rs4704397 — PDE8B PDE8B TSH-associated variant

Intronic variant in PDE8B (phosphodiesterase 8B) strongly associated with serum TSH levels; each A allele raises TSH by ~0.13 mIU/L through reduced cAMP degradation in thyroid tissue, with consequences for ovulatory function, implantation, and pregnancy maintenance

See your personal result for PDE8B

Upload your DNA data to find out which genotype you carry and what it means for you.

Upload your DNA data

Works with 23andMe, AncestryDNA, and other DNA test exports. Results in under 60 seconds.

PDE8B — When a Thyroid Brake Pedal Sits Too Heavy

The thyroid gland operates on a simple feedback loop: the pituitary releases TSH (thyroid-stimulating hormone), TSH binds receptors on thyroid follicular cells, and those cells respond by producing T4 and T3. The signal amplifier inside that thyroid cell is cyclic AMP (cAMP) — and the enzyme that silences it once it has done its job is phosphodiesterase 8B, encoded by PDE8B. A common variant near this gene, rs4704397, subtly impairs the cAMP-dampening step, leaving TSH signaling running slightly hotter and the thyroid tuned to a higher setpoint. The result is a persistent upward shift in TSH levels — small enough to be invisible on routine screening, large enough to matter for ovulation, implantation, and the earliest weeks of pregnancy.

The Mechanism

PDE8B11 PDE8B
Phosphodiesterase 8B; located at chromosome 5q13.3; encodes a high-affinity, rolipram-insensitive cAMP-specific phosphodiesterase highly expressed in thyroid tissue is the dominant cAMP-clearing enzyme in thyroid follicular cells. When TSH binds its receptor, cAMP floods the cell, activating PKA and triggering thyroid hormone synthesis. PDE8B's job is to hydrolyze that cAMP signal and restore the cell to baseline — essentially the "off switch" for TSH stimulation.

The rs4704397 variant lies in intron 1 of PDE8B and is thought to reduce gene expression or alter splicing efficiency, leading to less PDE8B activity. With less cAMP clearance, the thyroid follicular cell remains mildly hyperstimulated for longer, and the pituitary senses a slightly lower free T4 output. To compensate, the pituitary releases more TSH — enough to normalize free T4 into the reference range, but resulting in a persistent TSH elevation that is genetically encoded rather than driven by autoimmune damage, iodine deficiency, or frank hypothyroidism.

This mechanism was first established in 4,300 Sardinians with replication in 4,158 additional individuals22 first established in 4,300 Sardinians with replication in 4,158 additional individuals
Arnaud-Lopez et al., Am J Hum Genet 2008; discovery p=1.3×10⁻¹¹; overall replication p=1.9×10⁻²⁰; effect of 0.13 mIU/L TSH increase per A allele. Crucially, the association disappears entirely in people taking levothyroxine replacement — confirming the effect operates upstream at the pituitary-thyroid axis, not at end-organ level.

The Evidence

The foundational genome-wide association study by Arnaud-Lopez et al. (2008)33 Arnaud-Lopez et al. (2008)
Am J Hum Genet; n=8,458 total; discovery in 4,300 Sardinians, replication in Tuscans and Old Order Amish; rs4704397 in intron 1 of PDE8B; per-A-allele TSH increase of 0.13 mIU/L established rs4704397 as one of the strongest genetic determinants of serum TSH levels identified at that time, with effect sizes comparable to or larger than most subsequently discovered loci.

A meta-analysis of four cohorts totaling 2,557 participants44 meta-analysis of four cohorts totaling 2,557 participants
Taylor et al., Eur J Endocrinol 2011; associations longitudinally stable across 13 years (1981 vs 1994 measurements); A allele associated with TSH increase of 0.20 standard deviations per allele (p=1.64×10⁻¹⁰) and free T4 decrease of 0.07 standard deviations (p=0.023) confirmed that the TSH-raising effect is durable over time — not a transient phenomenon — and accompanied by a reciprocal, modest reduction in free T4.

The pregnancy implications were quantified in a cohort of 970 pregnant women assessed at 28 weeks55 970 pregnant women assessed at 28 weeks
Shields et al., J Clin Endocrinol Metab 2009; AA genotype median TSH 2.16 vs 1.73 mIU/L in GG; at the 4.21 mIU/L upper reference limit, AA women were 2.7× more likely to exceed it (9.6% vs 3.5%, p=0.004). This means that standard population-derived reference intervals classify a substantial proportion of genetically high-TSH women as subclinically hypothyroid even when their TSH reflects a genetic setpoint rather than thyroid disease — an important nuance for clinical interpretation.

The fertility connection was established in a case-control study of infertile subclinically hypothyroid women66 case-control study of infertile subclinically hypothyroid women
Mansuri et al., Int J Fertil Steril 2020; 60 infertile SCH cases vs 74 controls; AA genotype OR=3.84 (95% CI 1.86–8.01, p=0.0001) for infertility; A allele significantly overrepresented in cases (p<0.0001). The A allele appears to mark a subgroup of women whose high-normal or mildly elevated TSH reflects a fixed genetic setpoint, yet that setpoint is still high enough to impair the hormonal milieu of ovulation and implantation.

Separately, the Tromsø Study77 Tromsø Study
Jorde et al., Thyroid 2014; 8,938 participants without thyroid disease; AA genotype 0.29 mIU/L higher TSH than GG; AA also associated with HR 1.14 for myocardial infarction (95% CI 1.00–1.29) extends the clinical picture beyond fertility: the same TSH-elevating mechanism that impairs reproduction may mildly elevate cardiovascular risk over a lifetime.

Practical Implications

The thyroid-fertility link is well-established in clinical endocrinology. Hypothyroidism — even subclinical — impairs ovulation through effects on sex hormone-binding globulin, prolactin secretion, and LH pulsatility. Implantation depends on adequate free T4 in the endometrium, and early placental development is exquisitely sensitive to maternal thyroid status. Current fertility guidelines (ESHRE, ATA) recommend TSH below 2.5 mIU/L before conception and in the first trimester.

For AA carriers, TSH levels between 2.5 and 4.0 mIU/L may represent a genetically elevated setpoint rather than early thyroid failure — but the biological consequences for egg quality, luteal phase, and implantation may be the same regardless of cause. The practical response is monitoring TSH before and during conception attempts, maintaining awareness that "normal range" TSH may not be the optimal target for this genotype, and bringing genotype data to any infertility or preconception evaluation.

Interactions

PDE8B rs6885099 and rs2046045: These two additional PDE8B intronic variants are in moderate linkage disequilibrium with rs4704397 and were examined in the Yang 2015 pregnancy cohort (PMID 25822812). All three independently associated with subclinical hypothyroidism risk in pregnant women, suggesting a haplotype-level effect on PDE8B expression. Individuals carrying the full high-TSH haplotype (A at rs4704397 + non-reference alleles at rs6885099 and rs2046045) show the strongest TSH elevation. This compound haplotype is a candidate for a consolidated monitoring action but has not yet been separately analyzed for fertility outcomes.

Thyroid autoimmunity (TPO antibodies): The TSH-raising effect of the A allele is genetically mediated and independent of thyroid peroxidase antibody (TPOAb) status. However, carriers who also have elevated TPOAb face a compounded risk: the baseline TSH is already shifted upward by PDE8B genetics, and autoimmune thyroiditis pushes it further. Women with AA genotype and positive TPOAb represent a high-priority group for preconception TSH optimization.

Genotype Interpretations

What each possible genotype means for this variant.

GG Optimal TSH Setpoint Normal

Two G alleles — lowest genetic TSH setpoint, no PDE8B-related fertility impact

You carry two copies of the G (reference) allele at rs4704397 near PDE8B. This genotype is associated with the lowest genetically determined TSH setpoint at this locus — your thyroid's pituitary feedback loop operates at the most favorable baseline for this variant. Approximately 31% of people globally share this genotype; it is most common among Europeans (~38%) and least common in East Asian and African populations (~3–4%).

Your TSH levels at this locus are not influenced by the PDE8B-related mechanism, meaning this particular genetic factor is not contributing to any tendency toward elevated TSH or subclinical hypothyroidism.

AG Mildly Elevated TSH Setpoint Intermediate

One A allele — modestly raised genetic TSH setpoint; worth monitoring around fertility decisions

The TSH effect of the A allele is additive: one copy shifts TSH by approximately 0.13 mIU/L on average (Arnaud-Lopez 2008, PMID 18514160), with heterozygotes occupying an intermediate position. In a meta-analysis of 2,557 participants (Taylor 2011, PMID 21317282), the per-allele effect corresponded to 0.20 standard deviations in TSH and a reciprocal −0.07 SD in free T4.

For women trying to conceive or undergoing IVF, TSH values in the 2.5–4.0 mIU/L range that were previously considered acceptable are now recognized as potentially impairing — associated with reduced implantation rates and higher early pregnancy loss. This genotype warrants proactive TSH checking rather than waiting for symptoms.

AA Highest TSH Setpoint High Risk

Two A alleles — highest genetic TSH setpoint; fertility evaluation recommended if conceiving

The AA genotype at rs4704397 produces the strongest PDE8B-mediated TSH elevation. The mechanism: reduced PDE8B expression in thyroid follicular cells means cAMP is cleared more slowly after TSH stimulation, the thyroid cell remains mildly activated, and the pituitary compensates by raising its TSH setpoint to maintain adequate thyroid hormone output. The result is a right-shifted TSH distribution: AA individuals have normal thyroid function in the sense that free T4 is maintained within the reference range, but the cost is a persistently elevated TSH.

The fertility-TSH connection operates through at least three routes: (1) TSH-induced prolactin co-secretion can suppress GnRH pulsatility and impair LH-driven ovulation; (2) reduced free T4 availability — even within the reference range — reduces endometrial receptivity and trophoblast invasion; (3) adequate maternal T4 is the sole source of thyroid hormone for the embryo during the first 10 weeks before fetal thyroid function is established. For AA women attempting conception, the goal is to bring TSH below 2.5 mIU/L before conception and maintain it there through at least the first trimester.

The Tromsø Study also identified an AA-associated 14% higher myocardial infarction hazard ratio (HR 1.14, 95% CI 1.00–1.29) — a modest signal indicating the TSH-elevating mechanism may have long-term cardiovascular consequences independent of fertility.