The CAC score and the power of zero: The heart test that measures the disease, not the risk

A CAC score counts the calcified plaque in your heart's arteries. A score of 0 means the scan found none. It comes from a quick, low-dose CT scan with no contrast dye. You walk out with one number: the Agatston score, named after the doctor who came up with the scoring method back in 1990 ^[14]. Of all the tests in prevention cardiology, this one gets you closest to looking the disease straight in the eye.

Why does that matter so much? Most heart tests measure a risk factor: your cholesterol, your blood pressure, your age. A CAC scan measures the plaque actually sitting in your arteries right now. You are not estimating the odds. You are counting the damage.

The results come in bands. Zero is the one everyone hopes for:

The 2026 guideline that replaced the 2018 one goes further and uses your score to set how hard to treat ^[20]. More on that below.

So what does a zero get you? A long warranty. In the MESA study, researchers tracked 3,116 people who started at a CAC of 0 ^[1]. The "warranty period" on that zero (basically how long until about 1 in 4 people with a zero show some calcium, which tells you when a rescan becomes worth it) ran roughly 3 to 7 years, depending on sex and ethnicity. With diabetes it was shorter ^[1]. Even a full decade later, only about 8 percent had climbed past a CAC of 100. When you start from nothing, plaque builds slowly.

For survival, the picture is even longer. In a separate study of 9,715 people without symptoms, a zero kept the yearly death rate below 1 percent for almost 15 years in those at low or intermediate risk ^[15].

That is the power of zero. A CAC of 0 is one of the strongest "negative risk markers" in all of cardiology ^[10]. That means a result that actively pushes your risk down, not one that just fails to find a problem. It drops your short-to-mid-term risk hard. And it means a rescan every 3 to 7 years beats anxious yearly testing.

One caveat before we go on: a zero is not the same as zero risk. It tells you calcified plaque is absent today, not that nothing can ever go wrong. Soft plaque, smoking, diabetes: none of those show up in this number. We will get to all of it. For now, the headline holds. This is one of the best-evidenced longevity tests you can take that does not involve a supplement, because it measures the disease itself.

A high CAC score pushes your risk sharply upward. It changes the whole statin conversation. The bigger the number, the louder the alarm.

Take the extreme end first. In MESA, 257 people landed at a CAC of 1000 or higher. Compared with people at zero, even after the math accounts for the usual risk factors:

• Their risk of any cardiovascular disease ran 4.71 times higher.
• Their risk of coronary heart disease (clogged arteries feeding the heart itself) ran 7.57 times higher ^[2].

At a CAC of 1000, the study estimated about 3.4 major heart events (heart attack, stroke or cardiovascular death) per 100 people per year ^[2]. That is about the same as in a big trial of people who had already had a heart attack or stroke and were on a statin. A scan can show you are in that group before a single symptom shows up.

Doctors generally call a score high once it climbs above 100 Agatston units, or above the 75th percentile for your age and sex ^[5].

This is what flips decisions: statins pay off most where the calcium already is. MESA put a number on it. The 10-year "number-needed-to-treat" (how many people you have to put on the drug to prevent one heart event) was about 87 when CAC was 0, but only about 19 when CAC was above 100. Same pill, wildly different payoff ^[4].

The twist: plenty of people flagged for a statin had no calcium at all ^[3].

• Among people the guidelines actually recommended a statin for, 41 percent had a CAC of 0, with just 5.2 events per 1,000 person-years ^[3].
• Among everyone a statin was recommended or considered for, 44 percent had a CAC of 0, with 4.2 events per 1,000 person-years ^[3].

For a lot of them, the scan said "you are low risk after all."

The guidelines have caught up, and they keep leaning harder on CAC. The 2018 American cholesterol guideline (ACC/AHA) rated CAC Class IIa, meaning reasonable to use. It openly backed holding off on a statin when CAC is 0. That applied to adults aged 40 to 75 with an LDL cholesterol between 70 and 189 mg/dL. The exceptions: people with diabetes, current smokers, and anyone with a family history of early heart disease ^[5].

In 2026 the US societies replaced that guideline. The new ACC/AHA dyslipidemia guideline switches to the newer PREVENT risk calculator. For some groups defined by their calcium score, CAC is now a Class I tool, the strongest grade. The guideline reads your score against your age and sex and uses it to set how hard to treat and how low your LDL should go ^[20]. A 2026 MESA analysis adds a warning. In people who already cleared the treatment line on risk alone, heart events stayed higher even at CAC 0, so a zero should not generally cancel their statin ^[21].

Europe's ESC 2021 guideline rates CAC a notch lower, Class IIb, useful for nudging risk up or down near the treatment line ^[6]. And Germany's own Heinz Nixdorf Recall study found that CAC sharpens risk assessment beyond what both the European and American guidelines pick up on their own ^[12].

A CAC scan is a tie-breaker, not a routine check-up. The sweet spot is narrow. You have no symptoms, you are roughly 40 to 70, and your heart risk is borderline or intermediate. And you are stuck on a tough call about whether to start a statin. That is where the scan earns its keep.

What counts as intermediate risk? It depends on the calculator. The 2026 US guideline switched to the PREVENT calculator ^[20]. On that scale, a 2026 MESA analysis counted a 10-year risk of 3 to under 5 percent as borderline and 5 to under 10 percent as intermediate ^[21]. The older calculator behind the 2018 guideline put the borderline and intermediate band higher, around 5 to 20 percent ^[5]. If you are in that zone and you and your doctor honestly cannot decide whether to start a statin, a CAC scan can break the tie. The 2018 guideline treated this exact moment as Class IIa shared decision-making, meaning a reasonable thing to do together ^[5]. The 2026 MESA analysis found CAC helped most right here, when the statin call was uncertain ^[21].

This is not a test for everyone. A good doctor will talk some people out of it:

• Already high-risk? You should be on treatment no matter what the score says, so the scan rarely changes the plan.
• Truly low-risk and young? Your odds of a problem are already so small that a scan mostly buys you a bill and a little radiation, with no decision riding on it.
• Having symptoms like chest pain or breathlessness? A CAC score is the wrong tool. You need a CT angiography or a stress test, which hunt for actual blockages and starved blood flow, not just calcium.

There are also a few exceptions that override even a perfect zero. People with diabetes, current smokers, and anyone with a strong family history of early heart disease were specifically left out of the 2018 "hold the statin" rule, even at CAC 0 ^[5]. Why? Because those drivers keep churning out events whether or not calcium has shown up yet. A diabetic smoker with a zero score is not off the hook.

So the rule of thumb is simple. If your treatment decision is already made, skip it. If you are stuck in the middle and want data that mirrors your actual arteries, a CAC scan is one of the most useful 20-minute investments you can make.

A CAC scan is a quick cardiac CT without contrast dye. It takes minutes and delivers a low radiation dose: around 1 mSv on modern scanners that only fire during one slice of your heartbeat (prospectively-gated) ^[13]. That is roughly the natural background radiation you absorb over a few months.

The scan itself is almost boring, which is exactly what you want. You lie down, a few electrodes go on your chest, and the scanner snaps images timed to your heartbeat. No iodine dye is needed for the scoring, so no IV and no allergy worries. A computer turns the images into your Agatston score.

The question everyone actually asks is about the radiation, so here are the numbers. On those modern protocols the dose runs around 1 mSv. Older machines ran higher. In the MESA cohort, one analysis clocked an average of about 1.05 mSv ^[13]. An older modeling study that compared many scan protocols found a range of 0.8 to 10.5 mSv, with a median around 2.3 mSv ^[7]. That big spread came from old machines and settings. Newer scanners sit at the low end.

To put it in perspective: that low end is roughly the natural background radiation you soak up over a few months just by living on Earth. Low, but not nothing.

That leads to the practical takeaway on repeat scans. A zero comes with a 3 to 7 year warranty ^[1], which means there is no reason to rescan every year. Yearly imaging just piles on radiation for no new information, because plaque barely budges on that timescale once you start at zero. Space your scans out.

One last thing worth knowing. Agatston scores are reliable, but at very low values they wobble a bit from scan to scan. So do not lose any sleep over a literal "1 versus 0" difference. What matters clinically is the gap between a clean zero and a clearly positive score, not the gap between a 0 and a barely-there 1. Treat the very low end as a fuzzy zone, not a hard line.

This section covers Germany, Austria and Switzerland. Prices and coverage elsewhere, including the US, work differently.

In Germany, if you have no symptoms, you pay for a calcium score scan out of your own pocket. Practices offer it as an IGeL service (Individuelle Gesundheitsleistung, basically a self-pay extra), not something the public insurance (GKV) covers. Budget roughly 150 to 400 EUR. Some clinics quote as little as 150 EUR, others charge up to about 500 EUR, depending on the practice and the city. The IGeL-Monitor puts the CT alone, without consultation or report, at 135 to 380 EUR depending on the fee rate ^[17].

Why is it not covered? It comes down to one word: symptoms. In January 2024 the G-BA (the body that decides what German statutory insurance pays for) made CT coronary angiography, a heart CT with contrast dye, a covered service for people with suspected chronic coronary heart disease ^[18]. That decision covers people with symptoms. It explicitly does not cover heart CTs as an early-detection check for people who feel fine ^[17]. So if you just want a CAC scan to check your risk while feeling fine, you are outside what insurance will pay back.

There is also a legal catch. Under German radiation protection law, a CT used purely as an early-detection check in people without symptoms needs its own federal regulation. So far only low-dose lung-cancer screening for smokers has one, in force since July 2024. None exists for the heart. The IGeL-Monitor, run by the Medizinischer Dienst, therefore says CT early detection of narrowed heart arteries is not legally permitted, even though practices still offer it ^[17]. Before you book, ask your doctor why the scan is medically justified for you in particular.

Austria and Switzerland work the same way. Without a clear symptom-driven reason, a calcium scan is usually a private, pay-it-yourself workup. In Austria that often means going to a Wahlarzt (a doctor who works outside the public contract system) and paying directly. In Switzerland you are typically covering it yourself in CHF unless there is a covered medical reason.

So what should steer the decision in the DACH region, when you are the one paying? For doctors here, the European reference point is the ESC 2021 prevention guideline, which rates CAC a Class IIb risk modifier near the treatment threshold ^[6]. Translation: it is a sensible add-on for the intermediate-risk call, not a blanket recommendation for everybody.

The bottom line for your wallet: this is a few hundred euros or francs of your own money, best spent when you are honestly on the fence about a statin. If your decision is already clear either way, the money buys you reassurance at best and an unnecessary scan at worst. Treat it as a targeted spend, not another checkbox on an annual health menu.

A CAC scan has one limit you really need to understand: it only sees hardened, calcified plaque. Soft plaque, the younger and more rupture-prone kind that has not turned to chalk yet, is invisible to it. So a zero score can sit right next to real disease.

How often does that happen? In the SCOT-HEART trial, 642 patients with chest symptoms scored a CAC of 0. Still, 16 percent of them had soft, non-calcified plaque on a CT angiogram (a contrast-dye CT that also shows soft plaque) ^[8]. And that soft plaque is not harmless. In the same trial of people with stable chest pain, when soft, fatty plaque filled more than 4 percent of the artery's volume, the risk of a later heart attack ran about 4.65 times higher. That soft plaque predicted heart attacks on its own, over and above the calcium score and the narrowing ^[9]. In the ICONIC study of people with chest symptoms, 23 percent of those who later had an acute coronary syndrome (a heart attack or unstable angina) had a CAC of 0. The plaque behind their event looked much like the plaque in people with calcium, though that group was small, just 25 people ^[16].

One nuance that matters: this blind spot is biggest in people who already have symptoms or who are younger. In a Danish registry of people with chest symptoms, 58 percent of those under 40 with a real narrowing still scored a CAC of 0, against 5 percent of those aged 70 or older ^[19]. A zero is at its most reassuring exactly where the scan is built to be used, in symptom-free, middle-aged screening. It gets shakier in people with symptoms and in younger people.

Two blood tests pair well with CAC here. CAC is the plaque you have already built. ApoB and Lp(a) are the lifelong drivers that build it. ApoB counts every artery-clogging cholesterol particle in your blood. Lp(a) (say it "L-P-little-a," a cholesterol particle you mostly inherit) is set close to birth, so you only need to measure it once in your life. The 2026 US dyslipidemia guideline now recommends measuring it at least once in adult life ^[20]. A 2022 analysis showed Lp(a) and CAC each predict heart-disease risk on their own, in both MESA and the Dallas Heart Study ^[11]. The same study had a reassuring twist. Over 13 years, people with high Lp(a) but a CAC of 0 did about as well as people with neither. The real danger showed up when high Lp(a) met a CAC of 100 or more ^[11]. So a zero buys you time. High Lp(a) still works for life, though, so keep rescanning on schedule.

One limit sits in the evidence itself. All these CAC numbers come from big long-running studies (MESA) and CT-imaging trials (SCOT-HEART), not from a trial that randomly assigned people to CAC-guided care versus usual care and counted heart attacks. So the evidence is about predicting and reclassifying risk. Strong, but a different flavor of strong. And a zero never erases smoking, diabetes, or high blood pressure as ongoing drivers.