rs2943634 — IRS1

Intergenic variant near IRS1 at 2q36.3 associated with ischemic stroke risk, HDL cholesterol, and adiponectin levels; the A allele is protective

See your personal result for IRS1

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.

The 2q36.3 Stroke Locus — IRS1 Neighborhood Variant and Ischemic Stroke Risk

The rs2943634 variant sits in an intergenic region of chromosome 2q36.3, approximately 528 kilobases upstream of the IRS1 gene11 IRS1 gene
insulin receptor substrate 1, a central adaptor protein in insulin and IGF-1 signaling. Despite lying outside any protein-coding sequence, rs2943634 was independently replicated in genome-wide association studies for coronary artery disease22 genome-wide association studies for coronary artery disease
including the landmark Samani et al. 2007 NEJM GWAS and has since been shown to have a broader cardiovascular footprint including ischemic stroke, HDL cholesterol, and adiponectin.

The Mechanism

The SNP does not change any protein sequence. Instead, it likely acts as a regulatory variant33 regulatory variant
altering gene expression of nearby genes through enhancer elements or chromatin remodeling. The nearest candidate gene is IRS1, encoding a key scaffold protein that transmits insulin and IGF-1 receptor signals to downstream metabolic pathways including PI3K-Akt and MAPK. IRS1 activity influences adiponectin secretion by adipocytes44 adiponectin secretion by adipocytes
a cardioprotective adipokine that enhances insulin sensitivity, suppresses endothelial inflammation, and promotes reverse cholesterol transport.

The A allele is associated with higher circulating adiponectin and HDL cholesterol in an additive, dose-dependent manner. Crucially, when Arregui et al. adjusted for both adiponectin and HDL simultaneously55 Arregui et al. adjusted for both adiponectin and HDL simultaneously
in mediation analyses within the EPIC-Potsdam cohort, the stroke association persisted, suggesting the variant also engages independent protective mechanisms beyond these two biomarkers.

The Evidence

The 2q36.3 locus was first tagged in the Samani et al. 2007 NEJM genome-wide association study66 Samani et al. 2007 NEJM genome-wide association study
a joint analysis of the Wellcome Trust Case Control Consortium and the German MI Family Study, achieving genome-wide significance for coronary artery disease in European populations. The Coronary Artery Disease Consortium 200977 Coronary Artery Disease Consortium 2009
a large-scale multicenter replication study confirmed the locus in independent cohorts.

The most detailed functional analysis came from Arregui et al. 201288 Arregui et al. 2012
a nested case-cohort within the EPIC-Potsdam prospective study of 27,548 middle-aged German adults followed for a mean 8.2 years. The A allele was associated with [a 34% reduction in ischemic stroke risk | HR=0.66, 95% CI 0.50-0.87, P=0.003] in an additive model, while showing no significant association with myocardial infarction (HR=1.02, P=0.83). Adiponectin concentrations rose stepwise: CC 6.94, CA 7.27, AA 7.86 μg/ml (P=0.0002). HDL cholesterol followed the same pattern: CC 52.1, CA 53.1, AA 55.3 mg/dl (P=0.002).

In a broader cardiovascular context, the MORGAM prospective cohort study across N=33,282 participants99 MORGAM prospective cohort study across N=33,282 participants
spanning multiple European countries, with 571 incident stroke events also linked rs2943634 to HDL cholesterol and blood pressure. The Dlouha et al. 2025 study in heart transplant recipients1010 Dlouha et al. 2025 study in heart transplant recipients
examining accelerated coronary disease post-transplant found the C allele associated with OR 2.46 (95% CI 1.12-6.17) for early cardiac allograft vasculopathy, an accelerated form of CAD that develops uniquely in transplanted hearts.

A 2012 metabolic syndrome study van der Valk et al. in EPIC-NL (N=1,886)1111 van der Valk et al. in EPIC-NL (N=1,886)
testing IRS1 locus variants and metabolic syndrome components found the A allele associated with OR 0.88 (95% CI 0.79-0.97) protection against metabolic syndrome, with significant associations across HDL, triglycerides, and HbA1c.

The evidence is replicated across European cohorts but inconsistently in some smaller studies — the strongest and most consistent signal is for ischemic stroke and HDL cholesterol. Effect sizes are modest (common variant with population-attributable effects), fitting the polygenic architecture of cardiovascular disease.

Practical Implications

Carrying one or two C alleles does not determine your fate — this is one of many common variants that collectively elevate cardiovascular risk. The clinical relevance lies in recognizing that [this genotype lowers circulating adiponectin and HDL | both biomarkers with direct anti-atherogenic and anti-inflammatory effects on the vascular wall]. People with the CC genotype may need to work harder to achieve the same HDL levels as AA carriers, and should be more proactive about monitoring vascular risk factors.

Interactions

rs2943634 co-occurs with other cardiovascular GWAS loci. The chromosome 9p21.3 locus (rs10757274)1212 chromosome 9p21.3 locus (rs10757274)
one of the strongest genetic risk factors for coronary artery disease, near CDKN2A/B was studied alongside rs2943634 in the cardiac transplant study, and the two loci appeared to have additive effects on early allograft vasculopathy. The 6q25.1 locus (rs6922269)1313 6q25.1 locus (rs6922269)
an MRAS gene variant associated with CAD has also been grouped with rs2943634 in multi-locus cardiovascular risk studies.

Genotype Interpretations

What each possible genotype means for this variant.

AA Protective Genotype Beneficial

Lower stroke risk — protective variant at the 2q36.3 locus

The AA genotype sits at the favorable end of the adiponectin and HDL spectrum for this locus. Higher adiponectin has multiple cardioprotective effects: it enhances insulin sensitivity, suppresses endothelial NF-κB activation, and promotes macrophage polarization away from the pro-inflammatory phenotype that drives plaque formation. The stroke protection was independent of adiponectin and HDL in mediation analyses, suggesting additional vascular mechanisms. Note that AA at this locus does not eliminate cardiovascular risk — it simply means this particular genomic region is not contributing adversely.

AC Intermediate Risk Intermediate

Modestly reduced adiponectin and HDL at the 2q36.3 locus

The additive pattern of this variant means one C allele already shifts adiponectin and HDL in the unfavorable direction, though more modestly than two copies. The stroke hazard ratio for CA individuals lies between AA and CC — not independently reported in the available studies, but consistent with the additive model. This genotype is common enough that cardiovascular risk stratification should focus on traditional factors and biomarkers rather than this single SNP alone.

CC Risk Genotype High Risk

Higher stroke risk — homozygous for the 2q36.3 risk variant

The CC genotype is associated with reduced adiponectin secretion, a cardioprotective adipokine whose deficit is linked to endothelial dysfunction, increased vascular inflammation, and impaired reverse cholesterol transport. The stroke risk from this variant appears independent of HDL and adiponectin levels in mediation analyses, suggesting additional mechanisms — possibly involving IRS1-mediated insulin signaling in vascular smooth muscle or platelet function. A 2025 transplant study found CC individuals had OR 2.46 for early coronary artery disease post-transplant, a setting that amplifies underlying genetic vulnerabilities. The variant appears stroke-specific: no association with myocardial infarction was found in the EPIC-Potsdam study, suggesting the mechanism may preferentially affect the cerebrovascular bed.