Heart and blood vessels

What does actually look at the heart

None of this is on our menu, and most of it is not on the NHS unless you have symptoms. It is all privately available, and if you can afford it the information is genuinely good.

A coronary calcium score is a low-dose CT scan, a few minutes, no contrast dye and no needle. It counts the calcified plaque in your coronary arteries and gives you a number. A score of zero in someone without symptoms predicts a very low event rate for years afterwards, and a high score reclassifies your risk upwards more reliably than any calculator does. It is one of the few tests that shows you the disease itself rather than a risk factor for it.

But calcium is the old, settled, stabilised plaque. It is the scar rather than the wound. The soft plaque described above does not show up on it. That is the kind with a thin cap and a lipid core, the kind that tears, and it contains little or no calcium. Studies of people who scored zero on calcium have still found soft plaque on better imaging in around 1 in 10, and in some it was already narrowing the artery.

That matters because the soft plaque is the dangerous kind. So a zero calcium score is reassuring, not conclusive.

CT coronary angiography is the scan that sees it. Contrast dye, more radiation, more money, and it shows both the calcified and the non-calcified plaque along with how much of the channel is narrowed. Newer analysis software goes further and quantifies how much soft plaque you are carrying and what it is made of. Guidance reserves it for people with symptoms, largely on cost and radiation grounds rather than because the information is poor.

And there is a cheaper option. Carotid intima-media thickness is an ultrasound of the arteries in your neck. No radiation, no dye, considerably cheaper, and it measures how thick the artery wall has become. It is not the coronary arteries, but the same disease runs in both, so it gives you an early view of whether the process has started.

So a blood test tells you what is driving the process, and imaging tells you how far it has already got. They answer different questions, and the second one is not something anyone will offer you while you feel well.

What is actually happening

Arteries are lined with a single layer of cells called the endothelium. Particles carrying cholesterol cross that lining constantly, in everybody, and most of them pass back out again. That is totally normal.

The disease starts when they do not leave. Some particles bind to the proteoglycans, the mesh of sugar-coated proteins that forms the scaffolding inside the artery wall, and are held there. Once trapped, they are exposed to oxygen for longer than if they were in the bloodstream, and they oxidise. This is what changes them into something the immune system reads as damage rather than simple cargo.

This results in a clean-up operation that is doomed and cannot finish. Monocytes are recruited out of the blood into the wall, become macrophages, and begin eating the trapped particles. They have no mechanism for stopping, so they fill with fat until they are barely functional. These are foam cells, named for how they look. Foam cells die where they sit, and what they were holding is left behind. Over years that builds a soft core of debris, and smooth muscle then produces a fibrous cap over the top of it.

There are two futures for that plaque, and the dangerous one is not the obvious one. It can grow slowly and narrow the channel, which produces symptoms on exertion as blood supply becomes limiting. Or the cap can tear, exposing the core to flowing blood, which clots on contact. Most heart attacks come from the second route, and often from plaques that were not narrowing the artery very much at all. Which is why a test for narrowing and a test for the process are different questions.

Three features of that process decide what is worth measuring. It is driven by the number of particles crossing rather than the total weight of cholesterol they carry. It involves inflammation as much as it involves fat. And it runs for decades without producing a single symptom.

What the markers describe

ApoB counts the particles crossing into the wall, one molecule per particle, which is the number the process actually responds to. Lipoprotein(a) is an inherited variant that behaves as an extra and more troublesome particle. hs-CRP covers the inflammatory half. HbA1c matters because insulin resistance accelerates the whole thing. The lipid markers in the section below describe the transport system these particles belong to.

What is missing, and it matters

Blood pressure is one of the largest contributors to cardiovascular risk and it is not a blood test. Neither is smoking, diet, family history, age or body fat.

And blood pressure is where UK guidance openly prescribes differently by family origin. Very little in UK practice changes with ethnicity. This does. NICE's hypertension guideline, NG136, says "When choosing antihypertensive drug treatment for adults of Black African or African-Caribbean family origin, consider an angiotensin II receptor blocker (ARB), in preference to an angiotensin-converting enzyme (ACE) inhibitor." At the first step it goes further still. Somebody of Black African or African-Caribbean family origin without type 2 diabetes starts on a calcium-channel blocker at any age, where a white person under 55 would start on an ACE inhibitor or a receptor blocker.

The phenotype behind that is real and the explanation usually attached to it is not. High blood pressure in people of African background runs, on average, more salt-retaining and more dependent on blood volume, with lower levels of renin, the kidney enzyme those 2 drug classes act downstream of. But measuring a person's renin does not predict which drug will lower their pressure, and NICE gives the recommendation without publishing a mechanistic rationale of its own. The recommendation is also being argued with. English primary care records published in 2024 found receptor blockers going with more cardiovascular deaths than ACE inhibitors in Black patients, while angioedema, the swelling reaction that is the usual reason to avoid an ACE inhibitor, was less common on the receptor blocker. The guidance has not changed. NG136 says nothing different for South Asian people, who are routed by age and diabetes status exactly as a white person is.

How common high blood pressure is does vary a lot by group. Once age is accounted for it is highest in Black Caribbean, Black African and Pakistani adults, and lowest in Chinese and Other White adults, though the national figures behind that are published as experimental and a precise percentage for any single group is not one to quote with confidence. Older national survey data gives the shape more usefully. Black men were about twice as likely as white men to have high blood pressure, and South Asian men about as likely as Black men, while South Asian women were not raised at all and sat close to white women. That sex split is consistent and it gets dropped almost every time the claim is repeated.

Salt sensitivity is the part most often overstated. On a typical Western diet, blood pressure in people of African background does rise more for a given rise in salt intake. But when potassium intake comfortably exceeds current dietary targets the difference stops being detectable, and Black adults on average eat less potassium, so what gets read as an ethnicity effect may be a potassium effect. The explanation many people were taught, that salt retention was selected for during the transatlantic slave trade, has been rejected on both historical and physiological grounds and should not be repeated.

The NHS provides a target. The tool recommended for adults aged 25 to 84 without existing cardiovascular disease is QRISK3, and a 10-year score of 10% or more is the point at which a statin is offered. What QRISK3 combines is mostly things no blood test can see: age, sex, ethnicity, postcode, smoking, blood pressure, body mass index, family history, diabetes, kidney disease, atrial fibrillation. It also counts migraine, severe mental illness, lupus, erectile dysfunction, and treatment with steroids or certain antipsychotics. Cholesterol enters as a single ratio. A newer version called QR4 has been published and adds chronic obstructive pulmonary disease, Down's syndrome, learning disability and brain, lung, blood and oral cancer to that list, along with pre-eclampsia and postnatal depression in women, but it is not yet in the guideline. Scotland uses a different tool called ASSIGN, which builds in social deprivation by postcode and counts family history explicitly.

Ethnicity does far more inside QRISK3 than a single tick box. It is not a white and non-white switch. The tool takes 9 self-assigned categories and weights them separately, and those terms carry more weight than most of the other variables in the list. South Asian origin raises the calculated risk by roughly a third to three quarters, depending on the group and the sex, against an otherwise identical white profile. An uplift of that size can move a 10-year risk of about 6% up to about 10%, which is the line at which a statin is offered, on family origin alone.

The categories that lower the score are the ones almost nobody expects. Black Caribbean, Black African and Chinese origin all cut the calculated risk, by up to about a third. A 10% risk becomes roughly 7%. Why that is, and why it is not a flaw, is further down this page.

And the reference group is not white. It is white or not recorded. Ethnicity was missing for about 4 in 10 people in the records the model was built from, and anybody with no ethnicity coded in their GP record is scored as though they were white. So if you are of Pakistani origin and your record carries no ethnicity code, the uplift silently does not happen. An uncoded record is not a neutral default, it is a white default.

It also has boundaries. QRISK is not used for people with type 1 diabetes, chronic kidney disease, or familial hypercholesterolaemia. All three already carry enough risk that a general calculator would underestimate it and could wrongly reassure, so each is assessed on the condition itself rather than on a calculated percentage.

Having ethnicity in the model is not the same as the model being right for everybody. The previous version, QRISK2, predicted about half the risk South Asian women went on to have when it was checked against a multi-ethnic London population. QRISK3 rebuilt the ethnicity terms on far more records and much better ethnicity coding, but nobody has gone back and remeasured it. The 10% treatment line is the same for everybody whatever goes in.

A panel of markers is an input to that picture. It is not the whole picture, and nothing we sell produces a risk score. You can run QRISK3 yourself, since it is free and public, and the only 2 things in it you cannot fill in from memory are your blood pressure and your cholesterol. That is the real relationship between a blood test and a risk score. The test supplies 2 of the inputs and you supply the rest. Anyone telling you a blood test alone has established your cardiovascular risk is overselling it.

South Asian cardiovascular risk is real, and the figure you have probably heard is wrong. Coronary heart disease is roughly 70% more common in UK South Asian people than in white Europeans, and differences in body shape account for only part of that. The comfortable explanation, that it comes down to diabetes and central fat, does not survive UK data. Diabetes was more than 3 times as common in the South Asian group and the measured metabolic risk factors still did not explain the gap in coronary disease.

The excess is not spread evenly either. Pakistani and Bangladeshi origin carry materially more of it than Indian origin, which is why QRISK3 weights them separately.

The line about heart attacks arriving 10 years earlier does not hold. No UK source supports it. In UK data the average age at a first coronary event is about 2.3 years younger in South Asian people than in white Europeans. The larger figure usually quoted, closer to 6 years, comes from comparing South Asian countries with other countries rather than UK populations, and the same work puts it down to risk factors accumulating earlier rather than to anything intrinsic.

There is a genuine large age gap, and it is in the brain rather than the heart. British South Asian people have a first ischaemic stroke about 7 years earlier than white British people, and allowing for the usual stroke risk factors widens that gap instead of closing it. So the earlier-events claim is roughly right for stroke and several times too big for the heart.

In UK data Black people are at lower coronary risk, not higher. QRISK3 lowers the calculated cardiovascular risk of Black African and Black Caribbean patients by about a third against an otherwise identical white profile, and that is not a flaw in the tool. It is what the primary care records it was built from show. UK cohort data agrees. Coronary heart disease was about 35% less common in African Caribbean than in white European participants, the average age at a first coronary event was about 2 years older rather than younger, and the lipid profiles were more favourable rather than worse.

The excess in Black British populations is real and it sits in blood pressure and stroke. Stroke overall runs about 50% higher than in white Europeans in UK cohort data. Strokes caused by bleeding are where the large numbers sit, and they are also where Black African and Black Caribbean stop behaving alike. Bleeding into the brain tissue itself is close to 3 times more common in Black African people than in white people, and about 1.5 times more common in Black Caribbean people, with both gaps far wider under the age of 65.

Diabetes moves this more than ethnicity does. In people who already have diabetes, stroke risk in South Asian and African Caribbean groups runs at roughly twice that of a white European with diabetes, while in people without diabetes the ethnic difference is small. And atrial fibrillation, the other common cause of stroke, is less frequent in Black and South Asian people in the UK rather than more, so it is not what is driving the stroke excess.

So a cardiovascular conversation with a Black British patient built around heart attacks is aimed at the wrong target. Blood pressure and stroke are the target, and coronary disease is the part of the picture where the risk is lower.

National guidance is explicit that before an abnormal lipid result is treated as a primary problem, the secondary causes should be looked for and corrected. This includes things like excess alcohol, poorly controlled diabetes, an underactive thyroid, liver disease, and kidney disease that is losing protein. Every one of those is covered somewhere else in this guide, and an underactive thyroid is a recognised cause of a high cholesterol result on its own.