CIMT

Carotid intima-media thickness (CIMT) measures the combined width of the inner two layers of the carotid artery wall using ultrasound. No radiation, no contrast, relatively cheap. For decades it was the go-to surrogate endpoint in cardiovascular trials, and clinicians read a thinner CIMT after treatment as evidence of benefit. The evidence has since drawn a sharp line through that assumption.

What baseline CIMT tells you

A single snapshot of CIMT does carry prognostic weight. A meta-analysis of general-population studies found that each one-standard-deviation increase in common carotid IMT was associated with a relative risk of 1.26 for myocardial infarction and 1.32 for stroke [1]. Baseline CIMT, in other words, is a legitimate marker of subclinical atherosclerosis burden. The higher your number, the higher your background risk.

Where the surrogate breaks down

The problem arises when CIMT is used to track change, whether spontaneous or drug-induced.

The PROG-IMT collaboration pooled individual patient data from 16 cohorts (n=36,984) and found that CIMT progression over time did not predict myocardial infarction, stroke, or vascular death (HR 0.97), even though baseline CIMT in the same dataset still did [2]. Progression and prognosis are measuring different things, and one does not stand in for the other.

The picture is equally clear for treatment. A pooled analysis of 41 randomized controlled trials (n=18,307) found no relationship between treatment-induced CIMT regression and actual event reduction [3]. A drug can shrink CIMT and fail to prevent heart attacks. A drug can leave CIMT unchanged and still save lives. CIMT change is not a valid surrogate for clinical outcomes.

Why this matters for supplement and diet trials

Several plaque trials testing berberine, lycopene, Mediterranean diet patterns, and exercise have used CIMT change as their primary endpoint [4, 5, 6, 7, 8]. Given what PROG-IMT and the Costanzo analysis established, those CIMT reductions are weak evidence at best. They show the wall changed; they do not show patients did better. This is why LampHill rates the plaque claims from those interventions as "preliminary" rather than "supported."

A separate finding that does not rescue CIMT

One statistic circulates in discussions of plaque imaging: roughly each 1% reduction in coronary percent atheroma volume on intravascular ultrasound (IVUS) is associated with approximately 19% lower major adverse cardiovascular events [9]. This is sometimes invoked to argue that imaging-based plaque measures track outcomes. It does not apply here. Coronary IVUS atheroma volume and carotid IMT change are different measures in different vessels using different technology. The IVUS finding does not rehabilitate CIMT progression as a surrogate.

Practical context

CIMT has real limitations beyond the surrogate problem: it samples one vessel that may not reflect coronary disease, results vary with operator skill and equipment, and the measurement cannot distinguish soft from calcified plaque. For assessing subclinical atherosclerosis directly, coronary CT angiography with evaluation of low-attenuation (soft) plaque, or coronary artery calcium scoring, provides more actionable information.

If you've had a CIMT measurement, the baseline number is informative. If a trial or a clinician points to CIMT regression as proof that an intervention worked, that claim now requires a much stronger argument than the imaging alone can provide.

Where the evidence is contested

The disagreement is not about whether baseline CIMT predicts risk, both sides accept that. The dispute is whether the null result for CIMT change should permanently retire the measure from trials, or whether methodological issues (short follow-up, heterogeneous protocols, regression to the mean) explain the disconnect.

Some researchers argue that CIMT progression may still be biologically meaningful in high-risk subgroups or over longer timeframes than most trials used, and that pooling across heterogeneous populations dilutes a real signal. The PROG-IMT authors note their finding should prompt caution, not necessarily abandonment [2].

Others, including analysts who reviewed the 41-trial pool, read the data more definitively: if drug-induced CIMT regression across thousands of patients and dozens of trials cannot predict event reduction, the burden of proof for any individual trial reporting CIMT benefit is now high [3].

For practical purposes, the conservative reading, that CIMT change is insufficient on its own to conclude an intervention works, is where the weight of published evidence currently sits.

References

1. Lorenz 2007, Circulation — baseline CIMT predicts MI/stroke · PubMed 17242284
2. Lorenz 2012 PROG-IMT, Lancet — CIMT progression does NOT predict events (refuted surrogate) · PubMed 22541275
3. Costanzo 2010, JACC — drug-induced CIMT regression doesn't track outcomes (refuted surrogate) · PubMed 21126642
4. raw/physionic/physionic-berberine-shrinks-plaque-study.txt
5. raw/physionic/physionic-nattokinase*.txt
6. raw/physionic/physionic-mediterranean-diet-reversing-atherosclerosis.txt
7. Physionic — Why plaque in arteries not veins (GAG composition)
8. raw/physionic/physionic-exercise-hiit-plaque*.txt
9. Iatan 2023 — 23-study meta-analysis: 1% PAV → 19% MACE · PubMed 37647074