE evidence remains consistent with treatment benefit. T4LIFE did not test that population.

The resulting paradox: in anti-TPO+ women without SCH, levothyroxine does not help (T4LIFE); in anti-TPO+ women with SCH, levothyroxine probably does help — no specific post-T4LIFE RCT confirms this, but the biology supports it. The clinically relevant distinction is the underlying thyroid status, not the mere presence of antithyroid antibodies.

Wang et al. 2025: immune mechanism of thyroid-related recurrent miscarriage

Wang et al. (2025, Frontiers in Endocrinology) prospectively enrolled 576 euthyroid women with unexplained RPL. 17.5% (n=101) had thyroid autoimmunity (TAI+). Adjusted findings: TAI+ → higher frequency of subsequent secondary infertility + higher odds of additional pregnancy loss. Mechanism: Th1/Th17 imbalance, NK hyperactivity, reduced Treg.

Rashid et al. 2023 (PMC10619307) complement this: in anti-TPO+ RPL, Foxp3 and TGFβ were significantly lower, RORγT and IL-17 elevated. It is the same pro-inflammatory immune phenotype that, in L2.3, chronic cortisol induces via NF-κB/Th17. Thyroid autoimmunity and chronic stress converge on the same endotype that damages implantation — an L2.3 ↔ L3.2 connection that the literature does not make explicit.

Complete molecular/endocrine mechanism

Five pathways of thyroid-mediated luteal failure — original Lua Labs model

SCH (TSH 2.5-10 mIU/L) + reduced T3 + ± Anti-TPO(+)
                              │
        ┌─────────────────────┼──────────────────────────────────┐
        │                     │                                  │
        ▼                     ▼                                  ▼
   [PATHWAY 1]           [PATHWAY 2]                        [PATHWAY 3]
Elevated TSH →           Elevated TRH →                    Anti-TPO →
TSHR in luteal           lactotrophs →                     Th17↑ / Treg↓ →
cells →                  PRL ↑ (8-20% SCH) →               Hyperactivated
Dose-dependent           Kisspeptin ARC ↓ →                uterine NK →
luteal apoptosis         Suboptimal LH surge →             hostile endometrium →
(JEI 2023) →             Submature follicle →              implantation failure /
Premature CL             lower-quality CL                  early loss
regression               (upstream LPD)
        │
        ▼                     
   [PATHWAY 4]           [PATHWAY 5 — L2 CONNECTION]
Low free T3 →            Chronic cortisol (diurnal cortisol phenotype-A) →
TR in hGL underactive →  DIO2 ubiquitination (L3.1) →
Mediator protein ↓ →     additionally low tissue T3 →
StAR/CYP11A1 ↓ →         Amplifies PATHWAY 4
HSD3B2 ↓ →               + Cortisol competes with P4 at GR (L2.3)
P4 synthesis ↓           + CBP/p300 co-activators sequestered
(direct T3 mechanism)

Complete steroidogenic pathway — how low T3 affects each link:

Serum cholesterol (LDL/HDL)
        │
        ↓ [LDL-R ← TRE in promoter, T3 upregulates uptake]
Intracellular cholesterol
        │
        ↓ [StAR ← TRE in promoter, T3 upregulates synthesis; rate-limiting step]
Inner mitochondrial membrane
        │
        ↓ [CYP11A1/p450scc ← availability depends on StAR]
Pregnenolone
        │
        ↓ [HSD3B2 ← probable T3 modulation]
PROGESTERONE
        │
        → If SCH + Phenotype A: low P4 → GR saturated with cortisol (L2.3)
                               → FKBP51-PR inactivates the little that arrived
                               → Decidua does not receive sufficient P4 signal

TRH-PRL pathway — from SCH to anovulation:

SCH → insufficient T3/T4 tone
        │
        ↓ compensatory
Hypothalamic TRH ↑
        │
        ├──→ Thyrotrophs → TSH ↑ (expected effect — already in PATHWAY 1 and 4)
        │
        └──→ Lactotrophs → PRL ↑ (8-20% of SCH have elevated PRL)
                  │
                  ├── Mild PRL (within range, upper quartile):
                  │       → Kisspeptin ARC ↓ partially
                  │       → Preovulatory LH surge amplitude ↓ 15-25%
                  │       → Follicle ovulates with lower maturity
                  │       → Suboptimal CL from origin
                  │       → Peak luteal P4 ↓ 15-25% (not detectable with one test)
                  │       = "follicular-upstream" LPD
                  │
                  └── Moderate PRL (>25-30 ng/mL, in more pronounced SCH):
                          → Pulsatile GnRH significantly suppressed
                          → Spot LH "normal" on test
                          → BUT insufficient pulsatility
                          → Normogonadotropic anovulation

Anti-TPO → RPL pathway:

Anti-TPO (+) → systemic Th1/Th17-dominant autoimmunity
        │
        ├──→ Cytotoxic peripheral NK ↑ → uNK CD56+CD16+ ↑ (killer phenotype)
        │         → Invasive trophoblast as target
        │         → Loss weeks 4-8 (pre-placentation)
        │
        ├──→ Th17↑ (RORγT↑, IL-17↑) + Treg↓ (Foxp3↓, TGFβ↓) [Rashid 2023]
        │         → Tolerogenic endometrial window compromised
        │         → Insufficient decidualization
        │         → Blastocyst rejected
        │
        └──→ Complement C3 activation (ADCC)
                  → C3a/C5a → local decidual inflammation
                  → L1.3 connection: same mechanism as bacterial LPS →
                    TLR4 → NF-κB → granulosa apoptosis

Cross-synthesis with previous findings

The L2.3 "4-arm" model gains a fifth arm

In L2.3 we documented the 4-arm model of stress-related luteal failure:

1. P4 substrate deficit (cortisol diverts cholesterol)
2. GR-phase dyssynchrony (FKBP51-PR functionally withdraws P4)
3. NF-κB/Th17 inflammation (chronic cortisol)
4. Progesterobolomic collapse (L1.2)

L3.2 adds a fifth arm that does not appear in any L2 paper:

5. Insufficient T3 → luteal mediator protein ↓ → StAR/CYP11A1 ↓ → reduced P4 synthesis + elevated TSH → TSHR-mediated apoptosis → corpus luteum degrades prematurely (JEI 2023)

Bridge L3.2 → L2.4 (diurnal cortisol phenotype): susceptibility asymmetry between phenotypes

Phenotype A (HPA Hyperreactive) — double metabolic trap:

• High chronic cortisol → active DIO2 ubiquitination (Benvenga 2021, L3.1) → low tissue T3 with TSH/T4 in the normal-high zone
• With gray-zone TSH 2.5-3.5: DOUBLE HIT — reduced T3 from inhibited DIO2 + TSHR-mediated apoptosis of the corpus luteum (JEI 2023)
• The luteal cell simultaneously receives a deficiency signal (low T3 → StAR ↓) + a cytotoxic signal (elevated TSH → TSHR → apoptosis)
• Additionally: high cortisol activates intrafollicular HSD11B1 (L2.2) → high intrafollicular cortisol → saturated GR → FKBP51-PR inactivates the P4 that arrives (L2.3) → triple luteal-cortisol-receptor blockade

Phenotype B (Collapsed Allostatic Load) — different mechanism:

• Low cortisol → less DIO2 ubiquitination → peripheral T4→T3 conversion paradoxically more preserved
• BUT: low DHEA-S (L2.4) → fewer androgens → less follicular E2 → submature follicle → upstream LPD
• Phenotype B damage runs mainly through the follicle, not directly through the corpus luteum

Differential prediction: the same TSH of 3.0 mIU/L produces worse luteal deterioration in Phenotype A than in Phenotype B. This asymmetry is the core of H20.

Bridge L3.2 → L1.2 (progesterobolome): the invisible gut-thyroid circuit

A connection not articulated in the literature, emerging from the L1 + L3 cross-analysis:

SCH → constipation (item #3 of the thyroid-symptom phenotype, cardinal symptom of hypothyroidism) → longer intestinal transit time → lower substrate availability for Eggerthella lenta (cortisol→P4 conversion via 21-dehydroxylation, documented in L1.2) → less intestinal "rescue" P4 → further amplified LPD

Additionally: hypothyroidism affects proximal colon motility (enteric smooth muscle via Ca²⁺ channels and Na⁺/K⁺ ATPase) → disfavors Parabacteroides spp. (they require a normomotile environment) → fewer sulfatases → both progesterobolome pathways deteriorate simultaneously.

Negative feedback circuit: SCH → constipation → compromised progesterobolome → less P4 → LPD → less receptive endometrium → reduced fertility. Each mechanism is individually modest; the accumulated circuit is clinically relevant.

Bridge L3.2 → L2.3 (GR phase sensor): competition for co-activators

In L3.1 we documented that T3 and cortisol compete for transcriptional co-activators CBP/p300/SRC-1 in the endometrium. L3.2 extends this to the corpus luteum: under low T3 + high cortisol (Phenotype A), co-activators in luteal cells are preferentially recruited by GR (cortisol) instead of TR (T3) → double steroidogenic suppression:

• Pathway 1: less T3 → less active TR → less StAR/CYP11A1
• Pathway 2: cortisol competes for co-activators required by TR → even lower expression of T3-dependent genes even if T3 is marginally available

The "gray zone" TSH 2.5-4.0 + Phenotype A is functionally equivalent to a much higher TSH in terms of real luteal function.

The L2.3 ↔ L3.1 ↔ L3.2 triangle — now closed

L3.1 established that TSH varies intra-cycle with a mid-cycle peak. L3.2 specifies the exact damage mechanism: the mid-cycle TSH peak coincides with corpus luteum formation. If that peak is elevated (SCH), newborn luteal cells face a cytotoxic TSHR signal at their most vulnerable moment. L2.3 showed that luteal GR is a "phase sensor" with a changing role. Integration: the corpus luteum is attacked with triple timing — cytotoxic TSH during formation (L3.2) + cortisol competing through GR during maturation (L2.3) + low T3 limiting P4 synthesis throughout its lifespan (L3.1+L3.2). The triangle is closed.

Adaptogens: ashwagandha as an unrecognized peripheral thyroid modulator

Prior research on adaptogens describes that Withaferin A — the active principle of ashwagandha — acts on the Hsp90-GR-FKBP51 chaperone complex. The same Hsp90 complex also chaperones the thyroid hormone receptor TRβ. If ashwagandha lowers cortisol in women with an elevated diurnal cortisol phenotype, the expected consequence would be less DIO2 ubiquitination → greater T3 availability → better luteal progesterone synthesis; through that route, the adaptogen's benefit on the luteal phase could have an unrecognized peripheral thyroid component. This is a mechanistically plausible extrapolation that has not yet been verified in vitro.

Lua Labs Hypotheses

Hypothesis 19 — Subthreshold luteal deterioration as a reproducible temporal pattern

Statement: In women with an active cycle (25-42 years) and gray-zone TSH 2.5–4.12 mIU/L, luteal-phase deterioration is detectable as a reproducible temporal pattern — shorter luteal duration, greater symptom intensity on days 18-28, greater luteal morning cold intolerance, and greater inter-cycle variability in duration — before TSH exceeds the classic diagnostic threshold of 4.12 mIU/L. The hypothesis proposes that this longitudinal symptom pattern predicts subthreshold subclinical hypothyroidism with greater sensitivity than a single spot TSH measurement in the 2.5-4.12 range.

Proposed mechanism:

Gray-zone TSH 2.5-4.12 → available T3 in luteal cells ↓ proportionally
        │
        ├──→ StAR/CYP11A1 ↓ → suboptimal P4
        │         → Shorter luteal phase (−1.5 to −3 days vs euthyroid women)
        │         → Intensified luteal symptoms
        │           (insufficient P4 → less allopregnanolone → GABA-A ↓
        │            → anxiety/insomnia/irritability days 18-28)
        │
        ├──→ Mildly elevated TSH → TSHR-mediated apoptosis (JEI 2023)
        │         → CL regression 2-4 days premature
        │         → Early bleeding / subtly short cycle
        │
        └──→ Upper-quartile PRL → mildly reduced LH pulsatility
                  → Follicle 10-15% less mature (subtle, not anovulatory)
                  → Lower-quality CL from origin
                  → Peak luteal P4 −15-25% vs control
                  (not detectable with a single test, detectable with longitudinal symptoms)

Pattern distinction by phenotype (L2.3): luteal failure may be predominantly stress-mediated (high luteal stress phenotype with a mild luteal symptom pattern → responds to the Luteal-Phase Buffer from L2.3), subclinical-thyroid (low luteal stress phenotype with a marked luteal symptom pattern → responds to LTCP), or multifactorial when both coexist (the 5 arms of the model → combined L2.3 + LTCP). The longitudinal luteal-phase symptom pattern — shorter duration, greater symptom intensity on days 18-28, and greater inter-cycle variability — helps distinguish which of these mechanisms predominates.

Confidence level: Medium-High. Solid mechanism in the literature. What is original: that subthreshold luteal deterioration is observable from the longitudinal symptom pattern without measured TSH. Weakest link: morning cold intolerance has low individual specificity.

How to validate:

With formal study:

• Design: prospective observational, n=150, 6 cycles
• Outcome: TSH + FT3 + FT4 + anti-TPO on day 3-5; luteal duration by urinary LH; P4 on day 21-23
• Power: detect r=0.35 between the luteal symptom pattern and luteal P4 (alpha=0.05, beta=0.20, n=125)
• Sub-analysis: DIO2 Ala/Ala vs Thr/Thr — differential luteal symptom pattern with same TSH

Limitations:

• Estimating luteal duration from symptoms without temperature: precision ±2-3 days, sufficient for detecting the pattern but not clinical diagnosis
• Cold intolerance: low specificity — anemia, hypoglycemia, Raynaud also cause it
• In perimenopause (Carmen): cycles are frequently anovulatory → the pattern is not applicable in those cycles. Requires filtering estimated ovulatory cycles.
• Intra-cycle TSH variability (L3.1): day 3 TSH may not reflect the mid-cycle TSH peak that damages the corpus luteum
• Confounding with LPD from other causes (anovulatory PCOS, vitamin B6 deficiency, idiopathic luteal insufficiency)

Hypothesis 20 — NEW: "Cortisol-Thyroid Metabolic Trap: diurnal cortisol phenotype Phenotype A amplifies luteal damage from the thyroid gray zone through a dual DIO2 + TSHR mechanism"

Statement: In women with diurnal cortisol phenotype Phenotype A (HPA Hyperreactive, chronically high diurnal cortisol), the same TSH level in the gray zone (2.5–4.12 mIU/L) produces a significantly worse luteal phase than in women with diurnal cortisol phenotype Phenotype B or neutral diurnal cortisol phenotype, due to a triple amplification mechanism:

1. Chronic cortisol → DIO2 ubiquitination → even lower tissue T3 (above the basal reduction proportional to SCH)
2. Elevated TSH → TSHR-mediated apoptosis in luteal cells (JEI 2023), amplified by the high-cortisol intrafollicular environment (HSD11B1-CRHR1, L2.2)
3. Cortisol-T3 competition for CBP/p300 co-activators in granulosa → lower expression of T3-dependent steroidogenic genes even if T3 is marginally available

Proposed mechanism:

PHENOTYPE A (diurnal cortisol phenotype) + GRAY-ZONE TSH 2.5-4.12
                              │
              ┌───────────────┴──────────────────────────────┐
              ▼                                              ▼
HIGH CHRONIC CORTISOL (elevated CAR)              ELEVATED TSH (gray zone)
              │                                              │
              ├──→ DIO2 ubiquitination (Benvenga 2021)      │
              │         → Tissue T3 ↓↓ (DOUBLE HIT)         │
              │                                              │
              ├──→ Intrafollicular HSD11B1 ↑ (L2.2)         │
              │         → High intrafollicular cortisol      │
              │                                              │
              └──→ Co-activator competition                  ▼
                    CBP/p300 in granulosa           TSHR in luteal cells
                          │                                  │
                          ▼                                  ▼
              T3-dependent genes ↓↓                 Cytotoxic signaling
              (StAR, CYP11A1, HSD3B2)               → Accelerated luteal
              P4 synthesis highly compromised         apoptosis (JEI 2023)
                          │                          Premature CL regression
                          └──────────────┬───────────────────┘
                                         ▼
                           SEVERE LPD IN THE THYROID GRAY ZONE
                           (same TSH 3.0 → much worse LPD in diurnal cortisol phenotype-A
                            than in diurnal cortisol phenotype-B or diurnal cortisol phenotype-neutral with same TSH)
                                         │
                                         ↓
                           Low P4 reaches endometrium +
                           GR saturated with cortisol (L2.3) +
                           FKBP51-PR deactivates residual P4 →
                           NONFUNCTIONAL DECIDUA
                           → Implantation failure / early loss

Connection with H14 (L2.2): H14 proposed that reproductive fate under stress is determined by the interaction of CRH magnitude × local ovarian sensitivity × microbial buffer. H20 adds a fourth determinant: tissue thyroid status as an amplifier that can collapse the ovarian inverted U (Gershon 2025) into a monotonic descending slope. For Phenotype A + gray-zone SCH women, there is no beneficial inverted U — cortisol ubiquitinates DIO2, raises intrafollicular HSD11B1, AND TSH kills luteal cells simultaneously.

Confidence level: Medium. Individual links documented. Weakest link: whether DIO2 ubiquitination under moderate chronic cortisol (not Cushing's) adds a measurable effect on tissue T3 already reduced by mild SCH.

How to validate:

With formal study:

• n=80, stratified by TSH (< vs ≥ 2.5) × morning cortisol (< vs ≥ median)
• Outcome: mean luteal P4 (days 18-25) × 3 cycles
• Analysis: mixed-effects model with TSH×cortisol interaction term as predictor of mean luteal P4

Limitations:

• In vivo DIO2 ubiquitination under moderate cortisol (not Cushing's) has not been directly quantified in humans
• TSHR cytotoxic effect (JEI 2023) was demonstrated in vitro with pharmacologic TSH — relevance to TSH 2.5-4.12 in vivo needs confirmation
• Confounding by BMI, insulin resistance, underlying PCOS

Candidate formulation: "Luteal-Thyroid Coherence Protocol (LTCP)"

Target population: Sofía (28, active cycle, probable subclinical LPD), Carmen (47, perimenopause with sporadically ovulatory cycles)

Objective: Optimize the 5 vectors of the thyroid-mediated luteal failure model without pharmacology.

LTCP as complementary fifth arm: the Luteal-Phase Buffer (L2.3) addresses stress-GR-P4; the LTCP addresses SCH-T3-P4. They are complementary. The combination of both would cover the 5 arms of the luteal failure model documented in L3.2.

Regulatory status: 100% food-based + behavioral. GRAS. Zero prescription.

Individual variability

The same TSH of 3.0 mIU/L can produce a normal luteal phase or significant LPD. The determinants:

Genetic:

• DIO2 Thr92Ala (rs225014): Ala/Ala ~13-15% of population → tissue T3 ↓ ~30% with same TSH/T4. Maximum amplifier of the entire luteal cascade of H19/H20. Not characterized in Mexican/LATAM women.
• PGR PROGINS: lower PR efficiency → same suboptimal P4 produces even lower tissue response. Multiplicative effect when P4 is already low due to SCH.
• FKBP5 rs1360780 T/T (inherited L2.1/L2.3): "functional withdrawal of P4" without serum change. When P4 is already suboptimal due to SCH, T/T amplifies the deficient luteal phenotype to a clinically manifest level.
• NR3C1 BclI G/G (GR hypersensitive, inherited L2.3): greater cortisol competition with P4 in all tissues → amplifies the effect of low P4 from SCH.

Nutritional:

• Serum selenium < 70 µg/L: reduced DIO2 activity → insufficient T4→T3 conversion. Estimated 30-40% of urban LATAM women deficient (not well characterized in Mexico).
• Iodine: Silent deficiency from Himalayan/sea salt without iodine — unquantified urban "wellness" trend. A woman who switches to non-iodized salt without compensating with dairy/seafood can develop functional deficiency within months.
• Iron (ferritin < 30 µg/L): TPO is a hemoprotein — deficiency → reduced T4 synthesis → "functional" SCH with anatomically normal gland. ~35% of Mexican women 20-45 years with deficiency (ENSANUT 2022). Most prevalent modifiable LATAM nutritional determinant.
• Vitamin D < 20 ng/mL: VDR modulates NF-κB → reduces Th17. Deficiency ≥40% in urban LATAM. Aggravates the Anti-TPO → Th17/NK pathway.

Environmental/epigenetic:

• Chronic cortisol (diurnal cortisol phenotype Phenotype A): DIO2 ubiquitination → amplification of low tissue T3 (central H20 mechanism)
• Chronic systemic inflammation (VAT obesity, LPS dysbiosis): DIO3 upregulation → T3 inactivation → elevated rT3. Subclinical equivalent of NTIS.
• Endocrine disruptors: perchlorate, nitrate (unfiltered water), heated BPA (microwave plastics) — compete with iodine at the Na⁺/I⁻ pump or act as TR antagonists. Underestimated in LATAM.
• Urban acculturation: loss of selenium/zinc from the traditional Mesoamerican diet when adopting ultra-processed foods. Same vector as L1.4 for microbiome, now also relevant to peripheral thyroid function.

Medical history:

• Positive anti-TPO: double damage (endocrine + immune). Euthyroid anti-TPO+ women (TSH < 4.12) fall outside ASRM 2024 treatment but benefit from immunomodulatory interventions (anti-inflammatory diet, VD, omega-3 via food).
• Family history of Hashimoto's: HLA-DR3/DR4 polymorphisms increase risk of thyroid autoimmunity AND autoimmune infertility.
• Early miscarriages or "chemical pregnancies": history of ≥2 early losses with elevated thyroid-symptom phenotype and no recent anti-TPO/TSH is a high-yield signal for referral. In LATAM, "chemical pregnancies" are highly underdiagnosed — many women experience them as "I was a few days late and then it came very heavy."

Researcher Notes

The most urgent LATAM gap in L3 after this session:

1. Frequency of DIO2 Ala/Ala in Mexican women (no data in literature)
2. Serum selenium levels in urban LATAM women 20-45 years (not characterized)
3. Iodine content in commercial Himalayan salt brands available in Mexico (analytical laboratory — possible project costing < $500 MXN)
4. Natural history of the "gray range" 2.5-4.12 in infertile Mexican women with 12-month follow-up (Velázquez-García 2025 covers only TSH > 4.2)

Minor emerging secondary hypothesis (without formal number — requires L3.3): The combination "Phenotype B (diurnal cortisol phenotype) + anti-TPO+ + gray-zone SCH" produces silent pregnancy loss without severe LPD symptoms — because Phenotype B has low cortisol (fewer visible stress symptoms) but the immune pathway of anti-TPO operates independently of cortisol. "Calm women, without apparent stress" with repeated early losses may be in this invisible phenotype. L3.3 (thyroid autoimmunity and POI) can develop it.