CDKN2B-AS1 Genotype (rs10757278)

Some people build arterial plaque faster than their cholesterol numbers say they should. The chromosome region called 9p21.3 is one of the best-studied inherited reasons. It was one of the first major genome-wide signals for heart attack, and the signal has replicated most clearly in European and East Asian cohorts. Related variants in the same locus have also turned up in Saudi and Russian studies.

rs10757278 is one of the marker variants in that region. It lies inside CDKN2B-AS1, also called ANRIL, a long noncoding RNA gene. Knowing your genotype won't tell you whether you have plaque today. It tells you which direction to lean when you decide how early and how hard to look.

What This Variant Actually Is

CDKN2B-AS1 doesn't code for a protein. It makes an RNA molecule that acts as a switch, dialing the activity of nearby genes up and down. Those neighbors, CDKN2A and CDKN2B, help control when cells stop dividing. In arteries, that matters because plaque grows partly when smooth muscle cells and immune cells in the vessel wall multiply too much.

The 9p21.3 risk variants don't change a protein's sequence. They appear to act partly by changing how much of this RNA switch gets made. Work tracing which gene at the locus the variants influence points to ANRIL itself: a much larger share of the variation in its activity traces back to nearby DNA differences than for either of the two protein-coding neighbors. That makes ANRIL the leading candidate for how the risk is transmitted, though the exact chain from DNA to plaque hasn't been pinned down.

One wrinkle matters. The direction of the effect isn't simple. In people with premature coronary disease, risk genotypes at this locus went with lower ANRIL in blood, not higher. In some tumor studies, higher ANRIL levels are the bad sign. This is a context-dependent regulator, not a dial where one direction is always worse.

Coronary Artery Disease and Heart Attack

This is the association the variant is known for. The original Icelandic discovery study, which combined more than 17,000 people across Icelandic and US samples, found that people carrying two copies of the 9p21 risk variant had roughly 1.6 times the odds of heart attack compared with non-carriers, with a bigger effect in people whose disease started early. A single copy shifted the odds less. That finding replicated fast and widely, and rs10757278 specifically was flagged as a susceptibility marker for heart attack in people of European descent and later in South Koreans whose coronary disease had been confirmed by artery imaging.

What the locus appears to drive is plaque itself rather than the clot that ends in a heart attack. A collaborative analysis of patients who had their coronary arteries imaged found 9p21.3 risk variants tracked with how much disease showed up in those arteries, but not with whether someone who already had coronary disease went on to have a heart attack. Read that as: the variant seems to load the artery more than pull the trigger.

In Saudi patients, different variants inside the same gene, not rs10757278, carried odds around 2.1 and 2.2 for coronary disease. Adding them to a clinical model built on age, sex, and body mass index pushed its discrimination from 0.79 to 0.87, a measure where 0.5 is a coin flip and 1.0 is perfect. That is a real gain in a research setting. It has not been shown to change what anyone should do.

The association isn't universal. A Dutch cohort of nearly 8,000 older adults followed prospectively found no association between two common 9p21 variants and coronary heart disease or heart attack. Studies in elderly cohorts can miss genetic effects tied to early events, because the most susceptible people may already have had those events before enrollment.

Coronary Calcium

The cleanest imaging clue that this locus tracks plaque in people who feel fine comes from calcium scanning. In 1,688 asymptomatic Koreans getting routine health checks, a nearby variant in the same gene as rs10757278 came out on top of a genome-wide scan for severe coronary calcification. Carriers of the risk allele had about three times the odds of severe calcification, and that held after accounting for age, sex, high blood pressure, and diabetes.

rs10757278 itself replicated as a calcium-associated marker in a multi-ancestry study of people with type 2 diabetes, a group already at high vascular risk. The Korean study was small in its discovery phase, drawn from a single center, and enrolled health-check examinees rather than people with diagnosed disease. Treat the threefold figure as a signal of direction, not a number to apply to yourself.

If you carry a risk genotype, this is the finding that can change something concrete: move a coronary calcium scan earlier than your age alone would suggest, especially if family history points the same way.

Stroke

The 9p21.3 signal extends to stroke, but the most coherent signal is in large-artery atherosclerotic stroke. That's the subtype caused by plaque in the big vessels feeding the brain. Earlier studies that lumped all strokes together got less consistent answers.

In 5,262 hypertensive participants with DNA available from a blood-pressure trial, variants at this locus predicted stroke independently of classical risk factors and independently of whether the person had a coronary event. A large prospective Chinese stroke population found variants here tracked with both first stroke and recurrence. Meanwhile a Belgian analysis found a strong coronary signal and no association with ordinary cerebrovascular disease, and a family-based study of children with arterial ischemic stroke found rs10757278 was not a risk factor at all. Childhood stroke has different causes; atherosclerosis usually isn't one of them.

Type 2 Diabetes

One recent Chinese Han case-control study linked the G allele of rs10757278 to higher odds of type 2 diabetes. In that study, people carrying two copies had higher fasting glucose and HbA1c than people carrying none. Serum levels of the ANRIL RNA itself, a different blood measurement from this genotype test, ran lower in people with diabetes than controls and distinguished the groups in that dataset.

That diabetes finding is early. The coronary signal is much better established. The variant's effect on arteries does not seem to run through standard lipids: a meta-analysis of 9p21.3 variants covering more than 100,000 people found no statistically clear lipid effect for rs10757278 in the studies that reported this particular variant. The mechanism appears to be vessel-wall cell behavior more than cholesterol handling.

Why a Clean Lipid Panel Doesn't Rule This Out

Here is the part that catches people. You might reasonably assume that a genetic heart risk variant would show up somewhere in standard bloodwork. This one usually doesn't. The lipid finding above is the direct evidence: this variant can move plaque risk without meaningfully moving cholesterol.

So a person with a clean lipid panel, normal blood pressure, and no symptoms can still carry two copies of the risk allele. That doesn't prove plaque is there. It means the next useful question is anatomical, not biochemical. A calcium score or coronary CT angiogram can show whether inherited risk has become actual plaque.

Other Conditions Linked to This Region

The 9p21.3 region shows up in genome-wide scans for several conditions, which makes sense for a gene that helps regulate when cells stop dividing.

• Psoriasis: in a single case-control study of 586 people, the same G allele of rs10757278 carried about 1.9 times the odds of psoriasis.
• Glaucoma: variants in this gene correlate with optic nerve damage in people with primary open-angle glaucoma in both US and Japanese populations, and the alleles that shift disease risk also shift the shape of the optic nerve head.
• Cancer: other variants in this gene are tied to pancreatic cancer risk, and higher activity of the gene predicts worse survival in several cancers including liver, kidney, and endometrial. These are different variants and different tissues; rs10757278 specifically is a cardiovascular marker.

Don't read the cancer findings as applying to your genotype. Most concern either gene activity in tumor tissue or different variants at the same locus. The evidence tying rs10757278 to your health is mainly cardiovascular, with early metabolic data.

A One-Time Result You Use for Decades

Your genotype was set at conception and will read the same in ten years. There is no trend to track and no reason to repeat the test unless the call itself is in doubt. The value is entirely in what you do with it afterward.

What does need tracking is the phenotype. If you carry a risk genotype, the case for a coronary calcium scan in your forties gets stronger, and for a coronary CT angiogram if you have other risk factors stacked on top. ApoB and Lp(a) should be in your baseline regardless, since the 9p21.3 pathway runs alongside lipid risk rather than through it. Repeat imaging on a several-year cadence if your first scan shows anything, and keep annual lipid and inflammation markers as your near-term feedback loop.

What to Do With a Risk Genotype

A risk result is not a diagnosis and should not send you to the emergency room. It should change your threshold for looking. The pathway runs roughly like this: confirm the finding if the genotype came from a consumer chip rather than a clinical-grade lab, then order imaging rather than more blood tests, because imaging is what tells you whether the genetic risk has become actual plaque.

Pair the genotype with your full lipoprotein picture, ApoB and Lp(a) especially, and with hs-CRP as a read on vascular inflammation. The combination that warrants real action is a risk genotype plus any measurable calcium plus elevated ApoB. A risk genotype alone with a zero calcium score and clean lipids is a reason to keep watching on a tighter schedule, not to start treatment. If you have a family history of heart attack before 55 in men or 65 in women, bring the result to a preventive cardiologist or lipidologist; that combination is where 9p21.3 information has most consistently predicted early disease.

What This Test Can and Cannot Tell You

This is a research-grade risk marker. No guideline recommends genotyping asymptomatic people for it, and no trial has shown that knowing your result leads to earlier detection or better outcomes.

The best-designed test of its clinical value was blunt. In the Women's Genome Health Study, over 22,000 women were followed prospectively; 9p21.3 genotype was associated with incident cardiovascular disease, but adding it to traditional risk factors plus hs-CRP plus family history did not improve risk prediction. The signal is real. It is largely already captured by things you can measure more cheaply.

That is an argument against population screening, not against knowing your own result. If you are already ordering advanced lipid and imaging work and want to know which direction to lean on timing, the genotype is one durable input you will never have to buy again. Just don't expect it to override what your arteries actually show.

Where the Result Can Mislead

• Panel coverage: the assay reads the specific variant it was built to read. A non-risk result at rs10757278 says nothing about other variants in the same region or rarer inherited causes of early heart disease.
• Ancestry: risk allele frequencies differ across populations, and most effect estimates come from European and East Asian cohorts. The direction of risk has replicated broadly; the exact size has not been established for every group.
• Consumer chip versus clinical lab: direct-to-consumer reports use different platforms and occasionally miscall single positions. If a consumer result is driving a real decision, confirm it by a different method.
• Variant reports: a broader sequencing report may list a variant of uncertain significance near this region. That is not the same thing as carrying this known common risk allele.
• Germline versus tumor testing: this result should come from inherited DNA, such as buccal cells, saliva, or blood. Tumor-only sequencing can contain acquired changes and should not be used to infer this inherited result.
• Common variant, modest effect: this is a frequent variant that nudges the odds, not a rare mutation that dictates a disease. Plenty of people who carry the risk allele never develop significant plaque.