Clinical Utility of Carotid Intima–Media Thickness Measurement in Primary Care: A Narrative Review

Overview of major prospective cohorts

The clinical evidence base for CIMT rests on multiple large prospective cohorts rather than on any single study. Table 1 summarises the key cohorts and meta-analyses that have shaped the field, with study-level reference numbers shown next to each study name.

Table 1.

Major prospective cohorts and meta-analyses of carotid intima–media thickness (CIMT)

Histology of the intima–media complex

Elastic arteries—including the aorta and the carotid—comprise three layers. The tunica intima consists of a single endothelial layer with subendothelial connective tissue. The tunica media contains multiple layers of vascular smooth muscle cells and elastic laminae. The tunica adventitia is the outermost fibroelastic layer. The B-mode double-line pattern arises from two acoustic-impedance interfaces (Fig. 1): the lumen–intima interface and the media–adventitia interface.2,4

Fig. 1.

Carotid anatomy and the B-mode double-line pattern. (A) Schematic of the carotid system. The common carotid artery (CCA) bifurcates at the carotid bulb into the internal (ICA) and external (ECA) carotid arteries. The standard measurement segment is the far wall of the distal 1 cm of the CCA, immediately proximal to the bifurcation (blue dashed box). (B) On B-mode ultrasound, the far wall produces two parallel hyperechoic lines—the lumen–intima and media–adventitia interfaces—separated by a hypoechoic band corresponding to the media. The distance between the two hyperechoic lines defines the intima–media thickness (IMT). At the near wall the beam is obliquely incident, the double-line pattern is degraded, and measurement is unreliable. Original figure prepared by the authors.

The dual mechanism of CIMT thickening

CIMT increases through at least two distinct mechanisms.16,17 The first is non-atherosclerotic adaptive thickening, which reflects medial smooth-muscle hypertrophy and extracellular-matrix deposition in response to ageing, chronic hypertension, and altered shear stress. The wall thickens in a diffuse, concentric pattern. The second is atherosclerotic thickening, driven by endothelial dysfunction, subendothelial deposition of oxidised low density lipoprotein cholesterol (LDL-C), monocyte–macrophage infiltration, foam-cell formation, and a synthetic phenotypic switch of smooth-muscle cells. The resulting lesion is focal and eccentric.16 The reproducibility and clinical relevance of these two mechanisms differ markedly between carotid segments. Table 2 summarises segment-specific reproducibility and clinical relevance: the distal CCA, where adaptive thickening predominates and operator-dependent variability is lowest (inter-reader coefficient of variation [CV] 3–5%), is the standard measurement site; the carotid bulb and proximal internal carotid artery (ICA) are more plaque-prone but show poorer reproducibility (CV 8–15%).18

Table 2.

Reproducibility and clinical relevance of carotid measurement segments

This distinction has direct clinical implications. According to Finn et al.16, CIMT increase in the CCA is largely adaptive, whereas thickening at the bulb and the proximal ICA is more often atherosclerotic. This site-specific heterogeneity must inform both the choice of measurement site and the interpretation of findings. It also explains why CIMT and plaque convey intrinsically different clinical information.

CIMT, plaque, and coronary atherosclerosis

Autopsy correlations between CIMT and coronary atherosclerosis exist but are moderate (r=0.3–0.5).19 Carotid plaque is a stronger predictor of coronary events than CIMT alone. In the ARIC analysis, addition of plaque significantly improved the C-statistic over CIMT alone.20 The American Society of Echocardiography (ASE) therefore recommends concurrent plaque assessment whenever CIMT is measured.3

Why standardisation matters

CIMT is measured at the 0.01-mm scale. Operator, equipment, anatomical site, cardiac phase, and post-processing all introduce variability. In Polak et al.’s multicentre study, the inter-reader CV was 3–5% with a standardised protocol and 10–15% without one.21 The 2004 Mannheim consensus (revised 2011) and the 2008 ASE consensus established the current standards.3,22 The present review follows their integrated recommendations.

Equipment and acquisition

A 7–12 MHz linear-array transducer is used. Time-gain compensation, dynamic range, depth, and focal zone are adjusted to optimise far-wall resolution.3 The patient is supine with the head turned approximately 45° away from the side being examined. When electrocardiographic gating is available, images are acquired at end-diastole (R-wave peak).22 Longitudinal views are standard; transverse views serve to confirm plaque.

Anatomical site

The Mannheim consensus recommends the far wall of the distal 1 cm of the CCA, proximal to the bifurcation.22 The far wall is preferred for two reasons. First, the ultrasound beam is perpendicular to the wall, which yields the sharpest double-line pattern (Fig. 1). Second, near-wall images suffer from off-axis incidence, causing signal loss at the lateral margins and systematic measurement error.3

Calculation and reporting

Automated edge-detection software reduces the CV by approximately 50% compared with manual measurement and is strongly preferred.3,21 Both mean and maximum CIMT should be reported for each 1-cm segment, and the average of the two sides should be used clinically. In Lorenz et al.’s meta-analysis, each 0.1-mm increase in mean CIMT corresponded to a 10–15% increase in myocardial infarction risk and a 13–18% increase in stroke risk.7

Definition of plaque

Plaque is an atherosclerotic lesion and is conceptually distinct from a mere increase in CIMT. The Mannheim consensus defines plaque as any of the following: focal CIMT ≥1.5 mm protruding into the lumen, focal thickening ≥50% greater than surrounding CIMT, or focal thickening ≥0.5 mm greater than adjacent IMT.22 When plaque is identified, the report must include location, number, surface morphology (smooth, irregular, or ulcerated), echogenicity (hypoechoic plaques are lipid-rich and unstable; hyperechoic plaques are fibrotic and more stable), calcification, and the degree of stenosis by NASCET or ECST criteria.3,23

Age- and sex-related variation

CIMT increases nearly linearly with age. The annual rate of increase is 0.005–0.010 mm/yr in healthy adults and 0.010–0.030 mm/yr in those with risk factors.24,25 In the pooled analysis by Engelen et al.26 (24 cohorts, 24,871 individuals), mean CIMT in men rose from 0.54 mm in the fourth decade to 0.83 mm in the eighth. At any given age, men exceed women by 0.03–0.05 mm on average.

Absolute thresholds versus percentile-based interpretation

A single absolute cut-off (e.g., 0.9 or 1.0 mm) is simple but ignores age, sex, and ethnicity. The ASE consensus therefore recommends a percentile-based framework. Values above the 75th percentile (age-, sex-, and ethnicity-matched) indicate increased cardiovascular risk; values between the 25th and 75th percentile indicate average risk; and values below the 25th percentile suggest a “low vascular age.”3

Reference values for Korean adults

Authoritative nationwide reference values for CIMT in Korean adults are not yet available. Subgroup analyses from the Korean Genome and Epidemiology Study (KoGES) and the Korean Social Life, Health, and Aging Project (KSHAP), together with multi-centre health-screening data, provide indicative age- and sex-specific distributions.26 Representative pooled summaries are shown in Table 3 for illustrative use only. Because absolute values vary substantially with equipment, anatomical site, and the automated algorithm, each institution should ideally establish its own normative distribution.3,26

Table 3.

Representative (Illustrative) age- and sex-specific mean CIMT distribution in Korean adults

An integrated framework for clinical decisions

Quantitative CIMT, percentile, and plaque status can be integrated into five practical categories. The framework presented below is proposed by the present authors as a pragmatic synthesis of the ASE,3 Mannheim,22 2019 ESC/EAS,27 and 2023 European Society for Vascular Surgery (ESVS)23 statements together with the 2018 Korean Society of Neurosonology consensus on extracranial carotid duplex ultrasonography.28 It has not been validated in dedicated prospective trials. Fig. 2 outlines the proposed clinical algorithm.

Fig. 2.

Author-proposed clinical decision algorithm for CIMT interpretation in primary care. After bilateral carotid ultrasound performed under the Mannheim and American Society of Echocardiography (ASE) standards, plaque is first assessed (Mannheim criteria). When plaque is absent, patients are stratified by age- and sex-adjusted percentile of mean CIMT into Categories I–III. When plaque is present, stenosis grade and clinical symptoms determine assignment to Category IV or V. Each category is linked to specific lifestyle and pharmacological actions and to recommended follow-up intervals. The five-tier framework and the follow-up intervals shown are proposed by the present authors as a pragmatic synthesis of the ASE, Mannheim, 2019 ESC/EAS, 2023 ESVS, and 2018 Korean Society of Neurosonology consensus statements; they are not endorsed by any single guideline and have not been prospectively validated. The original figure has been redrawn with a larger font size in this revision to improve legibility. RF, risk factor; US, ultrasound; CEA, carotid endarterectomy; CAS, carotid artery stenting; ABPM, ambulatory blood-pressure monitoring. Original figure prepared by the authors.

Category I (Normal): mean CIMT below the age- and sex-adjusted 75th percentile and no plaque.

Category II (Borderline thickening): 75th–90th percentile, or absolute value 0.8–0.9 mm.

Category III (Abnormal thickening without plaque): ≥ 90th percentile or absolute value ≥0.9 mm, without plaque.

Category IV (Plaque-positive, non-stenotic): Mannheim-defined plaque with stenosis <50%.

Category V (Significant stenosis): stenosis ≥50% or neurological symptoms.

Category I: normal findings

A normal CIMT indicates a low absolute risk of future cardiovascular events but does not exclude them. In ARIC, the 10-year coronary event rate in adults in their fifties with a normal CIMT was approximately 4.5%.20 A single risk factor—smoking, diabetes, or hyperlipidaemia—can raise this rate substantially. Normal CIMT should therefore be interpreted as “low current burden of subclinical atherosclerosis,” not as a clean bill of health. Routine follow- up ultrasound is not recommended; reassessment is reserved for clinical change or the emergence of new risk factors.12,13

Category II: borderline thickening

Borderline thickening is where reclassification has the highest clinical yield. Naqvi et al. reported that 35–40% of intermediate-risk individuals (10-year ASCVD risk 5–20%) are reclassified after carotid imaging, often with a corresponding change in pharmacotherapy.14 Suggested actions are: (i) re-calculate ASCVD risk; (ii) intensify lifestyle measures, including a Mediterranean diet, at least 150 minutes per week of moderate-intensity aerobic exercise, and smoking cessation; (iii) individualise LDL-C, HbA1c, and blood-pressure targets; and (iv) consider repeat ultrasound after an interval of approximately 3 years, in line with the test–retest variability of CIMT and with current guideline emphasis on risk-factor control rather than serial imaging.12,13,28,29

Category III: abnormal thickening without plaque

CIMT above the 90th percentile or ≥0.9 mm without plaque indicates subclinical atherosclerosis. The patient should be moved up by one risk tier. Recommended evaluation includes fasting glucose and HbA1c; a full lipid panel with LDL-C, HDL-C, triglycerides, non-HDL-C, and apolipoprotein B; renal function; and ambulatory or home blood-pressure monitoring. In smokers, pharmacotherapy for cessation should be actively considered.13,27 When repeated, ultrasound is best performed at intervals of approximately 2–3 years, since shorter intervals largely capture measurement noise rather than true biological change.22,28

Category IV: plaque-positive

A Mannheim-defined plaque denotes established atherosclerosis. The 2019 ESC/EAS guidelines automatically classify such patients as high risk or higher.27 In Inaba et al.’s meta-analysis, the five-year cardiovascular event rate among individuals with carotid plaque was approximately 2.5-fold higher (95% CI, 2.0–3.1) than in those without.30 Recommended actions are: (i) detailed plaque characterisation (echogenicity, surface, neovascularisation, stenosis); (ii) an LDL-C goal of <70 mg/dL (high risk) or <55 mg/dL (very high risk); (iii) individualised consideration of antiplatelet therapy; and (iv) follow-up ultrasound at intervals of 1–2 years, consistent with the ESVS 2023 recommendation for surveillance imaging in non-stenotic carotid plaque under optimal medical therapy.23,27,28

Category V: significant stenosis or symptoms

Stenosis ≥50%—or any suspected symptom such as transient ischaemic attack, amaurosis fugax, or ipsilateral neurological deficit—exceeds the scope of primary care. Prompt referral to neurology or vascular surgery is required.23 The 2023 ESVS guidelines recommend optimal medical therapy with annual imaging follow-up for asymptomatic 50–69% stenosis when revascularisation (carotid endarterectomy [CEA] or carotid artery stenting [CAS]) is not performed. CEA or CAS should be considered for asymptomatic stenosis ≥70%, and for symptomatic stenosis ≥50%.23

Non-pharmacological measures

Lifestyle modification is the foundation of treatment across all categories. In the PREDIMED randomised trial, a Mediterranean diet reduced major cardiovascular events by 30% over five years.29 A pooled analysis showed an additional reduction in annual CIMT progression of approximately 0.008 mm/yr with the same dietary pattern.31 In ARIC, CIMT progressed approximately 1.5 times faster in current smokers than in non-smokers, and the rate returned to non-smoker levels within five years of cessation.32 A 5–10% weight reduction in obese patients is associated with slower CIMT progression and improved endothelial function.33 The quantitative effects of major interventions are summarised in Fig. 3.

Fig. 3.

Effect of major interventions on annual carotid intima–media thickness (CIMT) change. Positive values indicate progression (worsening), negative values indicate slower progression or regression (improvement). Continued smoking is associated with the most rapid progression. High-intensity statin therapy and angiotensin-converting enzyme inhibitors or angiotensin-receptor blockers produce the largest reductions in progression. Sources: ARIC,32 PREDIMED-Navarra,31 Bedi meta-analysis,34 Wang meta-analysis,39 METEOR.35 Original figure prepared by the authors based on the cited literature.

Statin therapy

Statins slow CIMT progression by lowering LDL through upregulation of LDL receptors. Their pleiotropic effects—anti-inflammatory action, endothelial restoration, and plaque stabilisation—also contribute. Bedi et al.’s meta-analysis (16 RCTs, 6,177 participants) reported a mean annual CIMT change of –0.012 mm/yr (95% CI, –0.018 to –0.006) in statin-treated patients, compared with +0.009 mm/yr in controls.34 In the METEOR trial, rosuvastatin 40 mg halted maximum CIMT progression to –0.0014 mm/yr over two years in low-risk individuals.35 In patients with plaque, the 2019 ESC/EAS guidelines recommend an LDL-C goal of <70 mg/dL (high risk) or <55 mg/dL (very high risk). If statin monotherapy is insufficient, ezetimibe (additional 18–22% LDL-C reduction) and PCSK9 inhibitors (additional 50–60%) should be added in stepwise fashion.27

Antiplatelet therapy

Routine aspirin for asymptomatic CIMT thickening is not recommended. In the ASPREE trial, low-dose aspirin in healthy older adults did not reduce cardiovascular events but increased major bleeding by 38%.36 ARRIVE and ASCEND reached similar conclusions.37,38 Antiplatelet therapy may still be individually considered in plaque-positive patients with multiple risk factors and in those with stenosis ≥50%.23

Blood-pressure and glycaemic control

Angiotensin-converting enzyme inhibitors and angiotensin-receptor blockers consistently slow CIMT progression by suppressing angiotensin-II–mediated vascular remodelling. In Wang et al.’s meta-analysis, these agents reduced annual CIMT progression by 0.005 mm/yr compared with calcium channel blockers.39 According to the 2022 Korean Society of Hypertension guidelines, target blood pressure is <140/90 mmHg in the general adult population and <130/80 mmHg in patients with diabetes, chronic kidney disease, or high cardiovascular risk.40 In type 2 diabetes, GLP-1 receptor agonists (liraglutide, semaglutide) and SGLT2 inhibitors reduce cardiovascular events beyond their glycaemic effect. In LEADER and EMPA-REG OUTCOME, major adverse cardiovascular events were reduced by 13% and 14%, respectively.41,42

Measurement variability

The most important concept in serial CIMT interpretation is the recognition of measurement variability. Intra-session CV is approximately 2–3%; test–retest CV is 5–8%.21 For a mean CIMT of 0.7 mm, this corresponds to a measurement error of 0.014–0.056 mm. Small short-term changes (<1 year) most often reflect measurement noise rather than true biological change. The Mannheim consensus defines a clinically meaningful change as an annual increase of ≥0.02 mm or the appearance of new plaque.22

Recommended follow-up intervals

Recommended follow-up intervals by category are summarised in Table 4. These intervals are author-proposed and integrate the 2008 ASE consensus,3 the 2011 Mannheim consensus,22 the 2019 ESC/EAS guidelines,27 the 2023 ESVS guidelines,23 and the 2018 Korean Society of Neurosonology consensus on extracranial carotid duplex ultrasonography.28 It should be emphasised that none of these source documents prescribe explicit follow-up schedules for Categories I–IV; the only directly cited interval is the ESVS 2023 recommendation of annual imaging for asymptomatic 50–69% stenosis under medical management.23 The intervals in Table 4 and Fig. 2 are therefore pragmatic suggestions, set wide enough that any meaningful CIMT change will exceed the test–retest variability of 5–8%.21,22 They have not been validated in randomised trials, and individual clinical judgement should take precedence.

Table 4.

Author-proposed follow-up intervals and risk-factor reassessment by CIMT category

Standardisation for follow-up

Meaningful longitudinal comparison requires (i) the same equipment and probe frequency, (ii) the same anatomical site and angle, (iii) the same operator whenever possible, and (iv) the same post-processing algorithm.3,22 Direct comparison of results obtained on different machines or in different institutions risks misinterpretation. Clinicians should note these limitations explicitly in the report.

Limited risk-reclassification capacity

In the USE-IMT meta-analysis, addition of mean CIMT to the Framingham risk score improved the net reclassification index (NRI) by only 0.8% (95% CI, 0.1–1.6%) in the overall population, and by only 3.6% in the intermediate-risk subgroup.10 Mean CIMT alone therefore offers limited incremental discrimination. Adding plaque assessment consistently outperforms CIMT alone.20,30 Combined use of the two indices is preferred in clinical practice.

Operator and equipment dependence

CIMT accuracy depends heavily on operator skill and equipment specifications. Non-standardised measurement undermines clinical decision-making.21 Before referring a patient, the primary care physician should verify that the requested study follows a Mannheim- or ASE-compliant protocol. The report should include (i) measurement site, (ii) mean and maximum CIMT, (iii) plaque presence and characteristics, and (iv) stenosis grade.

Optimal candidates

The available evidence supports CIMT testing in the following groups: (i) intermediate-risk individuals (10-year ASCVD risk 5–20%) in whom statin initiation is borderline; (ii) patients with a strong family history of early atherosclerotic disease; (iii) patients with suspected familial hypercholesterolaemia; (iv) patients with chronic inflammatory disorders (rheumatoid arthritis, systemic lupus erythematosus, psoriasis) at risk of accelerated atherosclerosis; and (v) pregnant or lactating women, and other young women, in whom CACS is undesirable or unavailable.13-15 Routine testing is not advised in patients with established atherosclerotic cardiovascular disease, in those at very low risk (10-year risk <5%), and in those with limited life expectancy.

Evolving clinical role of CIMT

The clinical role of CIMT has shifted markedly over three decades. Consistent predictive performance in ARIC, Rotterdam, and CHS in the 1990s established CIMT as a “novel marker of cardiovascular risk.”5,6,8 Many guidelines subsequently endorsed it as a risk-stratification tool in the general asymptomatic population. Two major criticisms emerged in the 2010s. First, measurement standardisation proved difficult, with operator-dependent variability of 0.05–0.10 mm common even in the same patient.21 Second, the USE-IMT meta-analysis reported an NRI of only 0.8%, calling the incremental value of CIMT alone into question.10 Subsequently, ACC/AHA, ESC, and the U.S. Preventive Services Task Force (USPSTF) ceased recommending CIMT as a screening tool, and CACS took its place.11-13,43 This trajectory should be read not as the obsolescence of CIMT but as the refinement of its appropriate niche.

The proposed five-tier framework is author-derived

The five-tier interpretive framework presented in Fig. 2 and Table 4, and the associated follow-up intervals, are explicitly an author-proposed pragmatic integration of existing evidence rather than a recommendation supported by any single guideline. The source documents from which it is drawn—the ASE consensus,3 the Mannheim consensus,22 the 2019 ESC/EAS guidelines,27 the 2023 ESVS guidelines,23 and the 2018 Korean Society of Neurosonology consensus on extracranial carotid duplex ultrasonography28—do not, in general, specify explicit follow-up intervals for Categories I–IV. Only the ESVS 2023 document specifies an explicit interval (annual imaging for asymptomatic 50–69% stenosis under medical management).23 The intervals proposed here were therefore set conservatively, wide enough that true biological change will exceed the test–retest variability of CIMT,21,22 and should be regarded as hypothesis-generating until validated in dedicated prospective studies.

Decoupling mean CIMT from plaque

A conceptual advance in recent years has been the explicit separation of mean CIMT from plaque as clinical signals. Spence and Hackam argued for “treating arteries instead of risk factors,” advocating titration of therapy to a measurable arterial phenotype.44 The relevant phenotype, however, is closer to plaque burden than to mean CIMT. In Inaba et al.’s meta-analysis (54,336 individuals), plaque conferred a relative risk for coronary events of 1.83 (95% CI, 1.55–2.16), whereas mean CIMT yielded only 1.30 (1.21–1.40).30 Plaque should therefore be reported qualitatively and quantitatively, not as a binary annotation appended to a thickness value.

Integration with other imaging and biomarkers

Single imaging markers have intrinsic limitations, and integrated multi-marker models are increasingly being studied. CACS outperforms CIMT for risk reclassification but carries radiation and cost penalties.11,43 Ankle–brachial index, pulse-wave velocity, lipoprotein(a), high-sensitivity C-reactive protein, and polygenic risk scores each contribute independent information.45-48 In a MESA post-hoc analysis, combining CACS, ABI, and hs-CRP improved NRI by 5–7% over any single marker.49 Future research should test how CIMT and plaque integrate into such models. In primary care, where cost and access dominate decision-making, CIMT may serve as an entry point to a multi-marker strategy.

Artificial-intelligence–based automation

Deep-learning automatic segmentation has the potential to resolve much of the standardisation problem. Menchón-Lara et al.50 reported inter-observer CV below 1.5% with a convolutional neural network.51 Laine et al. demonstrated that a U-Net model agreed with manual measurement within a mean absolute error of 0.024 mm.52 Recent clinical validation suggests that AI-assisted measurement (i) reduces acquisition time by approximately 60%, (ii) improves inter-reader agreement to 0.92, and (iii) raises non-expert performance to expert level.53 Such tools could become essential infrastructure for primary-care CIMT use. Multicentre, multi-ethnic validation remains a prerequisite. Training-data bias and the black-box nature of these models also need to be addressed before clinical deployment.

Special populations

Traditional risk equations were calibrated on general populations and therefore tend to underestimate risk in selected groups. CIMT can fill this gap. (i) Autoimmune disease: rheumatoid arthritis, lupus, and psoriasis are associated with accelerated atherosclerosis. CIMT can quantify subclinical disease.53 The EULAR recommendations include CIMT as an adjunctive risk-assessment tool in rheumatoid arthritis.54 (ii) Familial hypercholesterolaemia: cumulative LDL exposure begins early, and CIMT helps to guide intensity and follow-up of therapy.55 (iii) Chronic kidney disease: uraemic vascular ageing exceeds what traditional risk factors predict. KDIGO acknowledges CIMT as a useful tool.56 (iv) Women with a history of gestational hypertension or pre-eclampsia have elevated long-term cardiovascular risk, and CIMT can support post-partum surveillance.57 (v) People living with HIV experience accelerated atherosclerosis from chronic inflammation and antiretroviral therapy, and serial CIMT is supported by available data.58 Population-specific reference data are needed for each of these groups.

Patient education and behaviour change

CIMT also carries practical value beyond statistical prediction: a visible, quantitative result is a strong motivator for behaviour change. The “vascular age” concept has been associated with improved adherence to risk-factor management.59 In the BioImage study, patients who received image-based risk information showed 33% higher statin adherence than those who did not. Whether such behaviour changes translate into long-term event reduction requires further study.

Cost-effectiveness and the health-system perspective

CIMT testing in Korea costs approximately KRW 50,000–100,000, against KRW 150,000–250,000 for CACS and KRW 300,000–500,000 for coronary CT angiography. Because the reclassification yield of mean CIMT is modest, universal screening is unlikely to be cost-effective. Selective application in intermediate-risk patients and in special populations may, however, prove cost-effective. Formal health-economic modelling is needed. CIMT remains a non-reimbursed item in the Korean national health insurance system. Refining its indication should precede any discussion of reimbursement.

Limitations of this review and future directions

This review has several limitations. First, the Korean reference values cited represent pooled estimates from subgroup analyses of publicly available cohorts (KoGES, KSHAP); authoritative nationwide normative data covering all age strata, derived from a single standardised protocol, are not yet available. Table 3 should therefore be regarded as illustrative rather than as a reference standard. Second, this is a narrative rather than a systematic review and is subject to citation bias. Third, the proposed five-tier clinical categories and follow-up intervals integrate published guidelines and meta-analyses but, as emphasised above, are author-derived and have not been validated in dedicated prospective trials. Priority areas for future research include: (i) establishment of nationwide Korean reference values across the age spectrum; (ii) development of multidimensional risk models incorporating plaque volume, echogenicity, and neovascularisation; (iii) multicentre validation of AI-based automated CIMT systems; (iv) prospective studies on integration of CIMT with CACS, apolipoprotein B, lipoprotein(a), and polygenic risk scores; and (v) cost-effectiveness analyses in special populations (autoimmune disease, familial hypercholesterolaemia, chronic kidney disease, and adverse pregnancy outcomes).