A calcium score adds little across the board, but sorts borderline-risk patients sharply

Design

Prospective multi-ethnic cohort, risk-model performance analysis

Population

6,098 US adults aged 45 to 79 without cardiovascular disease (MESA)

Primary outcome

Ten-year fatal and non-fatal atherosclerotic cardiovascular disease

Effect

C statistic gain 0.02 (95% CI 0.01 to 0.03); borderline risk event rate 1.9% with CAC 0 vs 14.3% with CAC ≥300

The 2026 ACC/AHA multisociety dyslipidaemia guideline recommends selective coronary artery calcium (CAC) scoring for adults whose ten-year risk on the PREVENT equations is 3% to under 10%. This analysis of the Multi-Ethnic Study of Atherosclerosis tested what CAC actually adds to PREVENT in 6,098 adults aged 45 to 79 without cardiovascular disease, followed for ten years; 366 (6%) had an event.

PREVENT alone discriminated reasonably (C statistic 0.73). Adding CAC raised it by 0.02 (95% CI 0.01 to 0.03), with a net reclassification improvement of 0.095. Overall, that is a modest gain. The value sat in the borderline group (3% to under 5%): ten-year event rates were 1.9% with a score of zero, 3.9% with 1 to 99, 7.4% with 100 to 299 and 14.3% with 300 or more.

In practice, this supports scanning where a decision genuinely hangs on it, and not as a routine add-on. A borderline patient with a high score has a risk comparable to someone in the high-risk band; one with a zero score sits close to the low-risk group.

• Reserve CAC scoring for adults with PREVENT ten-year risk of about 3% to under 10% where statin choice is uncertain.
• Treat a borderline-risk patient with a CAC of 300 or more as high risk when discussing statins.
• A CAC of zero in a borderline-risk patient supports deferring a statin, with reassessment in a few years.
• Do not order CAC in patients already clearly above or below treatment thresholds; it rarely changes the decision there.

Why it matters

It gives the new guideline's calcium recommendation an evidence base built on the equations it now uses.

Don't overread it

MESA is a US cohort; the absolute rates may not transfer directly to South Asian patients, whose risk is underestimated by most equations.

The statistics, in plain English

The C statistic measures how often a model ranks the person who later has an event above one who does not; 0.73 to 0.75 is a small step. Net reclassification improvement of 0.095 means about one in ten people were moved into a more accurate risk category. The large spread in event rates within the borderline group (1.9% to 14.3%) is what makes CAC useful there.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.