ABSTRACT
Background
The cardio‐ankle vascular index (CAVI) is a blood pressure‐independent measure of arterial stiffness. While its association with cardiovascular disease (CVD) is well established, its prognostic value for non‐CVD mortality and its overall public health impact in the general population remain unclear. This study aimed to evaluate CAVI as a comprehensive predictor of cause‐specific mortality and to quantify its population attributable fraction (PAF).
Methods
In this 12.7‐year prospective cohort study, we examined 5994 community‐dwelling adults (aged 35–69 years) from the Japan Multi‐Institutional Collaborative Cohort Study. Baseline data (2005–2012) and mortality outcomes through December 2021 were analyzed. Participants were categorized into three groups: CAVI 1 (< 8.0), CAVI 2 (8.0–8.9), and CAVI 3 (≥ 9.0). Adjusted hazard ratios and 95% confidence intervals were estimated using Cox proportional hazards models. PAF was used to quantify the impact of elevated CAVI.
Results
During follow‐up, 295 deaths occurred. Compared with the CAVI 1 group, the CAVI 3 group showed increased risks of all‐cause (hazard ratio: 1.38, 95% confidence interval: 1.00–1.90), CVD (hazard ratio: 2.17, 95% confidence interval: 0.97–4.86), and non‐CVD mortality (hazard ratio: 1.26, 95% confidence interval: 0.88–1.79). PAFs were 8.9%, 23.4%, and 6.2% for all‐cause, CVD, and non‐CVD mortality, respectively.
Conclusion
Elevated arterial stiffness (CAVI ≥ 9.0) is an independent predictor of all‐cause mortality in the general Japanese population. Furthermore, a substantial proportion of CVD deaths (approximately 23.4%) could potentially be prevented by maintaining a CAVI below 9.0, highlighting the clinical utility of CAVI screening in primary prevention and nursing management.
Keywords: all‐cause mortality, arterial stiffness, cardio‐ankle vascular index, cardiovascular mortality, population attributable fraction
This 12.7‐year prospective study of 5,994 adults evaluated the cardio‐ankle vascular index (CAVI) as a mortality predictor. Compared to normal CAVI (< 8.0), elevated CAVI (≥ 9.0) significantly increased the risk for all‐cause mortality (HR: 1.38; PAF: 8.9%). Hazard ratios for cardiovascular (CVD) (HR: 2.17; PAF: 23.4%) and non‐CVD mortality (HR: 1.26; PAF: 6.2%) were elevated, but did not reach statistical significance. Elevated CAVI serves as an independent predictor of all‐cause mortality, functioning as a comprehensive biomarker for systemic vascular aging and early risk stratification.

1. Introduction
Arterial stiffness is an important indicator of cardiovascular health and is associated with cardiovascular diseases (CVD) and all‐cause mortality [1, 2]. The cardio‐ankle vascular index (CAVI) is a noninvasive measure of arterial stiffness that assesses arteriosclerosis independent of blood pressure at the time of measurement. Given that CAVI directly reflects arterial stiffness from the ascending aorta to the ankle arteries, it is considered a robust and reliable marker of vascular aging [3].
CAVI is calculated using the stiffness parameter β, which represents the logarithmic change in blood pressure required to increase arterial diameter. It was developed as an improved alternative to brachial‐ankle pulse wave velocity (baPWV) and is widely used as a marker of arterial stiffness. Unlike baPWV, which is influenced by blood pressure at the time of measurement, CAVI is designed to be less affected by blood pressure changes, thus providing a more accurate assessment of intrinsic arterial stiffness [4, 5]. This technical improvement allows CAVI to better reflect structural changes in the arterial wall, independent of acute hemodynamic conditions. By incorporating heart‐ankle pulse wave velocity, systolic blood pressure (SBP), diastolic blood pressure (DBP), and arterial diameter changes during the cardiac cycle, CAVI provides a comprehensive assessment of arterial stiffness from the aorta to the ankle [6].
Previous research extensively examined CAVI in clinical and high‐risk populations, demonstrating its association with cardiovascular events, such as myocardial infarction, stroke, and heart failure [7]. Similarly, in the prospective “Advanced Approach to Arterial Stiffness (TRIPLE‐A‐Stiffness) study,” Magnus et al. confirmed that elevated CAVI values were associated with cardiovascular morbidity and mortality. They further identified determinants of CAVI progression, including age, blood pressure, dyslipidemia, and glucose metabolism disorders [8]. However, these studies primarily focused on high‐risk patients, and the relationship between CAVI and overall or non‐CVD mortality in general populations has not been well established.
Recent studies suggest that arterial stiffness measured by CAVI may be related to various noncardiovascular conditions, potentially explaining its association with all‐cause mortality. Namekata et al. reported significantly higher CAVI values in patients with diabetes [5], whereas Gomez‐Marcos et al. found a relationship between CAVI and carotid intima‐media thickness in patients with Type 2 diabetes [9]. Nakamura et al. reported an association between the CAVI and serum cystatin C levels as a marker of kidney function [10]. Arterial stiffness has been linked to chronic inflammation [11], which is a risk factor for cancer development [12]. Therefore, CAVI may be a useful marker of systemic vascular health as it reflects multiple pathological processes that contribute to mortality beyond cardiovascular causes. While hazard ratios (HRs) provide valuable information regarding individual relative risk, they do not reflect the overall public health burden in a population, which also depends on the prevalence of the risk factor. Evaluating the population attributable fraction (PAF) is therefore crucial for clinicians and public health policymakers, as it quantifies the proportion of deaths that could potentially be prevented if elevated arterial stiffness (CAVI ≥ 9.0) were eliminated. However, to date, few prospective studies have quantified the PAF of elevated CAVI for both CVD and non‐CVD mortality in the general population.
Since 2005, a series of genome‐based cohort studies has been conducted in Kagoshima Prefecture in Amami Islands and Kagoshima mainland, Japan, as part of the Japan Multi‐Institutional Collaborative Cohort Study (J‐MICC Study) [13, 14]. This study aimed to assess associations between CAVI and all‐cause, CVD, and non‐CVD mortality in general populations.
2. Methods
2.1. Participants
This cohort study was conducted as part of the J‐MICC Study in Kagoshima Prefecture and included participants from the Amami Islands and mainland Kagoshima [14]. The baseline survey was conducted between 2005 and 2008 in one city and nine towns across five islands in the Amami region, and in 2012 in three cities in the Kagoshima mainland. After obtaining written informed consent, we recruited 7638 individuals aged 35–69 years from the general population. Participants were enrolled during routine health checkups conducted by the local government or private companies. The baseline survey comprised a self‐administered questionnaire; blood collection using serum, plasma, and buffy coat; urine collection; and examination of arterial stiffness using CAVI. Results of the routine health checkups were also obtained after acquiring informed consent.
Hypertension was defined as SBP ≥ 140 mmHg, DBP ≥ 90 mmHg, or intake of antihypertensive medication. Dyslipidemia was defined as TG level ≥ 150 mg/dL, LDL‐C level ≥ 140 mg/dL, or HDL‐C level < 40 mg/dL or the use of lipid‐lowering agents. Glucose intolerance was defined as FBG level ≥ 110 mg/dL or the use of antidiabetic medication.
Of the 7638 participants aged 35–69 years at baseline survey, 1664 participants were excluded due to missing CAVI data (n = 123) or difference between right and left CAVI values ≥ 1.0 (n = 74), ankle‐brachial index ≤ 0.9 (n = 72), extreme body mass index (BMI) values (BMI ≥ 50 or < 10 kg/m2) (n = 47), death within 1 year from baseline (n = 11), and previous history of ischemic heart disease, stroke, cancer or missing data on these diseases (n = 1317). Consequently, 5994 participants were included in the final analysis, including 3758 Amami island residents (1503 men and 2255 women) (Figure 1). Participants who moved out of study regions or who were deleted the authorization by a government official were censored. Mortality data were collected from regional public health offices and municipal records, ensuring a comprehensive follow‐up until December 31, 2021. The duration of follow‐up was calculated as the time from the date of the participant's baseline survey to their death, the censored, or December 31, 2021 according to the follow‐up procedures of each study region, whichever came first. The causes of death were confirmed using vital statistics data based on death certificates, with permission from the Ministry of Health, Labour and Welfare, Japan, and were coded according to the International Classification of Diseases and Related Health Problems, Tenth Revision (ICD‐10).
FIGURE 1.

A flow chart of the selection of study participants. Flow diagram of participant inclusion and exclusion in the present study. A total of 7638 eligible participants were screened, and 5994 were included in the final analysis after exclusions. BMI, body mass index; CAVI, cardio‐ankle vascular index.
Our study was approved by the Ethics Committee on Life Sciences and Genetic Analysis of Kagoshima University Graduate School of Medical and Dental Sciences (Approval number, #210178[G487]). Written informed consent was obtained from all participants prior to their inclusion in the study.
2.2. Data Collection and Measurements
2.2.1. Questionnaire
Data on lifestyle and medical history were collected using a structured questionnaire standardized by the J‐MICC Study [13, 14]. Data for the participants was collected using a structured questionnaire. Both self‐administered questionnaires were assessed by trained health professionals after interviewing the participants. Data common to both questionnaires were included in the analysis. Baseline assessments included demographic (age, sex, and residential area [Amami islands or Kagoshima mainland]), lifestyle (smoking and drinking habits [never, past, current]), and clinical data (BMI and medical history). Physical activity was assessed, and metabolic equivalents (METs) were calculated as previously described [15] and added to the analysis as a covariate. Physical activity was estimated as metabolic equivalent hours per day (MET・hours/day), based on the frequency and duration of daily and leisure‐time activities, where physical activity with MET values ≥ 3 were assessed. For daily physical activity, participants reported the average time spent per day in eight categories for each of the following daily activities: heavy physical work (assigned 4.5 METs) and walking (3.0 METs). Average time categories (assigned time in hours) were as follows: none (0), < 1 h/day (0.5), 1–< 3 h/day (2.0), 3–< 5 h/day (4.0), 5–< 7 h/day (6.0), 7–< 9 h/day (8.0), 9–< 11 h/day (10.0), and ≥ 11 h/day (11.0). Daily physical activity was estimated by multiplying the average time by the assigned MET value for each activity. For leisure‐time activity, participants reported the frequency and average duration for each activity: light‐intensity physical activity (e.g., walking and golf, assigned 3.4 METs as the median value) which does not make breathe harder than normal, moderate‐intensity physical activity which makes breathe somewhat harder than normal (e.g., jogging and swimming, 7.0 METs assigned as the median value), and vigorous‐physical activity which makes breathe much harder than normal (e.g., marathon running and martial arts, 10 METs assigned as the median value). The frequency and average duration were reported in five and six categories, respectively. The frequency categories (assigned daily average frequency) were as follows: none (0), 1–3 times/month (2/30), 1–2 times/week (1.5/7), 3–4 times/week (3.5/7), and ≥ 6 times/week (6/7). The average duration categories (assigned time in hours) were as follows: < 30 min (15/60), 30 min–< 1 h (45/60), 1–< 2 h (1.5), 2–< 3 h (2.5), 3–< 4 h (3.5), and ≥ 4 h (4.0). Leisure‐time physical activity was estimated by multiplying the frequency, average duration, and assigned MET value together for each activity. In the present study, total physical activity was then calculated by summing all daily and leisure‐time physical activities. The questionnaire was based on a similar one with good validity and reliability that was used in the Japan Public Health Center‐based prospective study [15].
2.2.2. Measurement of CAVI
CAVI was measured using a standardized method with Vasera VS‐1000 or VS‐1500 (Fukuda Denshi Co. Ltd., Tokyo, Japan), as described previously [16, 17, 18]. Briefly, cuffs were placed on both ankles and the upper arms, and electrocardiography electrodes were attached to both upper arms. In addition, a microphone was placed at the sternal angle for phonocardiography. CAVI was measured in the supine position. PWV was calculated by dividing the distance from the aortic valve to the ankle artery by the sum of the time intervals between the closing sound of the aortic valve and the notch of the brachial pulse wave and between the rise of the brachial pulse wave and the ankle pulse wave. To minimize the effect of cuff inflation on blood flow dynamics, pulse waves were measured with the cuff inflated to less than DBP (50 mmHg). Blood pressure was measured using an oscillometric device. SBP and DBP were measured from the right brachial artery. CAVI was calculated using the following equation:
where a and b are constants, and ρ is blood density. The mean CAVI values of right and left were used for the analysis.
Participants were classified into three CAVI groups as follows: CAVI 1 (CAVI < 8.0), CAVI 2 (8.0 ≤ CAVI < 9.0), and CAVI 3 (CAVI ≥ 9.0).
2.3. Statistical Analysis
Quantitative data are presented as mean ± standard deviation for data showing parametric distributions, whereas median and interquartile range were utilized for data showing nonparametric distributions. Differences in continuous variables were evaluated using the independent t‐test or the Mann–Whitney U test, depending on the normality of the data distribution. Differences in categorical variables were assessed using the chi‐square test.
Cox proportional hazards models were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between CAVI and mortality outcomes. Three sequential models with increasing levels of adjustment were constructed: Model 1 adjusted for age and sex; Model 2 further adjusted for BMI (by three categories, < 18.5, 18.5 to < 25.0, ≥ 25.0 kg/m2), smoking status (never, past, current), alcohol consumption (never, past, current), and residential area (Amami Islands vs. mainland Kagoshima); and Model 3 additionally adjusted for physical activity (METs) (by quartiles), comorbidities, including a history of diabetes mellitus, hyperlipidemia, and hypertension.
To quantify the population‐level impact, PAFs were calculated using Miettinen's formula: PAF = P(E|D) × (HR − 1)/HR, where P(E|D) represents the prevalence of CAVI ≥ 9.0 among deceased participants, and HR is the adjusted HR from the fully adjust–ed Cox proportional hazards model (Model 3).
Statistical analysis was performed using Stata, Version 18.0 for Windows (StataCorp LP, College Station, Tx). The forest plot was created using the ggplot2 package in R software (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
3.1. Baseline Characteristics by CAVI Categories
The study population (n = 5994) was categorized into three groups according to the CAVI values: CAVI 1 (CAVI < 8.0), CAVI 2 (8.0 ≤ CAVI < 9.0), and CAVI 3 (CAVI ≥ 9.0). Among the participants, 50.0% (n = 2995) were in the CAVI 1 group, 31.9% (n = 1910) in the CAVI 2 group, and 18.2% (n = 1089) in the CAVI 3 group. The mean CAVI values were 7.3 ± 0.6, 8.6 ± 0.3, and 9.7 ± 0.5 for the three groups, respectively (p < 0.001). Significant differences in demographic and clinical characteristics were observed across the CAVI groups. Mean age progressively increased with higher CAVI categories (53.7 ± 8.3 years; 60.0 ± 6.8 years; and 63.5 ± 4.9 years for CAVI 1, 2, and 3, respectively; p < 0.001). The proportion of males also increased with higher CAVI categories (33.9%; 43.0%; and 55.1% for CAVI 1, 2, and 3, respectively; p < 0.001). The proportion of participants from the Amami Islands decreased across the CAVI groups (CAVI 1, 71.7%; CAVI 2, 58.7%; CAVI 3, 45.8%; p < 0.001). Mean BMI showed a slight but significant decrease across the CAVI categories (CAVI 1, 24.2 ± 3.6 kg/m2; CAVI 2, 23.9 ± 3.2 kg/m2; CAVI 3, 23.6 ± 3.1 kg/m2; p < 0.001). Current drinking status was more prevalent in the highest CAVI category (CAVI 1, 51.7%; CAVI 2, 51.2%; CAVI 3, 57.9%; p = 0.001). Similarly, the proportion of former smokers increased with higher CAVI values (CAVI 1, 12.8%; CAVI 2, 17.4%; CAVI 3, 23.5%; p < 0.001), whereas the proportion of never‐smokers decreased (CAVI 1, 73.6%; CAVI 2, 69.8%; CAVI 3, 61.0%). A history of diabetes mellitus was observed in 3.0%, 6.4%, and 12.1% of participants in the CAVI 1, 2, and 3 groups, respectively (p < 0.001). Similarly, a history of hyperlipidemia (10.2%, 16.3%, and 19.9%; p < 0.001) and hypertension (16.5%, 29.0%, and 41.0%; p < 0.001) increased significantly across the CAVI groups, respectively. The number of deaths during the observation period was 100 (3.3%), 99 (5.2%), and 96 (8.8%) in the CAVI 1, 2, and 3 groups, respectively (p < 0.001). Mean observation periods were 13.2 ± 2.8, 12.4 ± 3.0, and 11.5 ± 3.0 years for the three CAVI groups, respectively (p < 0.001) (Table 1). The baseline characteristics of the study participants stratified by sex and region are summarized in Supplementary Table 1. Males exhibited less favorable lifestyle profiles compared to females, including significantly higher proportions of current smokers (27.9% vs. 4.0%, p < 0.001) and current drinkers (84.6% vs. 30.8%, p < 0.001), along with higher BMI and mean CAVI values (8.3 vs. 7.9, p < 0.001). The crude all‐cause mortality rate was correspondingly higher in males (7.1% vs. 3.5%, p < 0.001). Regarding regional differences, participants from Amami island were generally younger (mean age 55.1 vs. 61.6 years, p < 0.001) and had higher BMI and higher prevalences of current smoking and drinking compared to those from the Kagoshima mainland. Conversely, the Kagoshima mainland group had significantly higher mean CAVI values (8.4 vs. 7.9, p < 0.001), likely reflecting their older average age.
TABLE 1.
Baseline characteristics of participants by CAVI value.
| CAVI 1 (< 8.0) | CAVI 2 (≥ 8.0, < 9.0) | CAVI 3 (9.0 ≥) | p | |
|---|---|---|---|---|
| No. of participants, n (%) | 2995 (50.0%) | 1910 (31.9%) | 1089 (18.2%) | |
| CAVI, mean (SD) | 7.3 (0.6) | 8.6 (0.3) | 9.7 (0.5) | < 0.001 |
| Mean age (years), mean (SD) | 53.7 (8.3) | 60.0 (6.8) | 63.5 (4.9) | < 0.001 |
| Sex (male), n (%) | 1015 (33.9%) | 821 (43.0%) | 600 (55.1%) | < 0.001 |
| Region (Amami island), n (%) | 2146 (71.7%) | 1121 (58.7%) | 499 (45.8%) | < 0.001 |
| BMI (kg/m2), mean (SD) | 24.2 (3.6) | 23.9 (3.2) | 23.6 (3.1) | < 0.001 |
| Physical activity (METs‐hours/day), mean (SD) | 18.4 (14.3) | 19.2 (13.7) | 18.7 (13.0) | 0.13 |
| Alcohol drinking status, n (%) | ||||
| Never | 1402 (46.8%) | 899 (47.1%) | 435 (39.9%) | |
| Past | 46 (1.5%) | 34 (1.8%) | 24 (2.2%) | 0.001 |
| Current | 1547 (51.7%) | 977 (51.2%) | 630 (57.9%) | |
| Smoking status, n (%) | ||||
| Never | 2204 (73.6%) | 1331 (69.8%) | 664 (61.0%) | |
| Past | 383 (12.8%) | 332 (17.4%) | 256 (23.5%) | < 0.001 |
| Current | 408 (13.6%) | 245 (12.8%) | 169 (15.5%) | |
| History of DM, n (%) | 89 (3.0%) | 122 (6.4%) | 132 (12.1%) | < 0.001 |
| History of HL, n (%) | 303 (10.2%) | 311 (16.3%) | 217 (19.9%) | < 0.001 |
| History of HT, n (%) | 492 (16.5%) | 552 (29.0%) | 446 (41.0%) | < 0.001 |
| No. of death during observation, n (%) | 100 (3.3%) | 99 (5.2%) | 96 (8.8%) | < 0.001 |
| Observation periods (years), mean (SD) | 13.2 (2.8) | 12.4 (3.0) | 11.5 (3.0) | < 0.001 |
Note: p‐values were obtained by chi‐square test or one‐way analysis of variance*.
Abbreviations: BMI, body mass index; CAVI, cardio‐ankle vascular index; DM, diabetes mellitus; HL, hyperlipidemia; HT, hypertension; SD, standard deviation.
4. Mortality Outcomes
4.1. Association Between CAVI Values and Mortality
During the follow‐up period of 12.7 years (76 759 person‐years), 295 all‐cause deaths were recorded, including 46 CVD deaths, and 249 non‐CVD deaths were documented. For all‐cause mortality, individuals with CAVI ≥ 9.0 had a significantly higher risk than those with CAVI < 8.0, with an HR of 1.38 (95% CI: 1.00–1.90) in the fully adjusted model (Model 3). When comparing the impact of different risk factors in Model 3, diabetes mellitus showed the strongest association with all‐cause mortality (HR = 1.83, 95% CI: 1.27–2.62), followed by elevated CAVI values, while hypertension showed a nonsignificant association (HR = 1.24, 95% CI: 0.96–1.61) and hyperlipidemia showed no increased risk (HR = 0.78, 95% CI: 0.55–1.12). For CVD mortality, elevated CAVI showed a strong association, with individuals in the highest CAVI category having an HR of 2.17 (95% CI: 0.97–4.86) in Model 3. Comparing risk factors for CVD mortality, diabetes mellitus demonstrated the strongest association (HR = 2.37, 95% CI: 1.03–5.44), followed by elevated CAVI and hypertension (HR = 1.72, 95% CI: 0.92–3.22). For non‐CVD mortality, individuals with CAVI ≥ 9.0 had an HR of 1.26 (95% CI: 0.88–1.79) in Model 3. When comparing risk factors, diabetes mellitus showed the strongest association (HR = 1.73, 95% CI: 1.16–2.58), followed by elevated CAVI (Table 2, Figure 2). These findings demonstrate that elevated CAVI values are independently associated with increased risks of all‐cause mortality, and show positive trends for increased CVD and non‐CVD mortality. To confirm the robustness of our findings based on the mean CAVI values, we performed sensitivity analyses using the right CAVI and left CAVI independently. The results remained consistent with the primary analyses. In the fully adjusted model (Model 3), the highest category of right CAVI (≥ 9.0) was significantly associated with increased risks of all‐cause mortality (HR, 1.64; 95% CI, 1.19–2.26; p for trend = 0.003), CVD mortality (HR, 3.04; 95% CI, 1.34–6.90; p for trend = 0.051), and non‐CVD mortality (HR, 1.44; 95% CI, 1.01–2.05; p for trend = 0.213) compared to the reference group (< 8.0) (Supplementary Table 2A). Similar associations were observed when the left CAVI was utilized, with the highest category (≥ 9.0) showing significantly higher risks for all‐cause mortality (HR, 1.39; 95% CI, 1.01–1.92; p for trend = 0.043) and CVD mortality (HR, 2.31; 95% CI, 1.06–5.05; p for trend = 0.030) (Supplementary Table 2B). Kaplan–Meier survival analysis demonstrated that survival rates significantly worsened as the CAVI category increased. The group with the highest CAVI (CAVI 3: ≥ 9.0) exhibited a significantly higher incidence of all‐cause mortality compared to the lower CAVI groups (log‐rank test, χ 2 = 92.39, p < 0.001). Consistent with the findings for all‐cause mortality, cause‐specific analyses revealed that higher CAVI categories were significantly associated with increased risks of both CVD mortality (log‐rank test, χ 2 = 31.60, p < 0.001) and non‐CVD mortality (log‐rank test, χ 2 = 54.47, p < 0.001) (Supplementary Figure 1). These findings from the univariate survival analysis were consistent with those of the multivariate Cox proportional hazards models, confirming that elevated CAVI values were independently associated with an increased mortality risk, even after adjusting for potential confounders.
TABLE 2.
Hazard ratios and 95% CIs for all‐cause, CVD, and non‐CVD death according to CAVI value.
| No. of deaths | Person‐years | Hazard ratio (95% CI) | ||||||
|---|---|---|---|---|---|---|---|---|
| Model 1 | p for trend | Model 2 | p for trend | Model 3 | p for trend | |||
| All cause mortality | 0.012 | 0.022 | 0.055 | |||||
| CAVI 1 (< 8.0) | 100 | 39646.9 | Ref | Ref | Ref | |||
| CAVI 2 (≥ 8.0, < 9.0) | 99 | 23695.9 | 1.05 (0.78, 1.41) | 1.05 (0.78, 1.40) | 1.06 (0.79, 1.42) | |||
| CAVI 3 (9.0 ≥) | 96 | 12527.3 | 1.50 (1.10, 2.06) | 1.46 (1.06, 2.00) | 1.38 (1.00, 1.90) | |||
| DM | 1.78 (1.26, 2.53) | 1.76 (1.24, 2.51) | 1.83 (1.27, 2.62) | |||||
| HT | 1.23 (0.96, 1.57) | 1.25 (0.97, 1.61) | 1.24 (0.96, 1.61) | |||||
| HL | 0.92 (0.65, 1.29) | 0.89 (0.63, 1.26) | 0.78 (0.55, 1.12) | |||||
| CVD mortality | 0.022 | 0.026 | 0.051 | |||||
| CAVI 1 (< 8.0) | 13 | 39646.9 | Ref | Ref | Ref | |||
| CAVI 2 (≥ 8.0, < 9.0) | 13 | 23695.9 | 1.04 (0.46, 2.32) | 1.03 (0.46, 2.30) | 1.03 (0.46, 2.31) | |||
| CAVI 3 (9.0 ≥) | 20 | 12527.3 | 2.42 (1.10, 5.33) | 2.38 (1.07, 5.27) | 2.17 (0.97, 4.86) | |||
| DM | 2.05 (0.90, 4.66) | 1.99 (0.87, 4.55) | 2.37 (1.03, 5.44) | |||||
| HT | 1.53 (0.84, 2.79) | 1.56 (0.84, 2.91) | 1.72 (0.92, 3.22) | |||||
| HL | 0.26 (0.06, 1.08) | 0.24 (0.06, 1.00) | 0.18 (0.04, 0.78) | |||||
| Non‐CVD mortality | 0.082 | 0.126 | 0.213 | |||||
| CAVI 1 (< 8.0) | 87 | 39646.9 | Ref | Ref | Ref | |||
| CAVI 2 (≥ 8.0, < 9.0) | 86 | 23695.9 | 1.05 (0.77, 1.44) | 1.05 (0.77, 1.44) | 1.06 (0.77, 1.46) | |||
| CAVI 3 (9.0 ≥) | 76 | 12527.3 | 1.37 (0.97, 1.93) | 1.32 (0.93, 1.87) | 1.26 (0.88, 1.79) | |||
| DM | 1.73 (1.17, 2.55) | 1.71 (1.16, 2.53) | 1.73 (1.16, 2.58) | |||||
| HT | 1.17 (0.90, 1.54) | 1.19 (0.90, 1.58) | 1.16 (0.87, 1.54) | |||||
| HL | 1.06 (0.74, 1.51) | 1.04 (0.73, 1.49) | 0.94 (0.65, 1.36) | |||||
Note: Bold values indicate statistical significance.
Model 1: Adjusted for sex and age.
Model 2: BMI, smoking, alcohol drinking, and region were added to Model 1.
Model 3: Physical activity (METs), history of diabetes mellitus, hyperlipidemia, and hypertension were included in addition to the Model 2.
Abbreviations: CAVI, cardio‐ankle vascular index; CI, confidence interval; DM, diabetes mellitus; HL, hyperlipidemia; HT, hypertension; ref., reference.
FIGURE 2.

Forest plot of adjusted hazard ratios for all‐cause, cardiovascular, and noncardiovascular mortality according to Cardio‐Ankle Vascular Index (CAVI) categories. Hazard ratios (represented by points) and 95% confidence intervals (represented by horizontal bars) were estimated using multivariable Cox proportional hazards models. Participants were categorized into three groups based on their baseline CAVI values: CAVI1 (< 8.0; reference), CAVI 2 (8.0–8.9), and CAVI 3 (≥ 9.0). The vertical dotted line indicates the null effect (hazard ratio = 1.0). Models were adjusted for age, sex, body mass index, smoking status, alcohol consumption, residential area, physical activity (METs), comorbidities, including a history of diabetes mellitus, hyperlipidemia, and hypertension. CAVI, cardio‐ankle vascular index; CI, confidence interval; CVD, cardiovascular disease; HR, hazard ratio.
4.2. PAF of CAVI and Mortality Risk
We calculated the PAF to quantify the potential mortality reduction if high arterial stiffness (CAVI ≥ 9.0) was eliminated (Table 3). The prevalence of CAVI ≥ 9.0 among deceased individuals was 32.5% for all‐cause mortality, 43.5% for CVD mortality, and 30.5% for non‐CVD mortality. After adjustment, the PAFs were 8.9%, 23.4%, and 6.2% for all‐cause, CVD, and non‐CVD mortality, respectively. These findings indicate that approximately one‐fourth of cardiovascular deaths, as well as 8.9% of all‐cause and 6.2% of noncardiovascular deaths, can be attributed to high arterial stiffness. This highlights its broad significance as a potential target for preventive interventions beyond cardiovascular diseases.
TABLE 3.
Population attributable fraction of CAVI ≥ 9.0 for mortality outcomes.
| Adjusted HR (95% CI) | PAF (%) | |
|---|---|---|
| All cause death | 1.38 (1.00, 1.90) | 8.9% |
| CVD death | 2.17 (0.97, 4.86) | 23.4% |
| Non‐CVD death | 1.26 (0.88, 1.79) | 6.2% |
Note: The prevalence of CAVI ≥ 9.0 among cases, P(E|D), was 32.5% (96/295) for all‐cause death, 43.5% (20/46) for CVD death, and 30.5% (76/249) for non‐CVD death.
Abbreviations: CAVI, cardio‐ankle vascular index; CVD, cardiovascular disease.
5. Discussion
In this large‐scale, 12.7‐year prospective cohort study, we demonstrated that elevated CAVI was significantly associated with an increased risk of all‐cause, cardiovascular, and non‐CVD mortality in the general population. By quantifying these risks through PAFs, our study highlights that CAVI is not merely a cardiovascular risk indicator, but a comprehensive prognostic marker of systemic biological aging and whole‐body vascular health.
Our findings support and extend the prognostic significance of arterial stiffness. While several studies report an association between PWV and cardiovascular outcomes [19], CAVI offers distinct clinical advantages as it is less affected by blood pressure at the time of measurement [3]. Our results are consistent with previous reports linking elevated CAVI to mortality in hospital‐based cohorts [20] and cardiovascular events in patients with coronary artery disease [21]. Importantly, our study extends these findings to a healthy, community‐dwelling population. The robust dose–response relationship, with the highest CAVI category (≥ 9.0) consistently demonstrating the strongest mortality associations, reinforces its utility for early risk stratification in primary care.
Severe arterial stiffness (CAVI ≥ 9.0) was identified as a more potent predictor of CVD mortality (HR = 2.17) than established traditional risk factors, including diabetes mellitus (HR = 2.37) and hypertension (HR = 1.72). This finding suggests that CAVI captures the cumulative, end‐organ pathophysiological burden of various risk factors and aging on the arterial wall, which cannot be fully quantified by assessing traditional risk factors in isolation. Our PAF analysis translates this individual risk into public health impact: an estimated 23.4% of cardiovascular deaths and 8.9% of all‐cause deaths are attributable to elevated arterial stiffness. These results suggest that interventions targeting vascular aging could yield substantial survival benefits at the population level.
The mechanisms linking elevated CAVI to increased mortality are multifactorial. From a hemodynamic perspective, arterial stiffness causes left ventricular hypertrophy, impaired coronary perfusion, and increased cardiac afterload, precipitating myocardial ischemia and heart failure [22]. Furthermore, increased pulsatile stress damages the microcirculation of highly vascularized organs, particularly the brain and kidneys, promoting progressive organ dysfunction [23]. Crucially, the significant PAF (6.2%) observed for non‐CVD mortality suggests that the predictive value of CAVI extends far beyond cardiovascular risk to include systemic vascular alterations linked to multiple pathological processes. For instance, elevated CAVI correlates with target organ damage in Type 2 diabetes [9] and markers of renal dysfunction, such as serum cystatin C [10]. These relationships suggest shared pathophysiological pathways, wherein systemic microvascular damage drives multiorgan deterioration. Furthermore, arterial stiffness is increasingly being recognized as a clinical manifestation of chronic low‐grade inflammation [24] and biological aging. Orkaby et al. studied 2171 community‐dwelling older adults and found that prefrail and frail individuals had progressively higher carotid‐femoral pulse wave velocity compared with non‐frail participants (10.3 and 10.5 m/s vs. 10.0 m/s; p = 0.0002) [25]. A meta‐analysis pooling five studies demonstrated a standardized mean difference of 0.62 (p < 0.0001) in aortic stiffness between frail and non‐frail older persons [26]. In another study examining 5392 participants aged 55–78 years, arterial stiffness strongly correlated with age and physical function limitations, independent of blood pressure [27]. Further, arterial stiffness was identified as an emerging tissue biomarker for estimating individual biological age [28]. Chronic inflammatory states are fundamental to the progression of numerous noncommunicable diseases, including malignancies [12]. Large prospective studies provide strong evidence. In a study by Jiang et al. comprising 45 627 participants with 553 cancer cases, those in the highest arterial stiffness group (measured by baPWV) had a 51% increased cancer risk (HR = 1.51, 95% CI = 1.14–2.00) [29]. A study of patients with cancer undergoing coronary interventions identified malignancy as an independent predictor of cardiovascular events [30]. This shared inflammatory milieu offers a compelling mechanistic explanation for the observed association between elevated CAVI and non‐CVD/cancer‐related mortality. Collectively, CAVI functions as a composite biomarker, integrating the cumulative damage from metabolic dysregulation, renal impairment, and systemic inflammation.
From a clinical perspective, our findings have important implications for the practical application of CAVI. Given its non‐invasive, highly reproducible, and operator‐independent nature, CAVI is an ideal tool for large‐scale screening and can be easily incorporated into routine community health checkups. Furthermore, our study strongly supports the practical utility of a specific cutoff value. Using a threshold of CAVI ≥ 9.0 proved to be robust for identifying individuals at a significantly elevated risk of all‐cause, CVD, and non‐CVD mortality. As suggested by our PAF estimations, early identification of these high‐risk individuals (CAVI ≥ 9.0) during health screenings, followed by intensive and targeted management of their lifestyle and modifiable cardiovascular risk factors, could lead to a substantial reduction in mortality at the population level.
Our baseline data revealed an inverse association between higher CAVI and lower BMI. This observation aligns with important epidemiological findings reported in previous studies. For instance, Nagayama et al. demonstrated a negative correlation between CAVI and BMI, suggesting a phenomenon akin to the obesity paradox in the context of arterial stiffness [31]. Similarly, Kabutoya et al. highlighted the complex interplay between arterial stiffness, lower BMI, and adverse outcomes, particularly emphasizing that a lower BMI in older adults often serves as a surrogate marker for frailty and sarcopenia [32]. In our cohort, participants in the highest CAVI category (≥ 9.0) might encompass a larger proportion of frail or sarcopenic individuals, as indicated by their significantly lower BMI and older age. Because both physical frailty and advanced arterial stiffness are robust predictors of mortality, their coexistence could synergistically exacerbate the risk of death. Further studies incorporating direct assessments of frailty, such as grip strength or gait speed, are warranted to elucidate the precise mechanisms underlying the triad of high CAVI, low BMI, and increased mortality.
In the fully adjusted model (Model 3), the hazard ratios for CVD and non‐CVD mortality associated with elevated CAVI (≥ 9.0) were attenuated and did not reach statistical significance. This attenuation warrants epidemiological and methodological consideration. First, we employed the average of bilateral CAVI measurements rather than unilateral measurements. While unilateral measurements can be susceptible to localized subclinical arterial disease or measurement artifacts, utilizing the bilateral average provides a more stable and representative assessment of systemic arterial stiffness. This averaging process likely reduced individual measurement variations and smoothed out extreme values, resulting in more conservative and stable hazard ratio estimates. Second, the addition of physical activity as an adjustment variable in Model 3 contributed to the attenuation of CAVI's predictive value. Although baseline physical activity did not differ significantly among the three CAVI categories (p = 0.13), physical activity is known to influence mortality through biological pathways that overlap with those of arterial stiffness, such as systemic inflammation, endothelial function, and metabolic regulation. In a highly saturated multivariable survival model, adjusting for such lifestyle factors can lead to the sharing of variance among covariates. Consequently, this comprehensive adjustment controlled for residual lifestyle confounding but simultaneously led to a conservative estimation of the independent effect of CAVI, resulting in marginal statistical significance for cause‐specific mortality.
The strengths of this study include its long‐term prospective design, large community‐based sample, and the calculation of PAFs for cause‐specific mortality. A notable strength of our study is the confirmation of the prognostic utility of CAVI regardless of the measurement side. Our sensitivity analyses demonstrated that employing either the right or left CAVI independently yielded consistent results with the mean CAVI, showing significant associations with all‐cause and CVD mortality. In routine clinical practice, calculating the mean value of bilateral measurements might be cumbersome, or obtaining bilateral data might be challenging due to unilateral peripheral artery disease or measurement failure. Our findings suggest that a unilateral CAVI measurement (either right or left) is sufficiently reliable for risk stratification in general populations, thereby enhancing the practical applicability of CAVI in busy clinical settings. Another strength of this study is the inclusion of participants from geographically and demographically distinct areas, specifically the Kagoshima mainland and the remote Amami islands. As shown in our supplementary analysis, there was substantial heterogeneity in baseline characteristics across regions and sexes; males and island residents tended to have higher prevalences of lifestyle‐related risk factors such as smoking, drinking, and obesity, whereas mainland residents were older with correspondingly higher baseline arterial stiffness. Despite these varied sociodemographic and clinical backgrounds, adjusting for these factors—including sex and region—in multivariable Cox proportional hazards models did not attenuate the prognostic significance of CAVI. This highlights the robustness and broad generalizability of CAVI as an independent predictor of mortality in the general population, regardless of regional lifestyle disparities or gender differences. However, some limitations exist. CAVI was measured only at baseline, precluding the analysis of longitudinal vascular changes. Although we adjusted for physical activity in our multivariate models, detailed nutritional and dietary intake data were not included in the analysis. Given that dietary habits significantly influence cardiovascular health and overall mortality, the lack of dietary factors represents a limitation, particularly in the estimation of PAFs where residual confounding from lifestyle factors cannot be entirely ruled out. Residual confounding from unmeasured variables cannot be entirely ruled out. Finally, as the study was conducted in a Japanese cohort, the PAF values should be validated in other multi‐ethnic populations to confirm global generalizability.
6. Conclusion
Elevated CAVI is a valuable, independent predictor of all‐cause mortality in the general population. Although its independent associations with cardiovascular (CVD) and non‐CVD mortality did not reach statistical significance after adjustment for comprehensive clinical and lifestyle factors, elevated CAVI consistently displayed a strong prognostic trend. By capturing the cumulative burden of systemic inflammation, metabolic dysregulation, and aging, CAVI serves as a highly valuable tool for comprehensive health assessment, outperforming some traditional risk factors. The substantial PAFs associated with elevated CAVI highlight the urgent need to integrate arterial stiffness measurements into primary care and public health strategies. Future translational research should explore whether interventions directly targeting systemic vascular aging can mitigate the mortality risks identified in this study.
Author Contributions
S.T., I.A.Z., C.K., and T.T. conceived the general structure of the study. S.T., I.A.Z., and A.N. analyzed data, wrote the first draft of the manuscript and drafted its revised versions. All authors critically revised the different versions of the manuscript and approved its final version. S.T. and C.K. are the guarantors of this work and, as such, have full access to all data in the study and take responsibility for the integrity of data and the accuracy of data analysis.
Funding
This study was supported by Grants‐in‐Aid for Scientific Research for Priority Areas of Cancer (No. 17015018) and Innovative Areas (No. 221S0001) and by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant (No. 16H06277 and 22H04923 [CoBiA]) from the Japanese Ministry of Education, Culture, Sports, Science and Technology.
Ethics Statement
This study was approved by the Ethics Committee on Life Sciences and Genetic Analysis of Kagoshima University Graduate School of Medical and Dental Sciences (Approval number, #210178[G487]).
Consent
Written informed consent was obtained from all participants prior to their inclusion in the study.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Supplementary Figure 1. Kaplan–Meier plots of survival rates for all‐cause, cardiovascular disease (CVD), and non‐CVD mortality by CAVI value. Kaplan–Meier survival analysis demonstrated significant differences in survival rates across the three CAVI categories: CAVI 1 (< 8.0), CAVI 2 (≥ 8.0, < 9.0), and CAVI 3 (≥ 9.0). Log‐rank tests revealed highly significant differences in survival distributions among the three CAVI groups for all‐cause mortality (χ 2 = 92.39, p < 0.001), CVD mortality (χ 2 = 31.60, p < 0.001), and non‐CVD mortality (χ 2 = 54.47, p < 0.001).
Supplementary Table 1. Baseline characteristics by sex and region.
Supplementary Table 2A. Hazard ratios and their 95% CIs for all‐cause, CVD and non‐CVD death according to right_CAVI value.
Supplementary Table 2B. Hazard ratios and their 95% CIs for all‐cause, CVD and non‐CVD death according to left_CAVI value.
Acknowledgments
The authors would like to thank all patients and investigators who participated in this study. We would like to thank Editage (www.editage.jp) for English language editing. During the preparation of this work, the author used Google Gemini to improve language readability and assist with formatting R code for data visualization. Additionally, the Graphical Abstract was created with the assistance of NotebookLM (Google LLC). After using these tools, the author reviewed and edited the content as needed and took full responsibility for the content of the publication.
Data Availability Statement
The datasets generated and analyzed during the current study are not publicly available due to ethical restrictions and the protection of participant privacy. However, the data are available from the corresponding author upon reasonable request and subject to approval by the relevant ethics committee.
References
- 1. Vlachopoulos C., Aznaouridis K., and Stefanadis C., “Prediction of Cardiovascular Events and All‐Cause Mortality With Arterial Stiffness: A Systematic Review and Meta‐Analysis,” Journal of the American College of Cardiology 55, no. 13 (2010): 1318–1327. [DOI] [PubMed] [Google Scholar]
- 2. Mitchell G. F., Hwang S. J., Vasan R. S., et al., “Arterial Stiffness and Cardiovascular Events: The Framingham Heart Study,” Circulation 121, no. 4 (2010): 505–511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Shirai K., Utino J., Otsuka K., and Takata M., “A Novel Blood Pressure‐Independent Arterial Wall Stiffness Parameter; Cardio‐Ankle Vascular Index (CAVI),” Journal of Atherosclerosis and Thrombosis 13, no. 2 (2006): 101–107. [DOI] [PubMed] [Google Scholar]
- 4. Shirai K., Hiruta N., Song M., et al., “Cardio‐Ankle Vascular Index (CAVI) as a Novel Indicator of Arterial Stiffness: Theory, Evidence and Perspectives,” Journal of Atherosclerosis and Thrombosis 18, no. 11 (2011): 924–938. [DOI] [PubMed] [Google Scholar]
- 5. Namekata T., Suzuki K., Ishizuka N., and Shirai K., “Establishing Baseline Criteria of Cardio‐Ankle Vascular Index as a New Indicator of Arteriosclerosis: A Cross‐Sectional Study,” BMC Cardiovascular Disorders 10 (2011): 11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Hayashi K., Yamamoto T., Takahara A., and Shirai K., “Clinical Assessment of Arterial Stiffnesswith Cardio‐Ankle Vascular Index: Theory and Applications,” Journal of Hypertension 33, no. 9 (2015): 1742–1757. [DOI] [PubMed] [Google Scholar]
- 7. Miyoshi T., Ito H., Shirai K., et al., “Predictive Value of the Cardio‐Ankle Vascular Index for Cardiovascular Events in Patients at Cardiovascular Risk,” Journal of the American Heart Association 10, no. 16 (2021): e020103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Bäck M., Topouchian J., Labat C., et al., “Cardio‐Ankle Vascular Index for Predicting Cardiovascular Morbimortality and Determinants for Its Progression in the Prospective Advanced Approach to Arterial Stiffness (TRIPLE‐A‐Stiffness) Study,” eBioMedicine 103, no. April (2024): 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gómez‐Marcos M. T., Recio‐Rodríguez J. I., Patino‐Alonso M. C., et al., “Relationship Between Intima‐Media Thickness of the Common Carotid Artery and Arterial Stiffness in Subjects With and Without Type 2 Diabetes: A Case‐Series Report,” Cardiovascular Diabetology 12 (2011): 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Nakamura K., Iizuka T., Takahashi M., et al., “Association Between Cardio‐Ankle Vascular Index and Serum Cystatin C Levels in Patients With Cardiovascular Risk Factor,” Journal of Atherosclerosis and Thrombosis 16, no. 4 (2009): 371–379. [DOI] [PubMed] [Google Scholar]
- 11. Cagnina A., Chabot O., Davin L., Lempereur M., Maréchal P., and Oury C., “Atherosclerosis, an Inflammatory Disease,” New England Journal of Medicine 340, no. 2 (1999): 115–126. [Google Scholar]
- 12. Coussens L. M. and Werb Z., “Inflammation and Cancer,” Nature 420, no. 6917 (2002): 860–867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. “The Japan Multi‐Institutional Collaborative Cohort Study (J‐MICC Study) to Detect Gene‐Environment Interactions for Cancer The J‐MICC Study Group*,” http://www.biobankjp.org/index.html. [PubMed]
- 14. Takeuchi K., Naito M., Kawai S., et al., “Study Profile of the Japan Multi‐Institutional Collaborative Cohort (J‐Micc) Study,” Journal of Epidemiology 31, no. 12 (2021): 660–668. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Kikuchi H., Inoue S., Odagiri Y., et al., “Intensity‐Specific Validity and Reliability of the Japan Public Health Center‐Based Prospective Study‐Physical Activity Questionnaire,” Preventive Medicine Reports 20 (2020): 1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Kubozono T., Miyata M., Ueyama K., et al., “Clinical Significance and Reproducibility of New Arterial Distensibility Index,” Circulation Journal 71, no. 1 (2007): 89–94. [DOI] [PubMed] [Google Scholar]
- 17. Nakamura K., Tomaru T., Yamamura S., Miyashita Y., Shirai K., and Noike H., “Cardio‐Ankle Vascular Index Is a Candidate Predictor of Coronary Atherosclerosis,” Circulation Journal 72, no. 4 (2007): 598–604. [DOI] [PubMed] [Google Scholar]
- 18. Ueyama K., Miyata M., Kubozono T., et al., “Noninvasive Indices of Arterial Stiffness in Hemodialysis Patients,” Hypertension Research 32, no. 8 (2009): 716–720. [DOI] [PubMed] [Google Scholar]
- 19. Ben‐Shlomo Y., Spears M., Boustred C., et al., “Aortic Pulse Wave Velocity Improves Cardiovascular Event Prediction: An Individual Participant Meta‐Analysis of Prospective Observational Data From 17,635 Subjects,” Journal of the American College of Cardiology 63, no. 7 (2014): 636–646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Kubota Y., Maebuchi D., Takei M., et al., “Cardio‐Ankle Vascular Index Is a Predictor of Cardiovascular Events,” Artery Research 5, no. 3 (2011): 91–96. [Google Scholar]
- 21. Sato Y., Nagayama D., Saiki A., et al., “Cardio‐Ankle Vascular Index Is Independently Associated With Future Cardiovascular Events in Outpatients With Metabolic Disorders,” Journal of Atherosclerosis and Thrombosis 23, no. 5 (2016): 596–605. [DOI] [PubMed] [Google Scholar]
- 22. Mitchell G. F., “Arterial Stiffness and Hypertension: Chicken or Egg?,” Hypertension 64, no. 2 (2014): 210–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. O'Rourke M. F. and Safar M. E., “Relationship Between Aortic Stiffening and Microvascular Disease in Brain and Kidney: Cause and Logic of Therapy,” Hypertension 46, no. 1 (2005): 200–204. [DOI] [PubMed] [Google Scholar]
- 24. Ussell R. and Oss R., “Atherosclerosis‐A Inflammatory Disease,” New England Journal of Medicine 340 (1999): 115–126. [DOI] [PubMed] [Google Scholar]
- 25. Orkaby A. R., Lunetta K. L., Sun F. J., et al., “Cross‐Sectional Association of Frailty and Arterial Stiffness in Community‐Dwelling Older Adults: The Framingham Heart Study,” Journals of Gerontology. Series A, Biological Sciences and Medical Sciences 74, no. 3 (2018): 373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Piotrowicz K., Gryglewska B., Grodzicki T., and Gąsowski J., “Arterial Stiffness and Frailty ‐ A Systematic Review and Metaanalysis,” Experimental Gerontology 1 (2021): 153. [DOI] [PubMed] [Google Scholar]
- 27. Brunner E. J., Shipley M. J., Witte D. R., et al., “Arterial Stiffness, Physical Function, and Functional Limitation: The Whitehall II Study,” Hypertension 57, no. 5 (2011): 1003–1009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Palombo C. and Kozakova M., “Arterial Stiffness, Atherosclerosis and Cardiovascular Risk: Pathophysiologic Mechanisms and Emerging Clinical Indications,” Vascular Pharmacology 1, no. 77 (2016): 1–7. [DOI] [PubMed] [Google Scholar]
- 29. Jiang Y., Xing A., Hidru T. H., et al., “The Association Between Arterial Stiffness and Cancer Occurrence: Data From Kailuan Cohort Study,” Frontiers in Cardiovascular Medicine 10 (2023): 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Tabata N., Sueta D., Yamamoto E., et al., “A Retrospective Study of Arterial Stiffness and Subsequent Clinical Outcomes in Cancer Patients Undergoing Percutaneous Coronary Intervention,” Journal of Hypertension 37, no. 4 (2019): 754–764. [DOI] [PubMed] [Google Scholar]
- 31. Nagayama D., Imamura H., Sato Y., et al., “Inverse Relationship of Cardioankle Vascular Index With BMI in Healthy Japanese Subjects: A Cross‐Sectional Study,” Vascular Health and Risk Management 1, no. 13 (2016): 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Kabutoya T., Hoshide S., and Kario K., “Asian Management of Hypertension: Current Status, Home Blood Pressure, and Specific Concerns in Japan,” Journal of Clinical Hypertension 22, no. 3 (2020): 486–492. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure 1. Kaplan–Meier plots of survival rates for all‐cause, cardiovascular disease (CVD), and non‐CVD mortality by CAVI value. Kaplan–Meier survival analysis demonstrated significant differences in survival rates across the three CAVI categories: CAVI 1 (< 8.0), CAVI 2 (≥ 8.0, < 9.0), and CAVI 3 (≥ 9.0). Log‐rank tests revealed highly significant differences in survival distributions among the three CAVI groups for all‐cause mortality (χ 2 = 92.39, p < 0.001), CVD mortality (χ 2 = 31.60, p < 0.001), and non‐CVD mortality (χ 2 = 54.47, p < 0.001).
Supplementary Table 1. Baseline characteristics by sex and region.
Supplementary Table 2A. Hazard ratios and their 95% CIs for all‐cause, CVD and non‐CVD death according to right_CAVI value.
Supplementary Table 2B. Hazard ratios and their 95% CIs for all‐cause, CVD and non‐CVD death according to left_CAVI value.
Data Availability Statement
The datasets generated and analyzed during the current study are not publicly available due to ethical restrictions and the protection of participant privacy. However, the data are available from the corresponding author upon reasonable request and subject to approval by the relevant ethics committee.
