Relationship between Vascular Tortuosity Rate and Atherosclerotic Risk Factors in Patients with Carotid Artery Stenosis Authors: Masanori Tsuji, Fujimaro Ishida, Yoichi Miura, Takenori Sato, Yume Suzuki, Hideki Kanamaru, Kazuhiro Furukawa, Takeshi Okada, Keiji Fukazawa, Ryuta Yasuda, Yasuyuki Umeda, Naoki Toma, Hidenori Suzuki Published: 2026 ABSTRACT AbstractPurposeThere is a significant association between carotid artery (CA) stenosis and atherosclerosis, with those patients often presenting with carotid tortuosity. However, the relationship between atherosclerotic risk factors and carotid tortuosity remains controversial. Previous assessments of tortuosity have predominantly utilized qualitative 2D methods. This retrospective study aimed to quantitatively evaluate the tortuosity of CAs in 3D geometry and to investigate the relationship with atherosclerotic risk factors.MethodsA total of 168 CAs in 96 patients diagnosed with at least either-side CA stenosis on magnetic resonance angiography (MRA) between 2011 and 2019 were included in this single-center study, following the exclusion of post-treatment CAs and those with poor imaging quality. The 3D geometries of CAs were segmented using MRA and the centerlines were generated. Vascular tortuosity rate (VTR) was calculated as the curved distance divided by the linear distance between the specified 2 points. The relationship between VTRs and 8 atherosclerotic risk factors—age, body mass index, systolic blood pressure, hemoglobin A1c, high-density lipoprotein, total cholesterol (TC), non-fasting serum triglyceride (TG), and Brinkman index—was analyzed.ResultsHigher VTRs were associated with ≥ 65 years of age, body mass index ≥ 25 [kg/m2], serum TC ≥ 220 [mg/dL], and non-fasting serum TG ≥ 175 [mg/dL] in common CAs, and body mass index ≥ 25 [kg/m2], and systolic blood pressure ≥ 140 [mmHg] in internal CAs.ConclusionQuantitative assessment of carotid tortuosity revealed significant associations with several atherosclerotic risk factors, including obesity, hypertension, advanced age, and dyslipidemia, suggesting that VTRs may reflect the presence of these risk factors.Keywords: atherosclerosis, carotid artery stenosis, vascular tortuosity rate Introduction There is a strong association between carotid artery (CA) stenosis and atherosclerosis.1,2 Patients with CA stenosis also tend to have more carotid tortuosity compared to healthy individuals. However, past reports have been inconclusive as to the relationship between carotid tortuosity and atherosclerosis, with some studies showing an association between them,3 while the others showing no association.4 The reported morbidity of carotid tortuosity ranges widely from 6% to 66%.5 Several factors contributed to these discrepancies. First, different measurement methods such as color doppler ultrasonography, magnetic resonance angiography (MRA), and digital subtraction angiography (DSA) have been used, with most of tests being qualitative.3–7 Qualitative tests, which classified the tortuosity of CAs as flexion, coiling, and kinking,4,6 may result in lower inter-rater reliability due to ambiguous classification. Secondly, many previous reports evaluated CAs in the 2D, neglecting the 3D structure of the actual CA.3–5 This 2D evaluation has risks of underestimating the tortuosity present in the complex 3D structure. Several previous studies reported 3D tortuosity metrics in intracranial arteries using non-contrast MRA.8 In addition, the reproducibility of carotid tortuosity measurements across multiple software platforms has been demonstrated using contrast-enhanced computed tomography angiography (CTA).9 Furthermore, Chen et al. evaluated internal CA (ICA) tortuosity using 3D CTA with curved planar reformats.10 However, most of these studies9,10 relied on contrast-enhanced CTA rather than non-contrast MRA8 and mainly focused on methodological validation, and the relationship between carotid tortuosity and atherosclerotic risk factors remains unclear. This retrospective study aimed to clarify the association between carotid tortuosity and atherosclerotic risk factors by performing 3D quantification from the common CA (CCA) to the ICA using non-contrast MRA and commonly available software. Materials and Methods This retrospective study was approved by the Ethics Committee at Mie University (approval number H2020-020). The requirement for written informed consent was waived due to the retrospective nature of the study, and an opt-out approach was applied in accordance with institutional policies. Although this study was conducted at a single institution, several co-authors from other institutions contributed through technical advice or by referring some of the study patients. Patient population Between November 2011 and September 2019, patients diagnosed with CA stenosis who underwent non-contrast MRA using a single 3.0T MRI scanner (Ingenia; Philips Medical Systems, Best, the Netherlands) at our institution were considered for inclusion in this study. CA stenosis was identified based on MRA findings in routine clinical practice. The assessment on MRA was qualitative. The degree of stenosis was subsequently quantified using a standardized method, as described below. At our institution, all MRI examinations requested from the Department of Neurosurgery, including both brain and cervical imaging, are routinely performed using this system, and no selection of MRI system was made. Cervical MRA was performed in patients with suspected CA stenosis on ultrasonography and in those referred from other institutions after abnormalities were identified on prior imaging studies. In addition, cervical MRA is included as part of the standard imaging protocol for routine brain MRI, allowing incidental detection of CA stenosis, and such cases were also included in this study. Both symptomatic and asymptomatic cases were included. In this study, patients with at least unilateral CA stenosis were included, and both CAs (bilateral CAs) in each patient were analyzed. Exclusion criteria were (1) CAs after carotid endarterectomy or CA stenting, (2) CAs with poor imaging quality, in which vascular geometry and the centerline could not be extracted, primarily due to severe CA stenosis, and (3) patients with a history of connective tissue diseases that could affect vascular tortuosity.11 Thus, 5 post-treatment CAs and 16 CAs with too severe CA stenosis were excluded from 220 CAs in 110 patients. No patients had a history of connective tissue diseases. Consequently, a total of 199 CAs in 109 patients were analyzed in this study. Quantitative evaluation of carotid tortuosity Cervical MRA was performed using a 3.0T MRI scanner (Ingenia; Philips Medical Systems) with the following scan parameters: time-of-flight (TOF) sequence, sensitivity encoding (SENSE) factor 3.1, TR = 25 ms, TE = 3.4 ms, flip angle = 18°, FOV = 200 × 200 mm2, matrix size = 432 × 303, slice thickness = 0.5 mm, and number of slices = 190. No contrast agent was used. Cervical MRA was extracted as digital imaging and communications in medicine datasets to Mimics Innovation Suite version 24.0 (Materialise Japan, Yokohama, Japan). Arterial lumen was segmented using a thresholding technique adapted from our previous computational fluid dynamics (CFD)–based hemodynamic analysis of cerebral aneurysms,12,13 in which the threshold was determined based on intra-arterial CT values obtained from CTA. In the present study, the threshold was applied to signal intensities obtained from TOF MRA and was set at 40% of the intensity range between the maximum intravascular signal within the CCA and the minimum extravascular background signal. Here, the extravascular background signal was defined as the signal intensity measured in non-vascular regions adjacent to the target artery, such as surrounding soft tissue outside the vessel lumen. This segmentation approach is consistent with the standardized method routinely used in our CFD research group and previously reported.12,13 Using this method, 3D vascular structures of the CAs were reconstructed (Fig. 1A). The centerlines of the 3D vessel geometries were generated using Mimics Innovation Suite version 24.0 (Fig. 1B),14,15 ensuring low inter-examiner error. The centerlines were then exported to 3-matic version 17.0 (Materialise Japan) as stereolithography files (Fig. 1C). The centerlines from the proximal end of the CCA to the distal end of the ICA and the external CA (ECA) near the CCA bifurcation were retained, while the other branch centerlines were deleted (Fig. 1D). The proximal point of the CCA, the distal point of the ICA, and the branching point of the ECA from the CCA on the centerline were defined as the points c, i, and b, respectively (Fig. 1E). The vascular tortuosity rate (VTR) was defined as the curved distance on the centerline divided by the linear distance between the 2 points (Fig. 1F). For the notation of the VTRs, for example, VTRbc was defined as the VTR between points b and c. Visualization of the process of measuring VTR. Extraction of 3D structure of CAs as stereolithography based on MRA (A). Centerline setting (B). Extraction of centerline only (C). Keeping the centerline in the proximal end of the CCA to distal end of the ICA and the ECA near the CCA bifurcation, and deleting the other centerlines (D). The proximal point of CCA, the distal point of ICA, and the ECA branching point from the CCA on the centerline are defined as points c, i, and b, respectively (E). VTR between 2 points is defined as the curved distance on the centerline divided by the linear distance between the 2 points (F). This figure shows the measurement of VTRbi between positions b and i.CA, carotid artery; CCA, common carotid artery; ECA, external carotid artery; ICA, internal carotid artery; MRA, magnetic resonance angiography; VTR, vascular tortuosity rate. In measuring VTRs, 3 issues needed to be addressed. First, it was necessary to define the measurement range of the CAs accurately. Ideally, point i should correspond to the location where the ICA enters the skull base. If the linear distance between points b and i was too short, VTRbi could not properly assess tortuosity. However, as this was a retrospective study and the imaging coverage of cervical MRA was not consistent across patients, point i could not always be set at the anatomically ideal location. Therefore, the measurement range was determined based on the maximum overlapping imaging range available among all included cases. According to McNamara et al.,16 the median of linear distance from the proximal end of the ICA to the point where the ICA entered the base of skull was 72.1 (48.0–98.1) [mm] based on the measurements of 172 CAs using 3D CTA. In this study, the linear distances between these points were defined as bc (between points b and c), bi (between points b and i), and ci (between points c and i). Thus, CAs with a bi of less than 48 mm were excluded. Because applying this bi threshold preferentially retained arteries with longer bi, the relative proportion of bc within the measured segments became shorter. Therefore, the cutoff value for bc was set at 24 mm (half of that for bi), and arteries with distances less than 24 mm were excluded. Secondly, it was necessary to examine the influence of the anatomical factors on carotid tortuosity.5 To address this issue, the differences in VTR between the right and left sides were examined, and the correlation between the VTRs of both sides was analyzed. Thirdly, it was theoretically considered that the longer the linear distance between bc, bi, and ci, the more tortuous the CAs in those portions would be. Therefore, we examined whether there were any significant differences in the linear distance between each of these points. The degree of CA stenosis was measured by the North American Symptomatic Carotid Endarterectomy Trial (NASCET) method17 on 3D rotational DSA obtained using the Allura system (Philips Medical Systems). Risk factors for atherosclerosis We selected 9 categories based on the guidelines and reports on atherosclerotic risk factors.18–23 Specifically, the following 9 atherosclerotic risk factors were included: age ≥ 65 years old;18,24 body mass index (BMI) ≥ 25 [kg/m2];18 systolic blood pressure (SBP) ≥ 140 [mmHg];7 serum low-density lipoprotein (LDL) ≥ 140 [mg/dL];18 serum high-density lipoprotein (HDL) < 40 [mg/dL]; non-fasting serum triglyceride (TG) ≥ 175 [mg/dL];18 serum total cholesterol (TC) ≥ 220 [mg/dL];18 serum hemoglobin (Hb) A1c ≥ 7 [%] measured by the National Glycohemoglobin Standardization Program method;20,25 and Brinkman index (BI) ≥ 600.18,26 Nine atherosclerotic risk factors were initially defined. However, if no patients met the predefined criterion for a given factor, that factor was excluded from the statistical analysis. The medication history and laboratory values at the time of admission were used as references, and VTRs were compared between the high-risk and low-risk groups for atherosclerotic risk factor according to predefined reference values for each category. Patients who met the cutoff value for a given risk factor were assigned to the high-risk group, whereas those who did not meet the criterion were assigned to the low-risk group. Because vascular tortuosity is unlikely to regress with pharmacological treatment, patients receiving medication for a given risk factor were classified into the high-risk group, even if their laboratory values were within the reference range. Statistical analysis Measurement values were expressed as medians (interquartile ranges [IQR]), and categorical variables were expressed as numbers and percentages. Statistical analysis was performed with EZR (Easy R).27 Differences in VTRs between right- and left-sided CAs and between groups stratified according to predefined clinical cutoff values for each atherosclerotic risk factor were assessed using the Brunner–Munzel test. Correlation analyses between right- and left-sided VTRs were performed using Spearman’s rank correlation coefficient in patients in whom both CAs were included in the analysis. Spearman’s rank correlation coefficient was used to evaluate multicollinearity among atherosclerotic risk factors showing significant associations with VTRs on univariate analyses. P-values less than 0.05 were considered significant. Results Of the 199 CAs in 109 patients, 14 CAs in 10 patients had a linear distance between points b and i of less than 48 mm and 19 CAs in 13 patients had a linear distance between points b and c of less than 24 mm: they were therefore excluded. Finally, 168 CAs from 96 patients were included in the analysis, and their clinical characteristics are shown in Tables 1 and 2. The linear distances among the measurement points were as follows: bc, median 43 mm (IQR 35–50); bi, median 67 mm (IQR 62–74); and ci, median 109 mm (IQR 108–113). These values represent the distances calculated from CAs that met the predefined inclusion criteria. Among the 9 predefined atherosclerotic risk factors, no patients met the criterion for elevated serum LDL levels (≥ 140 mg/dL) at admission; therefore, the remaining 8 risk factors were included in the statistical analysis. Table 1Background characteristics of patients included in the analysisPatient characteristicsValuesNumber of patients96Age, years72 (68–77)Male91, 94.8%BMI, kg/m223.1 (21.5–24.9)Current smoker11, 11.5%SBP, mmHg131 (118–142)Serum HDL*, mg/dL48 (41–58)Serum LDL*, mg/dL95 (80–112)Non-fasting serum TG*, mg/dL110 (76–170)Serum TC*, mg/dL163 (138–188)Values are presented as a median (interquartile range) or the number of patients (a percentage of all patients).*The number of patients with available data for each risk factor: serum HDL, 81; serum LDL, 82; non-fasting serum TG, 94; and serum TC, 93.BMI, body mass index; Hb, hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride. Background characteristics of patients included in the analysis Values are presented as a median (interquartile range) or the number of patients (a percentage of all patients). *The number of patients with available data for each risk factor: serum HDL, 81; serum LDL, 82; non-fasting serum TG, 94; and serum TC, 93. BMI, body mass index; Hb, hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride. Table 2Characteristics of carotid arteries included in the analysisCarotid arteries characteristicsNumber of vessels and percentage of the totalNumber of carotid arteries168Symptomatic carotid artery28, 16.7%Right-sided carotid arteries79, 47.0%Degree of carotid artery stenosis by the NASCET method< 30 %77, 45.8%30–49 %8, 4.8%50–69 %19, 11.3%70–99 %64, 38.1%NASCET, North American Symptomatic Carotid Endarterectomy Trial. Characteristics of carotid arteries included in the analysis NASCET, North American Symptomatic Carotid Endarterectomy Trial. As shown in Fig. 2, VTRbc was significantly higher in patients with ≥ 65 years of age, BMI ≥ 25 [kg/m2], serum TC ≥ 220 [mg/dL] and non-fasting serum TG ≥ 175 [mg/dL]; and VTRbi as well as VTRci were significantly higher in those with BMI ≥ 25 [kg/m2], and SBP ≥ 140 [mmHg]. Serum HbA1c, serum HDL, and BI were not related to any of VTRs. Among the items that showed significant differences, none exhibited a significantly longer linear distance between any of the 2 points. Representative cases demonstrating low VTRs with minimal atherosclerotic risk factors, and those with high VTRs with multiple risk factors, are presented in Fig. 3. Comparison of VTRs between low- and high-risk groups for each atherosclerotic risk factor. The graphs show the statistical evaluation of VTRs between the low-risk and high-risk groups for each atherosclerotic risk factor. The error bars represent the interquartile range. Statistical significance was assessed using the Brunner–Munzel test: *P < 0.05; **P < 0.01; ***P < 0.001. (A) Age; (B) BMI; (C) SBP; (D) HbA1c; (E) HDL; (F) TC; (G) Non-fasting serum TG; (H) BI. No patients met the criterion for elevated serum LDL levels (≥ 140 mg/dL); therefore, LDL was excluded from the analysis. VTRbc indicates the vascular tortuosity rate between points b and c; VTRbi, between points b and i; and VTRci, between points c and i.BI, Brinkman index; BMI, body mass index; Hb, hemoglobin; HDL, high-density lipoprotein; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride; and VTR, vascular tortuosity rate. Differences in vascular structures between representative low and high atherosclerotic risk cases. An 82-year-old man with minimal atherosclerotic risk factors (BMI of 22.7, SBP of 132 mmHg [no antihypertensive medication], TC of 136 mg/dL, and non-fasting serum TG of 96 mg/dL [no statin therapy]) shows that VTR values are 1.02 for VTRbc, 1.09 for VTRbi, and 1.07 for VTRci in his right carotid artery (A). A 77 year-old man with multiple atherosclerotic risk factors (BMI of 27.5, SBP of 141 mmHg [on antihypertensive medication], TC of 162 mg/dL, and non-fasting serum TG of 115 mg/dL [on statin therapy]) shows a tortuous vascular structure and that VTR values are 1.05 for VTRbc, 1.34 for VTRbi, and 1.22 for VTRci in his left carotid artery (B). In (A) and (B), the surface-rendered images are on the left and the corresponding centerlines of the vessels are on the right. Compared to the case with a low risk of atherosclerosis (A), the case with a high risk (B) exhibits a more tortuous vascular structure and higher VTR values.BMI, body mass index; CA, carotid artery; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride; and VTR, vascular tortuosity rate. In the comparison of VTRs between the right- and left-sided CA, VTRbi and VTRci were significantly larger on the left side (Table 3). The comparison of VTRs between the right- and left-sided CAs was performed using all analyzed CAs (168 CAs from 96 patients). The correlation coefficients between right- and left-sided VTRbc, VTRbi, and VTRci were 0.42, 0.68, and 0.67, respectively, in 55 patients in whom bilateral CAs were included in the analysis, suggesting that VTR was influenced by anatomical factors. Non-fasting serum TG levels were weakly correlated with BMI and serum TC levels, when correlation was examined between the categories with significantly different VTRs on univariate analyses (Table 4). A supplementary analysis including arteries excluded by the cutoff criteria was also conducted, confirming the appropriateness of the thresholds (Supplementary Fig. 1, Supplementary Tables 1 and 2). Table 3Comparison of VTRs between the right- and left-sided carotid arteriesVTRsRight-sided CAs (n = 79)Left-sided CAs (n = 89)P Value*Correlation coefficientVTRbc1.02 (1.02–1.04)1.03 (1.02–1.05)0.210.42VTRbi1.11 (1.08–1.19)1.15 (1.09–1.24)0.040.68VTRci1.08 (1.06–1.13)1.11 (1.07–1.16)0.020.67Values are presented as a median (interquartile range).*P values were estimated using the Brunner–Munzel test, and correlation coefficients were calculated using Spearman’s rank correlation coefficient between right- and left-sided CAs.CA, carotid artery; VTRbc, vascular tortuosity rate between points b and c; VTRbi, between points b and i; VTRci, between points c and i. Comparison of VTRs between the right- and left-sided carotid arteries Values are presented as a median (interquartile range). *P values were estimated using the Brunner–Munzel test, and correlation coefficients were calculated using Spearman’s rank correlation coefficient between right- and left-sided CAs. CA, carotid artery; VTRbc, vascular tortuosity rate between points b and c; VTRbi, between points b and i; VTRci, between points c and i. Table 4Spearman’s rank correlation coefficients between the atherosclerotic risk factors with significant differences of VTRs on univariate analysesVariable pairCorrelation coefficient (P value)*Age vs. BMI− 0.07 (P = 0.30)Age vs. SBP− 0.08 (P = 0.25)Age vs. Serum TC− 0.04 (P = 0.60)Age vs. Non-fasting serum TG0.01 (P = 0.88)BMI vs. SBP− 0.01 (P = 0.86)BMI vs. Serum TC0.05 (P = 0.51)BMI vs. Non-fasting serum TG0.26 (P < 0.01)SBP vs. Serum TC0.15 (P = 0.04)SBP vs. Non-fasting serum TG0.06 (P = 0.37)Serum TC vs. Non-fasting serum TG0.34 (P < 0.01)*Correlation coefficients and P-values were calculated using Spearman’s rank correlation coefficient.BMI, body mass index; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride; and VTR, vascular tortuosity rate. Spearman’s rank correlation coefficients between the atherosclerotic risk factors with significant differences of VTRs on univariate analyses *Correlation coefficients and P-values were calculated using Spearman’s rank correlation coefficient. BMI, body mass index; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride; and VTR, vascular tortuosity rate. Discussion This study demonstrated that carotid tortuosity was significantly associated with several atherosclerotic risk factors, including higher age, BMI, lipid levels, and blood pressure (Fig. 2A, B, C, F, G). These findings suggested that carotid tortuosity, quantified using a standardized, non-contrast MRA-based 3D method, may serve as a noninvasive marker of systemic atherosclerosis. Wang et al. reported that carotid tortuosity is more likely to occur on the left side because the left CCA originates from the aortic arch and follows a nearly straight path from its origin to the skull.5 Although VTRbi and VTRci were significantly greater on the left side, significant bilateral correlations were also observed for these parameters (Table 3). These findings suggest that carotid tortuosity may be influenced by both laterality-related anatomical characteristics and atherosclerotic risk factors. Because VTRbi and VTRci reflect tortuosity over longer arterial segments than VTRbc, these parameters may more sensitively capture anatomical differences in vascular tortuosity, including possible contributions from segments involving the ICA. However, further studies are needed to clarify the anatomical characteristics and influence of atherosclerotic risk factors underlying these laterality differences. VTRbc was also related to some of atherosclerotic risk factors, that is, ≥ 65 years of age (Fig. 2A), BMI ≥ 25 [kg/m2] (Fig. 2B), serum TC ≥ 220 [mg/dL] (Fig. 2F), and non-fasting serum TG ≥ 175 [mg/dL] (Fig. 2G). Although VTRci is a combination of VTRbc and VTRbi, associated factors of VTRci were the same as those of VTRbi, and were BMI ≥ 25 [kg/m2] (Fig. 2B), and SBP ≥ 140 [mmHg] (Fig. 2C). As the atherosclerotic risk factors should influence bilateral CAs similarly, thus, bilateral carotid tortuosity may be associated with patients with ≥ 65 years of age (Fig. 2A), BMI ≥ 25 [kg/m2] (Fig. 2B), SBP ≥ 140 [mmHg] (Fig. 2C), and dyslipidemia, especially serum TC ≥ 220 [mg/dL] (Fig. 2F) and non-fasting serum TG ≥ 175 [mg/dL] (Fig. 2G). Several reports have examined the mechanism by which atherosclerotic risk factors contribute to carotid tortuosity. Aging, for instance, is associated with fragmentation and reduction of elastic fibers within the walls of the CAs.28 This leads to an increase in carotid tortuosity with age.28 Our study found that VTRbc was predominantly influenced by aging, suggesting that aging augmented hemodynamic factors–induced CA tortuosity associated with degenerated elastic fibers (Fig. 2A). Increased intra-abdominal pressure has been identified as a potential mechanism for carotid tortuosity related to obesity.5,28–30 Higher BMI is associated with increased visceral fat, which raises intraabdominal pressure,30 leading to upward deviation of the diaphragm and thoracic aorta, and consequently the tortuosity of the CAs.28,29 Previous studies also reported a positive correlation between the carotid tortuosity and BMI.5 These findings were consistent with our study, where all VTRs were significantly higher in the high BMI group (≥ 25 kg/m2) (Fig. 2B). The association between hypertension and carotid tortuosity has remained inconclusive, with some reports suggesting a relationship due to increased endoluminal pressure caused by elevated blood pressure,3,7 while others have found no such association.4,5 However, all of these studies used 2D evaluation methods. Our study performed a 3D evaluation and demonstrated the association between hypertension and carotid tortuosity (Fig. 2C). Although the relationship between non-fasting serum TG and carotid tortuosity has not been previously reported, our study found a weak correlation between non-fasting serum TG and BMI (Table 4), suggesting that carotid tortuosity in patients with higher non-fasting serum TG levels (Fig. 2G) may be partly attributable to obesity. Furthermore, this study identified a relationship between elevated serum TC levels and carotid tortuosity (Fig. 2F). Further studies are needed to explore the causal relationships. Previous studies have investigated vascular tortuosity using different imaging modalities and anatomical segments. Shang et al. evaluated the association between tortuosity from the CCA to the ICA and acute ischemic stroke using MRA.31 Chen et al. assessed ICA tortuosity using contrast-enhanced CTA with curved planar reformats and examined its relationship with cerebral small vessel disease.10 In contrast, the present study evaluated carotid tortuosity using non-contrast MRA, focusing on the distal CCA and proximal ICA around the carotid bifurcation, and investigated its association with individual atherosclerotic risk factors, which have not been comprehensively assessed in previous studies. These findings highlighted the potential utility of carotid tortuosity, as assessed by non-contrast MRA, as a noninvasive marker of systemic atherosclerosis in routine clinical practice. The reproducibility of VTR measurement in this study was supported by the use of Mimics software for vascular reconstruction, which has demonstrated interobserver variability of less than 5%.32 To ensure reliability, arteries that did not meet predefined anatomical criteria were excluded. Centerlines were then consistently extracted using automated processing in 3-matic. In the present study, 3D morphological evaluation of CAs was performed using image processing techniques originally developed for CFD analyses at our institution.12,13 These techniques enabled accurate reconstruction of vascular geometry and extraction of centerlines from non-contrast MRA data. While recent studies using CFD have provided insights into the local hemodynamic environment of CA disease,33,34 few have examined morphological characteristics such as carotid tortuosity in relation to systemic atherosclerotic risk factors. Our findings highlighted the potential value of tortuosity metrics as anatomical markers reflecting systemic vascular pathology and may complement hemodynamic analyses in future comprehensive assessments. From a clinical perspective, quantitative assessment of carotid tortuosity using non-contrast MRA may serve as a practical and noninvasive imaging marker that reflects systemic atherosclerotic risk. Increased tortuosity metrics may help identify patients with a higher burden of atherosclerotic risk factors, even in the absence of significant stenosis, and may facilitate individualized risk stratification. Furthermore, tortuosity assessment could also complement conventional measures during patient evaluation and follow-up. There are several limitations in this study. First, being a retrospective study, the examination items related to the atherosclerotic risk factors were not consistent. This inconsistency may have affected the accuracy of the associations observed. Secondly, the degree of CA stenosis and the presence of symptoms varied, and this study did not clarify the causal relationship between CA stenosis and tortuosity. Prospective studies in patients with no or mild CA stenosis are needed to establish this relationship. Lastly, this study was based on MRA, which might not exactly visualize the geometry of CAs. Further studies should consider using 3D CTA and 3D rotational DSA for more accurate assessment. Despite these limitations, our study’s 3D assessment of carotid tortuosity may be more reliable than previous 2D evaluation.3–5,28–30 Quantitative evaluation of 3D CA tortuosity may provide important insights or breakthroughs in understanding the relationship between carotid tortuosity and atherosclerosis, which remains inconclusive.3,4 Conclusion Our study suggested that the presence of specific atherosclerotic risk factors—such as obesity, hypertension, advanced age, and dyslipidemia—could be assessed based on VTRs. Funding Statement This study was supported by the Suzuken Memorial Foundation (grant number 23-079). Conflicts of Interest: The authors declare that they have no conflicts of interest. Supplementary Materials: Supplementary Fig. 1 Comparison of VTRs between low- and high-risk groups for each atherosclerotic risk factor including arteries excluded by the cutoff criteria. The graphs show the statistical evaluation of VTRs between the low-risk and high-risk groups for each atherosclerotic risk factor, including arteries that were excluded by the predefined cutoff criteria in the main analysis. The error bars represent the interquartile range. Statistical significance was assessed using the Brunner–Munzel test: *P < 0.05; **P < 0.01; ***P < 0.001. (A) Age; (B) BMI; (C) SBP; (D) HbA1c; (E) HDL; (F) TC; (G) non-fasting serum TG; (H) BI. No patients met the criterion for elevated serum LDL levels (≥ 140 mg/dL); therefore, LDL was excluded from the analysis. VTRbc indicates the vascular tortuosity rate between points b and c; VTRbi, between points b and i; and VTRci, between points c and i. BI, Brinkman index; BMI, body mass index; Hb, hemoglobin; HDL, high-density lipoprotein; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride; VTR, vascular tortuosity rate. Supplementary Table 1 Comparison of VTRbi values between bi < 48 mm and bi ≥ 48 mm groups. Supplementary Table 2 Comparison of VTRbc values between bc < 24 mm and bc ≥ 24 mm groups. Supplementary Materials Supplementary figures are available online. Supplementary Table 1Comparison of VTRbi values between bi < 48 mm and bi ≥ 48mm groups Comparison of VTRbi values between bi < 48 mm and bi ≥ 48mm groups Supplementary Table 2Comparison of VTRbc values between bc < 24 mm and bc ≥ 24mm groups Comparison of VTRbc values between bc < 24 mm and bc ≥ 24mm groups References