The Coronary Calcium Score: A Look Inside the Arteries to Refine Risk

The coronary calcium score is one of the few cardiovascular risk tools that looks directly inside the arteries to measure established disease rather than estimating risk from external factors applied to population averages. For the right patient at a genuine decision point, it can resolve an otherwise uncertain treatment question efficiently and without contrast or significant radiation. It also carries a specific, well-documented limitation for women that changes how the result must be interpreted, and ignoring that limitation is how a useful test becomes a misleading one.

The Mechanism

Coronary artery calcium is deposited within atherosclerotic plaques as they mature. The calcification process is biologically analogous to bone mineralization: calcium and phosphate are deposited in the fibrous cap and necrotic core of developing plaques through mechanisms involving osteoblast-like vascular smooth muscle cells and regulatory proteins including osteopontin, osteocalcin, and matrix Gla protein (MGP). The presence of coronary calcium therefore indicates established, advanced atherosclerosis, not merely the risk factors that lead to it. This is the fundamental distinction between CAC scoring and other risk estimation approaches: it measures what is already there, not what the probability distribution of a population with similar risk factors would predict.

The scan itself is a non-contrast, ungated or prospectively gated low-dose CT examination of the chest, typically delivering a radiation dose of 1 to 3 millisieverts, comparable to a standard mammogram. The output is the Agatston score, which weights each calcium deposit by its density (HU-based multiplier of 1 to 4) and area, summing across all four coronary arteries to produce a single unitless number. A score of zero means no detectable calcified plaque; scores of 1 to 99 indicate mild disease; 100 to 399 indicate intermediate-to-high disease; and scores of 400 or above indicate extensive atherosclerosis with high near-term event risk.

The important mechanical nuance is that calcium is not itself the cause of most acute coronary events. Plaque rupture, which precipitates most myocardial infarctions, typically occurs at sites of soft, lipid-rich, non-calcified plaque, not at highly calcified lesions. Calcified plaque is actually more mechanically stable than soft plaque. The calcium score predicts events because calcium indicates the overall atherosclerotic plaque burden, including the softer non-calcified plaques elsewhere in the same arteries, not because the calcium itself is dangerous. This distinction matters clinically when a patient asks why a high calcium score can coexist with a “normal” coronary artery scan that shows no obstructive stenosis: the calcium reflects total disease burden, not flow obstruction specifically.

A second mechanistic point relevant to women: calcium develops in plaques progressively as they mature. Newer or smaller plaques, including those that are clinically significant in terms of rupture risk, may carry very little calcified component, particularly in younger patients. This means that a low or zero calcium score in a patient with early or predominantly non-calcified plaque does not mean the arteries are clear of disease; it means the disease that is there has not yet calcified substantially. In women, who on average develop coronary calcium approximately a decade later than men with equivalent risk factor burden, this scenario is not rare.

What the Evidence Shows

The foundational dataset for CAC scoring is the Multi-Ethnic Study of Atherosclerosis (MESA), a prospective cohort of 6,814 men and women free of cardiovascular disease at enrollment, recruited from six US communities and followed for cardiovascular events from 2000 onward. MESA enrolled participants across four self-reported ethnic groups (White, Black, Hispanic, and Chinese American) and performed CT-based CAC scoring at baseline, enabling risk prediction analyses that could directly test how calcium score added to or refined traditional risk factor-based models.

The 2004 Greenland et al. analysis from MESA, published in Circulation, established that adding the CAC score to Framingham Risk Score (FRS) substantially improved the C-statistic for 10-year cardiovascular event prediction from approximately 0.72 to 0.81, a gain that is considered clinically meaningful for a single imaging biomarker. Subsequent analyses confirmed that this improvement held across all MESA ethnic subgroups, though the calibration and the score-specific event rates differed by group.

The MESA data on zero CAC scores are particularly important for clinical application. Persons with a CAC score of zero had a 10-year coronary heart disease event rate of approximately 1 percent, even among those classified as intermediate risk by FRS. This very low event rate in zero-CAC individuals supports deferring intensive pharmacotherapy (particularly statins) in those patients and managing with lifestyle measures, pending rescanning in three to five years. The 2022 MESA publication by Blaha et al. confirmed that the zero-CAC group maintained very low event rates even at 10- to 15-year follow-up in the absence of intervening events or new risk factor development.

The translation of these findings into clinical guidance came with the 2018 ACC/AHA cholesterol guidelines, which explicitly incorporated CAC scoring into the decision pathway for patients at borderline (5 to 7.5 percent 10-year ASCVD risk) and intermediate (7.5 to 20 percent) risk where the statin treatment decision is uncertain after a risk discussion. A CAC score of zero in these patients supports deferring statin therapy; a score of 100 or above supports initiating treatment; and a score of 1 to 99, particularly in patients aged 55 or older or with other risk-enhancing features, may support treatment initiation depending on the clinical context.

The women-specific evidence is where important caveats emerge. MESA data showed a consistent sex difference in calcium score distributions at any given age: at age 55, for example, the 75th percentile CAC score for men was approximately 100 Agatston units, while for women it was approximately 20. This calcium lag means that age-based CAC percentile reference ranges (which compare a patient’s score to others of the same sex and age) are essential for correct interpretation: a score of 50 in a 55-year-old woman is above the 75th percentile for women her age and should be interpreted as elevated, not moderate, even though the same absolute score in a 55-year-old man falls near the 50th percentile.

The non-obstructive disease limitation in women is documented most clearly in the WISE study (Women’s Ischemia Syndrome Evaluation), a National Heart, Lung, and Blood Institute-funded multicenter prospective cohort that enrolled 936 women referred for coronary angiography for suspected ischemia. WISE found that approximately 60 percent of symptomatic women referred for angiography had no obstructive coronary artery disease (defined as stenosis of 50 percent or more) on the invasive test. Yet these women had significantly elevated rates of major adverse cardiac events over five-year follow-up compared to low-risk women without symptoms, driven by microvascular dysfunction, coronary vasospasm, and other non-obstructive pathologies. A CAC score in most of these women would have been low or zero, because non-obstructive disease and microvascular dysfunction do not produce the advanced calcified plaque the scan detects. Using the CAC score as a reason not to further evaluate symptomatic women in this pattern would cause systematic underdiagnosis.

A 2015 analysis by Shaw et al. in JACC: Cardiovascular Imaging analyzed CAC scoring in women from the CONFIRM registry and found that the relationship between CAC score and cardiac event rates was modified by sex: women at any given Agatston score had higher event rates than men at the same score, consistent with the interpretation that a given calcium burden in women reflects greater underlying plaque vulnerability or total plaque burden than the same score in men. This finding, while requiring replication in additional datasets, suggests that thresholds developed in mixed-sex or male-dominant cohorts may need sex-specific calibration.

Coronary CT Angiography: When Anatomy Matters More Than Calcium

For women where the clinical question requires anatomical characterization beyond what a CAC scan provides, specifically those with moderate risk and stable symptoms where non-calcified plaque is a meaningful possibility, coronary CT angiography (CCTA) offers a complementary diagnostic path that addresses a different and sometimes more relevant question.

CCTA generates cross-sectional images of the coronary arteries using iodinated contrast and allows direct visualization of both the lumen and the vessel wall. Unlike the CAC scan, which detects only calcified atherosclerotic lesions, CCTA can identify non-calcified, low-attenuation plaque, the lipid-rich material that carries higher vulnerability to rupture but does not register on the non-contrast calcium scoring scan. Women are known to have a higher proportion of non-calcified versus calcified plaque than men at similar risk levels, meaning the CAC scan systematically underestimates total plaque burden to a greater relative degree in women.

The SCOT-HEART trial (Scottish Computed Tomography of the Heart), published in the New England Journal of Medicine in 2018 by Newby and colleagues, randomized 4,146 patients with stable chest pain to CT angiography plus standard care versus standard care alone. At five years, the CCTA group had a significantly lower rate of fatal and non-fatal myocardial infarction (HR 0.59; 95% CI 0.41 to 0.84). The benefit was attributed to improved identification of coronary artery disease, with more appropriate statin therapy, antiplatelet treatment, and revascularization in patients who needed it, and more confident ruling-out of coronary disease in those who did not. 5 / Solid

The practical distinction from a CAC scan is the question being answered. The CAC score answers: “Is calcified plaque detectable in the coronary arteries?” CCTA answers: “What is the total atherosclerotic burden, calcified and non-calcified, and is there obstructive anatomy?” For the intermediate-risk asymptomatic woman whose management decision depends on plaque burden, CCTA provides a more complete answer when mixed or non-calcified plaque is the likely phenotype. For the symptomatic woman where the exercise ECG is uninformative, CCTA can confirm or exclude obstructive anatomical disease in a way that directly informs downstream management.

Limitations include contrast requirement, radiation exposure (low with modern scanners), and inability to directly quantify ischemia, CCTA shows anatomy, not function. The choice between CCTA and functional stress testing depends on the specific question being asked.

What to Do This Week

1. If your clinician has told you that your 10-year ASCVD risk is in the borderline or intermediate range (roughly 5 to 20 percent) and you are uncertain about whether to start a statin, ask specifically whether a CAC scan would change the decision. If the plan would be the same regardless of the result, the scan is not indicated. If the result would genuinely change the recommendation, it is the right test for that question.

2. If you receive a CAC score of zero, understand the specific meaning: no detectable calcified plaque in the coronary arteries and low near-term risk of obstructive coronary events. This supports deferring statin therapy in appropriate intermediate-risk asymptomatic patients with rescanning in three to five years. It does not mean your arteries are disease-free and it does not evaluate microvascular disease, vasospasm, or non-calcified plaque.

3. If you have symptoms, including chest pain, exertional dyspnea, or reduced exercise tolerance, do not accept a CAC score as an evaluation or dismissal of those symptoms. The test was not designed for symptom evaluation and a low result cannot exclude the non-obstructive coronary disease and microvascular dysfunction that is common in symptomatic women. Pursue the appropriate symptom evaluation regardless of your CAC score.

4. If you are younger than 50 and have a low or zero CAC score, recognize that the limitations of the test are most pronounced in this age group. Women in their 40s have very low calcification rates even in the presence of non-trivial atherosclerosis, and the calcium lag relative to men is largest in the decade before menopause. A zero score in a young symptomatic woman is particularly poor evidence of absence.

5. Ask your clinician to interpret your score using sex-specific and age-specific percentile reference ranges, not just the absolute Agatston number. A score of 50 means different things at different ages and in different sexes. The calcium score has sex-specific reference tables derived from MESA that allow the result to be placed in the appropriate context.

The coronary calcium score, used for the right question in an asymptomatic woman at borderline risk, is a direct and efficient look at calcified plaque burden that genuinely refines prevention decisions. Used as a blanket reassurance for symptoms, as a substitute for symptom evaluation, or without sex-specific interpretation, it can mislead in specific and consequential ways. Matching the test to the question it was built to answer is what makes it a valuable clinical tool rather than a source of false confidence.

The Women’s Signal Check is fifteen questions mapping the female cardiovascular risk pattern, including reproductive history, microvascular signals, and the factors standard risk calculators do not capture. It produces a specific starting point for your next clinical conversation.