Coronary Artery Calcium Scoring: What the Evidence Actually Supports

What is established

Coronary artery calcium scoring is a short, low-dose, non-contrast CT scan of the heart. It counts and quantifies calcified plaque in the coronary arteries and produces a single number, the Agatston score, that has been used in research and clinical practice for over two decades. As a procedure it is well characterized: the scan itself is quick, does not require contrast dye, and carries a radiation dose well below a standard diagnostic CT.

What is also established is that CAC is not a symptom test. It says nothing about whether a plaque is about to rupture today. It is a marker of cumulative atherosclerotic burden, used to refine a person's long-term cardiovascular risk estimate alongside cholesterol, blood pressure, smoking history and family history, not to replace them.[1]

What the evidence for this specific use actually shows

The clearest evidence comes from the Multi-Ethnic Study of Atherosclerosis (MESA), a large prospective cohort that has followed thousands of adults for over a decade. Analysis of nearly 6,800 participants found that CAC substantially improves discrimination when added to conventional risk factors, and that the same absolute CAC score carries a different risk meaning depending on a patient's underlying risk-factor profile. In other words, CAC does not replace the traditional risk calculation, it sharpens it.[2]

A separate MESA analysis testing 13 candidate risk markers against one another found that a CAC score of zero was the single strongest negative risk marker of the group, over a 10-year follow-up period. For a patient whose risk-factor profile suggests they might benefit from a statin but the decision is genuinely uncertain, a CAC of zero is one of the more informative pieces of reassurance available in preventive cardiology.[3]

This body of evidence is the basis for the 2018 AHA/ACC cholesterol management guideline, which formally endorses measuring CAC specifically to help decide on statin therapy in adults at borderline or intermediate risk when that decision is unclear after standard risk assessment.[1]

The trap: improved prediction is not the same as improved outcomes

Every source above is a risk-prediction or risk-reclassification study, or a guideline built on that kind of evidence. None of them is a randomized trial that took patients, split them into a CAC-guided decision arm and a standard-care arm, and measured whether fewer people in the CAC arm actually had heart attacks or died. Improving a statistical model's ability to sort patients into risk categories is a genuinely useful step, but it is a different claim from proving that acting on the score changes what happens to the patient.[2][1]

What remains uncertain: the DANCAVAS null finding

The most direct outcomes test of population-level cardiovascular screening that included CAC comes from the Danish Cardiovascular Screening trial (DANCAVAS), a randomized trial of over 47,000 men aged 65 to 74. After 5.6 years, inviting men to a comprehensive screening program, which combined CAC scanning with screening for abdominal aortic aneurysm, peripheral artery disease and atrial fibrillation, did not significantly reduce all-cause mortality (hazard ratio 0.95, p=0.06). A modest reduction in stroke was seen, but the primary mortality result was a null finding.[4]

DANCAVAS bundled several screening modalities together, so it cannot isolate what CAC specifically contributed, and it studied only older Danish men, which limits how far the result generalizes to a broader adult population making earlier statin decisions. But it is the closest thing available to a hard-outcomes test of calcium-inclusive screening at a population level, and its honest reading is that inviting an entire population to comprehensive screening did not clearly extend life within the trial's follow-up window.[4]

What this means for you

CAC scoring is treated as what the evidence supports: a genuinely useful, physician-interpreted input for a specific, narrow decision, not a general wellness scan and not a promise of outcome. It is most useful when your cardiovascular risk profile puts the statin decision in a genuinely grey zone, where a result of zero or a meaningfully elevated score can change what your physician recommends next.[1]

Diagnostics come first, and a physician decides what the number means for that individual patient, in the context of their full risk profile, not against a fixed protocol or a one-size score threshold. The DANCAVAS result is part of that conversation too: a scan can inform a decision well without it following that scanning everyone, or scanning routinely without a clear clinical question, has been shown to change how long someone lives.[4]