Lua Labs Report — TSH as a global modulator of the menstrual cycle

Date: 2026-05-27 Researcher: Lua Labs (Scientist) Classification: Neuroendocrine Line: L3 — Hypothalamic-pituitary-thyroid (HPT) axis and reproductive function Sub-topic: 3.1 — TSH as a global modulator of the menstrual cycle (formal opening of L3)

External sources

Mintziori G, Veneti S, Kassi E, Liberopoulos E, Macut D, Goulis DG, Lambrinoudaki I. (2024). EMAS position statement: Thyroid disease and menopause. Maturitas, 185, 107991. DOI: 10.1016/j.maturitas.2024.107991. PMID: 38658290.
ASRM Practice Committee. (2024). Subclinical hypothyroidism in the infertile female population: a guideline. Fertility and Sterility, 121(6), 977–987. DOI: 10.1016/j.fertnstert.2024.01.029. Ref: S0015-0282(23)02109-X.
Aghajanova L, Lindeberg M, Carlsson IB, Stavreus-Evers A, Zhang P, Scott JE, Hovatta O, Skjöldebrand-Sparre L. (2009). Receptors for thyroid-stimulating hormone and thyroid hormones in human ovarian tissue. Reproductive BioMedicine Online, 18(3), 337–347. PMID: 19298732. DOI: 10.1016/S1472-6483(10)60091-0.
Sun SC, Hsu PJ, Wu FJ, Li SH, Lu CH, Luo CW. (2010). Thyrostimulin, but not thyroid-stimulating hormone (TSH), acts as a paracrine regulator to activate the TSH receptor in mammalian ovary. Journal of Biological Chemistry, 285(6), 3758–3765. PMC2823517. DOI: 10.1074/jbc.M109.066266.
Luongo C, Dentice M, Salvatore D. (2023). Deiodinases and their intricate role in thyroid hormone homeostasis (and update on Type II deiodinase polymorphisms). Nature Reviews Endocrinology / Endocrinol Metab (updated review). PMID: 37150515 (rev. 2023 EnM 38(1):95–106). DOI: 10.3803/EnM.2022.1631.
Carrillo-Lozano E, Sayán-Hidalgo NM, Vásquez-Cruz 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), 417. PMC8001256. DOI: 10.3390/diagnostics11030417.
López-Caudana AE, et al. (2024). Evaluation of the risk of hypothyroidism and its clinical manifestations using the Zulewski scale in a Mexican population sample. Frontiers in Endocrinology, 15, 1416663. PMC11362067. PMID: 39220359. DOI: 10.3389/fendo.2024.1416663.
Plowden TC, Schisterman EF, Sjaarda LA, Perkins NJ, Silver R, Radin R, Kim K, Galai N, DeVilbiss EA, Mumford SL. (2018). Thyroid hormones and menstrual cycle function in a longitudinal cohort of premenopausal women (BioCycle Study). Paediatric and Perinatal Epidemiology, 32(1), 1–12. PMID: 29517803. DOI: 10.1111/ppe.12428.
Krassas GE, Poppe K, Glinoer D. (2010 / Poppe 2008 update). Thyroid function and human reproductive health. Endocrine Reviews / Human Reproduction Update, 14(4), 357–369. DOI: 10.1093/humupd/dmm045. (Inherited from pre-lab report 2026-05-14, used here as a mechanistic anchor.)
Khouri NK, Toloza FJK, Maraka S, et al. (2025). Reverse T3 in patients with hypothyroidism on different thyroid hormone replacement regimens. PLOS ONE, 20(5), e0325046. DOI: 10.1371/journal.pone.0325046.
Benvenga S, Klose M, Vita R, Feldt-Rasmussen U. (2021). Cushing's syndrome effects on the thyroid. International Journal of Molecular Sciences, 22(6), 3131. PMC8033054. DOI: 10.3390/ijms22063131. (Clean model of extreme hypercortisolism → DIO1/DIO2 inhibition + reversible T3 decline after surgery.)
Bagheripuor F, Tehrani FR, et al. (2024). Global research trends in endometrial receptivity from 2000 to 2024: bibliometric analysis (with a dedicated section on T3-TR signaling in endometrial receptivity). Frontiers in Medicine, 11, 1465893. PMC11558532. DOI: 10.3389/fmed.2024.1465893.

Baseline knowledge (what I know before searching)

The hypothalamic-pituitary-thyroid (HPT) axis appears, on the surface, to be a classic negative feedback circuit: TRH (PVN) → TSH (pituitary thyrotrophs) → thyroid → T4 (main secreted output) → peripheral T4→T3 conversion by 5'-deiodinase type I (DIO1, liver/kidney) and type II (DIO2, peripheral tissues including ovary and endometrium), with DIO3 inactivating T3→T2 and T4→rT3 as a tissue-level brake. The biologically active hormone is T3, which binds to TRα1, TRα2, TRβ1 and TRβ2 — nuclear receptors whose primary function is transcriptional regulation over hundreds of genes in the range of “basal metabolism, thermogenesis, growth, differentiation, insulin sensitivity.”

But that summary radically underestimates what TSH and T3 do in the reproductive system. There are six mechanistic facts that any serious reader of the cycle must internalize before looking at the new literature:

The TSH receptor (TSHR) is not exclusive to the thyroid. It is expressed in granulosa, theca, oocyte, endometrium (glandular epithelium and stroma), trophoblast and adipocyte. This means circulating TSH — even within “normal ranges” — is sending a direct signal to reproductive tissues, not only regulating the thyroid.
The ovary has a second source of TSHR ligand: thyrostimulin (GPA2/GPB5). It is a heterodimeric glycoprotein produced locally that activates TSHR with greater affinity than TSH itself and operates in an intra-ovarian paracrine/autocrine manner. In L3 this reconfigures the model: the ovary has its own local TSH-like loop, parallel to the systemic one, similar to how in L2.2 we documented paracrine ovarian CRH.
TSH is pulsatile and has a circadian pattern with a nocturnal peak (22:00–02:00). High nocturnal cortisol (hyperreactive HPA → diurnal cortisol phenotype Phenotype A, L2.4) suppresses the nocturnal TSH peak through central inhibition of TRH. In other words: the cortisol pattern that in L2 we characterized as “Phenotype A” has a direct consequence on TSH dynamics before any thyroid pathology.
T4→T3 conversion by DIO2 is subject to rapid ubiquitin-proteasome post-translational regulation. Glucocorticoids (cortisol, dexamethasone), inflammatory cytokines (IL-6, TNF-α), bile acids and prolonged fasting increase DIO2 ubiquitination → degradation → less tissue T3 without serum TSH or T4 changing. This produces the pattern “normal TSH + normal T4 + low T3 + hypo symptoms” that classical endocrinology calls “non-thyroidal illness syndrome” (NTIS), but which is actually a regulatory mechanism of peripheral inhibition.
TSH has intra-menstrual cyclic variation. TSH is higher at mid-cycle and lower in the early follicular and late luteal phases. The amplitude is not trivial (~10–20% within-woman) and, combined with pulsatility, means that a single TSH measurement is diagnostically weak for mild SCH — a point the clinical literature recognizes but routine practice ignores.
The diagnostic threshold for SCH is political, not mechanistic. The typical upper limit of normal (4.0–4.5 mIU/L) reflects the 97.5th percentile of historical cohorts that included subjects with undetected subclinical thyroid autoimmunity. Cohorts with anti-TPO/anti-Tg excluded show a 97.5th percentile closer to 2.5–3.0 mIU/L. The debate over “TSH 2.5 vs 4.0” as a threshold in infertile women is exactly this.

Regarding the cycle: T3 acts on at least four reproductive nodes simultaneously:

Hypothalamus/pituitary: Elevated TRH (in hypoT) co-releases prolactin — high PRL suppresses GnRH pulsatility and replicates a functional amenorrheic phenotype. This is the mechanism taught in classical endocrinology, and it is correct.
Granulosa: T3 + FSH synergize on CYP19A1 (aromatase) → greater follicular A4→E2 conversion. Without sufficient intracellular T3 (saturated or degraded DIO2), FSH is functionally less potent — producing functional FSH resistance independent of circulating FSH level.
Corpus luteum: T3 maintains the expression of StAR, CYP11A1, HSD3B for P4 biosynthesis. Hypothyroidism (clinical or subclinical) → documented luteal insufficiency with reduced mid-luteal P4 (direct bridge to L2.3, GR phase sensor).
Endometrium: TSHR and TRs are expressed; T3 regulates integrins (αvβ3), LIF (leukemia inhibitory factor), HOXA10 — all canonical markers of the implantation window. Endometrial tissue T3 failure → shortened window, blastocyst does not implant even if embryology is correct.

Known individual variability: Thr92Ala polymorphism in DIO2 (rs225014) — the Ala/Ala genotype (~13–15% of the population) has reduced DIO2 activity → lower tissue T3 with identical serum TSH/T4. Women with Ala/Ala and SCH respond worse to T4 monotherapy (LT4 → DIO2 → T3) because conversion is limited. It is the most relevant polymorphism for thyroid nutrigenomics and connects with L3.4 (insulin resistance, elevated BMI Thr92Ala) and L6 (nutrigenomics).

This is the foundation. What the recent papers contribute is: how large the effect is in LATAM humans, how common digitally trackable dysfunction is, how the picture changes in perimenopause, and how separable the “thyroid” signal is from the “stress-cortisol” signal when both modulate the same pathways.

Findings from recent papers

Mintziori et al. 2024 (EMAS position statement, Maturitas) establishes the contemporary framework: thyroid and menopause overlap symptomatically to such a degree that the average clinician confuses one for the other. It lists an explicit convergence of symptoms: menstrual irregularity, mood alteration, increased sweating, sleep disturbance, alopecia, deteriorated quality of life. The official position is: every woman in the perimenopausal transition should be evaluated for thyroid dysfunction, and every woman with thyroid dysfunction near age 45–55 should be evaluated for perimenopause. It is the first time a European medical society has formalized the symptomatic overlap as clinically actionable. Even more relevant: they document that underdiagnosis is the rule, not the exception in this age range.

ASRM 2024 guideline (Fertility and Sterility) partially settles the threshold debate: for the infertile population, TSH > 4.12 mIU/L → treat; between 2.5–4.12 mIU/L → treat only if anti-TPO positive. Three non-obvious points: (a) the threshold applies to the population upper limit adjusted for age and iodine sufficiency, not a universal number; (b) ASRM explicitly recognizes that evidence in the 2.5–4.0 range without anti-TPO does not support universal treatment — a step back versus the more aggressive 2015 guidelines; (c) in euthyroid anti-TPO-positive women, the group is heterogeneous and the decision is individualized. The "gray zone 2.5–4.0 mIU/L" concentrates a significant proportion of symptomatic women and is where clinical management remains controversial.

Carrillo-Lozano et al. 2021 (Diagnostics, Mexico, n=1,496 infertile women) is the strongest regional data that exists for LATAM: SCH prevalence 40.7% if threshold is 2.5 mIU/L vs 14.7% if threshold is 4.1 mIU/L in infertile Mexican women. The gap between thresholds is enormous. Although this is an infertile cohort (with selection bias), the data confirms two things: (1) there is a massive reservoir of Mexican women with TSH in the functional gray zone, (2) the decision about how to categorize them determines whether “10–15%” or “30–40%” of the population is a candidate for some intervention. López-Caudana 2024 (Frontiers Endocrinol, Mexico, Zulewski scale) complements this with non-infertile data: it applied the Zulewski clinical scale to the general Mexican population and found that symptoms compatible with subclinical hypothyroidism are detectable in a significant proportion before any biochemical confirmation. The combined takeaway: Mexico has a problem of symptomatic SCH that is larger than the “official” prevalence suggests, and the suboptimal diagnostic tool is the cause.

Aghajanova et al. 2009 (RBM Online) remains the canonical paper documenting TSHR + TRα1/TRα2/TRβ1 receptors in human granulosa, theca and endometrium. It is not recent, but it is foundational — all 2020+ literature assumes its content. Sun et al. 2010 (JBC) adds the second ligand: thyrostimulin (GPA2/GPB5) activates ovarian TSHR with higher affinity than TSH and operates in a paracrine manner. The updated review MDPI 2025 IJMS (Molecular and Functional Insights into Thyrostimulin) confirms that thyrostimulin is more potent than TSH on TSHR and operates autocrine/paracrine in the ovary, pituitary and CNS. This matters for L3: it means ovarian TSHR signaling is not only a function of serum TSH; there is a local thyrostimulin–TSHR axis that clinical endocrinology simply does not measure.

Plowden et al. 2018 (BioCycle Study) remains the cleanest longitudinal data on TSH across the cycle: n=259 healthy women, regular ovulators, 1–2 complete cycles, multiple measurements. Key findings: TSH varies intra-cycle with a mid-cycle peak and a nadir in late menstruation; total T4 (not free T4) correlates positively with maximum luteal progesterone levels and maximum follicular E2. Implication: total T4 integrated across the cycle is a better predictor of luteal function than any single point measurement. This challenges the clinical practice of “one single TSH measurement on day X.”

Khouri et al. 2025 (PLOS ONE, n=810 women with hypothyroidism on replacement) measures rT3 in treated patients: 11.0% have elevated rT3, with the highest rate (20.9%) on T4 monotherapy. In other words: even with TSH “normalized” by LT4, one in five women is converting T4 toward rT3 (the inactivating DIO3 pathway) instead of T3 (the activating DIO2 pathway). Although this is not a study of untreated women, it opens the door to a relevant phenomenon: normal TSH does not guarantee adequate tissue T3 when the deiodination pattern is biased toward rT3 — exactly the pattern documented in hypercortisolism, chronic inflammation and NTIS.

Benvenga et al. 2021 (IJMS, Cushing's effects on thyroid) is the clean model of the cortisol→thyroid arrow: in Cushing's (massive hypercortisolism), T3 drops disproportionately to T4, and after curative surgery T3 rises before and more than T4 — interpreted as recovery of DIO1 + DIO2 activity after the cortisol brake is removed. In other words, chronic cortisol functionally suppresses T4→T3 conversion, not only through a central mechanism (TRH→TSH) but through peripheral inhibition of deiodinases. This closes the L2→L3 bridge that L2.4 had anticipated.

Bagheripuor 2024 (Frontiers Med, endometrial receptivity bibliometric) confirms that T3-TR signaling is a growing node in endometrial receptivity, with publications increasing year over year linking endometrial tissue T3 with expression of HOXA10, LIF, integrins, MUC1. Its symptomatic correlate — luteal phase with early spotting, cycles that ovulate but do not conceive, early biochemical loss — is what manifests clinically even though endometrial receptivity itself is not directly measurable.

Luongo/Colella 2023 (Endocrinol Metab review, DIO2 polymorphism) documents that Ala/Ala (rs225014) homozygotes represent ~13–15% of the general population; they have slightly reduced serum T3 and, more importantly, reduced tissue T3 in muscle, liver and CNS with no difference in serum TSH/T4. Meta-analyses associate them with higher BMI + higher fasting glucose + persistent cognitive symptoms on LT4 monotherapy. There are no LATAM-specific studies on the frequency of Ala/Ala, but the global prevalence suggests that ~1 in 7 LATAM women has it. This is the first L6 polymorphism that fully enters an earlier line — it anticipates that thyroid nutrigenomics (L6.x) will have real traction.

Complete molecular/endocrine mechanism

What the literature, when cross-referenced with baseline knowledge, makes explicit is that TSH is not “the hormone that says whether your thyroid works.” TSH is a node in a bidirectional system where reproductive output depends simultaneously on (a) serum TSH, (b) the local paracrine ligand thyrostimulin in the ovary, (c) local T4→T3 conversion mediated by DIO2 (modifiable by cortisol, cytokines, bile acids), (d) TRα/TRβ transcriptional sensitivity (modifiable by epigenetics, polymorphisms), and (e) the temporal pattern (nocturnal pulsatility, intra-cycle variation) that a single measurement ignores.

Diagram of the extended model — L3.1 “TSH as a modulator of four simultaneous fronts”:

                    Hypothalamus
                  CRH-PVN (L2.1) ──┐
                                    │ inhibits TRH
                  TRH-PVN ──────────┴───┐
                       │                │
                       ▼                ▼
                    TSH (pituitary)   PRL (pituitary)
                       │                │
                       │                └── suppresses pulsatile GnRH ──┐
                       │                                                  │
            ┌──────────┼────────────────┐                                 │
            │          │                 │                                 ▼
            ▼          ▼                 ▼                              ovulation
        THYROID     OVARY          ENDOMETRIUM              ◄─────────────┘
            │     (TSHR + TR        (TSHR + TR
        serum T4    paracrine        local DIO2/DIO3
            │       thyrostimulin)    integrins, LIF,
            ▼          │              HOXA10)
        DIO1/DIO2      │
        T4 → T3        │
        (liver,        │
         tissues)      ▼
            │     T3 + FSH → CYP19A1 → follicular E2
            ▼     T3 → StAR/CYP11A1 → luteal P4
        systemic T3
            │
            ▼
        nuclear TR-alpha/beta → transcription
                                  in granulosa, endometrium
                                  corpus luteum

Modulation by cortisol/HPA (explicit L2 → L3 bridge):

Chronically high cortisol (diurnal cortisol phenotype Phenotype A, L2.4)
    │
    ├── inhibits TRH-PVN ────────────► TSH ↓ flattened nocturnal pulse
    │
    ├── ubiquitinates DIO2 ─────────► tissue T3 ↓ (misleading “normal” TSH)
    │
    ├── upregulates DIO3 ───────────► rT3 ↑ (shift of T4 to inactive pathway)
    │
    └── inhibits endometrial HSD11B2 ─► active endometrial cortisol ↑
                                        competes with T3 for transcriptional
                                        co-activators (CBP/p300)

Resulting phenotype: woman with normal TSH, normal T4,
“hypothyroid” symptoms + “perimenopausal” symptoms + mildly
irregular cycle + shortened luteal phase — all simultaneous,
all correlated with elevated cortisol but none
“diagnosable” with standard panels.

Variability by DIO2 Thr92Ala polymorphism:

Ala/Ala (~13-15% pop)
    │
    ▼
DIO2 catalytically ↓ ~30%
    │
    ▼
Tissue T3 ↓ with serum T4 = (Ala) vs (Thr)
    │
    ├── Granulosa: attenuated T3+FSH synergy → functional FSH resistance
    ├── Endometrium: suboptimal receptivity
    ├── Corpus luteum: P4 ↓
    └── CNS: persistence of cognitive symptoms on LT4 monotherapy

The synthesized point: TSH is a “system pressure indicator,” not a “precise thermostat of thyroid status.” In L3 we will treat TSH as a global modulator of four reproductive fronts (pituitary-PRL, granulosa-aromatase, luteal-P4, endometrium-receptivity), where each front has its own layer of local regulation (thyrostimulin, DIO2 ubiquitination, TR polymorphisms, transcriptional competition with GR).

Cross-synthesis with previous findings

With L2.4 (adrenal allostasis): L2.4 documented that chronic cortisol inhibits TSH centrally and 5'-deiodinase peripherally. L3.1 expands this: a hyperreactive diurnal cortisol pattern predisposes to a peripheral hypo-functional thyroid profile with normal TSH — that is, "normal TSH + low T3 + hypo symptoms + anxiety/insomnia" emerges as a predictable biotype derived from the literature, without requiring direct measurement of cortisol or T3.

With L2.3 (P4 vs cortisol in GR, phase sensor): L2.3 established GR as a cycle phase sensor. L3.1 adds an additional competitor to the same node: nuclear T3 competes with cortisol for transcriptional co-activators (CBP/p300/SRC-1) in luteinized granulosa and endometrium. Under chronically high cortisol + low tissue T3, transcription of P4-dependent genes in the endometrium fails through a dual pathway: P4 substrate deficit (L2.3) and T3 modulator deficit (L3.1). This is new and connects L2 with L3 in a single molecular mechanism.

With L2.6 (adaptogens): L2.6 documented that ashwagandha (Withaferin A) recalibrates GR. There is emerging literature suggesting that ashwagandha also modulates TSH and T4 in a euthyroid direction (raises TSH and T4 in subclinical women, does not change them in euthyroid women). This must be examined critically in a future session (not in L3.1) — the risk is that we are extending an herbal mechanism to a different axis without a specific RCT. Pending: evaluate the evidence for ashwagandha → thyroid with the same RCT standard we applied in L2.6.

With L1.6 (gut-brain-ovary axis, neurobolome, vagal-tone phenotype): The gut microbiome modulates bacterial deiodination of T4 and T3 and enterohepatic reabsorption of T3-sulfate and T3-glucuronide. Dysbiosis → less enteric T3 recirculation → lower systemic T3 with stable TSH/T4. The vagal-tone phenotype (L1.6) could predict peripheral functional thyroid dysfunction without measuring it — hypothesis to test in L3.5.

Lua Labs Hypotheses

Hypothesis 19: “TSH as a global multi-node modulator — a symptomatically healthy menstrual cycle requires simultaneous coherence across four thyroid fronts, and the typical dysfunction in LATAM women is not ‘subclinical hypothyroidism’ but sub-threshold multi-nodal desynchrony”

Statement: In LATAM women with symptomatically irregular cycles and TSH within the “normal range” (0.5–4.0 mIU/L), the source of the symptomatic phenotype is not detectable TSH elevation, but the coexistence of sub-threshold thyroid-reproductive stressors: intra-cycle TSH variation, attenuated peripheral T4→T3 conversion, tissue-level DIO2 vulnerability, and lower enteric T3 recirculation.

Proposed mechanism:

Mid-cycle serum TSH (physiological Plowden 2018 peak) in a woman with high thyroid symptom load → transiently pushes toward the threshold of functional subclinical hypothyroidism → transient reduction of late follicular E2 + attenuation of the LH peak.
Chronic cortisol (diurnal cortisol phenotype Phenotype A, L2.4) → peripheral DIO2 ubiquitination → reduced tissue T3 in granulosa → CYP19A1/aromatase less responsive to FSH → delayed ovulation + corpus luteum with suboptimal steroidogenesis (bridge to L2.3 luteal insufficiency).
Reduced endometrial tissue T3 (local DIO2 + Ala/Ala polymorphism) → suboptimal expression of HOXA10, LIF, integrin αvβ3 → shortened implantation window → peri-menstrual spotting, possibility of unrecognized biochemical pregnancy loss.
Microbiome with low diversity (low dietary-diversity phenotype, L1.3) + post-antibiotic dysbiosis → reduced enteric T3 recirculation → low systemic T3 with stable TSH/T4.

Confidence level: Medium. Each individual link is documented in the literature. What is original to Lua Labs is the sub-threshold multi-nodal integration and the proposal that the aggregate phenotype is not capturable with point-in-time TSH. The main weakness: no RCT has tested the integrated model; the evidence is piece-by-piece.

How to validate:

Through a prospective sub-study: n ≈ 120 perimenopausal women (40–55 years), 90 days, with three TSH measurements (mid-follicular, mid-cycle, mid-luteal), fT3/fT4/rT3, four-point salivary cortisol on two occasions, and daily symptom and cycle logging. Compare the predictive capacity of a model based on symptomatic and temporal pattern against a single TSH measurement and against a complete thyroid panel. Prediction: the integrated model identifies the symptomatic phenotype better than point-in-time TSH alone.

Limitations:

The “microbiome → enteric T3 recirculation” link is the weakest. The literature is mechanistic animal evidence with limited human confirmation.
thyroid-symptom phenotype self-report is noisy. Thyroid symptoms massively overlap with perimenopause (Mintziori 2024) — the model requires co-adjustment by hormonal stage.
It does not measure local thyrostimulin in the ovary — a layer that exists mechanistically but that no clinical assay measures.

Candidate formulation (dietary + behavioral, ZERO pharmacological supplements)

Name: “Thyroid-Cycle Coherence Foundation” — extension of the Adrenal Allostasis Protocol L2.6 with a thyroid vertex.

Compounds / foods:

Controlled dietary iodine (not supplement): ~150 µg/day via iodized salt in Mexico (Mexican NOM mandates iodization at 20–40 mg/kg) + fish/seafood 2x/week + pastured dairy — do not exceed 300 µg/day (risk of worsening underlying Hashimoto, Teng et al. 2006).
Dietary selenium: 1–2 Brazil nuts/day (≈ 100–150 µg) or tuna/sardines 2x/week + daily organic egg + sunflower seeds. Mechanism: DIO1/DIO2 cofactor + thyroid glutathione peroxidase + anti-TPO reduction (Fan 2014 meta-analysis).
Zinc + iron: pumpkin seeds, black beans (already part of the Mesoamerican Ancestral Buffer from L1.4) — secondary deiodinase cofactors.
Continuous fermentable fiber: same L1.4-L1.5 foundation (nopal, beans, green plantain, nixtamal RS2) — supports enteric T3 recirculation.
Behavioral: avoid morning caffeine within 4 hours of levothyroxine intake (relevant only for those already on LT4 treatment); morning sunlight exposure 7–9 AM (synchronizes the circadian TSH rhythm); dinner → breakfast window ≥ 12 hrs (reduces nocturnal inflammation and supports DIO2); evening alcohol restriction (acute alcohol depresses TSH and T4).

Target population: Carmen (47, perimenopause) primary; Sofía (28, symptomatic irregular cycle with suspected PCOS) secondary; Rosa (55, postmenopause with persistent fatigue) tertiary — Rosa requires medical evaluation of TSH/T4 before any behavioral intervention, due to risk of undiagnosed post-menopausal Hashimoto (Mintziori 2024).

Complementary mechanisms: (1) iodine + selenium + zinc provide substrates+cofactors for T4→T3 production and conversion; (2) fermentable fiber maintains enteric T3 recirculation through microbiota; (3) circadian alignment + dinner-breakfast window reduces DIO2 ubiquitination by nocturnal cortisol/inflammation; (4) evening alcohol restriction preserves the physiological nocturnal TSH peak.

Regulatory status: 100% GRAS foods and behavioral practices. Zero pharmacological supplements. Zero therapeutic dose recommendations.

Requires validation: Prospective RCT n ≥ 80, 6 months, in women with high thyroid symptom load + TSH 1.0–4.0 mIU/L (normal-gray zone) and cycle irregularity. Compare control diet vs Thyroid-Cycle Coherence diet. Primary endpoint: ≥ 25% reduction in thyroid symptom load + cycle regularization (SD ≤ 3 days). Sub-analysis: stratification by diurnal cortisol phenotype.

Individual variability

Genetics:

TSHR polymorphisms: 22+1 variants linked to PCOS-like presentation (Prudente 2023, already cited in L2.2) that could modulate ovarian TSHR response. Cross-talk L2 ↔ L3.
DIO1 / DIO3 polymorphisms: less studied; high tissue DIO3 → T4 → rT3 diversion (Khouri 2025 suggests that phenotype in ~11% of women treated with LT4).
HLA-DR3, CTLA-4, PTPN22: Hashimoto susceptibility factors. L3.3 will explore in depth.

Epigenetics:

DIO2 promoter methylation in skeletal muscle modulated by insulin and chronic inflammation (bridge L3.4).
Endometrial TSHR methylation is understudied; open hypothesis that it predisposes to euthyroid recurrent implantation failure.

Environmental:

Iodine in LATAM: Mexico has iodized salt by NOM, but natural salt consumption patterns (without iodine) in rural areas + growing use of non-iodized “Himalayan salt” in the urban middle class → possible re-emergence of subclinical deficiency.
Selenium in LATAM: high soil variability. Andean soils high in selenium (Bolivia/Peru/volcanic Mexican highlands) vs deficient sedimentary soils (Yucatán). Not characterized.
Endocrine disruptors: BPA, phthalates, PCB, PFAS → all decrease serum TSH and T4 and compete for TR. L5 (hormonal epigenetics) will capture this.

Hormonal stage (interaction with thyroid-symptom phenotype):

Sofía (28, active cycle): TSH shows wider intra-cycle variation; mid-cycle peak bias is more visible.
Carmen (47, peri): maximal thyroid ↔ peri symptomatic confusion; TSH gradually rises with age (physiological “TSH-creep” effect).
Rosa (55, post): post-menopausal Hashimoto prevalence doubles; T4 declines physiologically; symptoms become chronic.