Estimated Pulse Wave Velocity for Cardiovascular Risk Assessment in Rural Populations: The ENAH Study Authors: Marija Domislović, Ana Jelaković, Vladimir Prelević, Danilo Radunović, Lana Gellineo, Krešimir Đapić, Mirjana Fuček, Josipa Josipović, Jelena Kos, Viktor Domislović, Ivan Pećin, Mirta Abramović Barić, Krunoslav Capak, Bojan Jelaković Published: 2026 ABSTRACT AbstractIntroductionArterial stiffness, typically measured by pulse wave velocity (PWV), is a recognized biomarker of cardiovascular (CV) risk and mortality. Our study assesses the predictive value of arterial stiffness measured by estimated pulse wave velocity (ePWV) for fatal and nonfatal outcomes in a general rural adult population.MethodsThis prospective observational cohort study was conducted within the ENAH project, which initially included 3,305 adults, with 2,232 participants followed up over a median of 11.3 years. Baseline demographic, clinical, and laboratory data, including blood pressure, anthropometry, and biochemical analyses of blood and urine, were collected using standardized protocols. The primary outcome was all-cause mortality, while secondary outcomes comprised nonfatal CV, cerebrovascular, and renal events, analyzed using Kaplan-Meier survival curves and multivariable regression models.ResultsIn this cohort of 1,175 adults followed for a mean of 11.3 years, 263 participants died (22.4%), and 171 (14.5%) experienced a nonfatal event (myocardial infarction, atrial fibrillation, heart failure, stroke, transient ischemic attack, or dialysis). Kaplan-Meier analysis showed lower survival in participants with ePWV of >10 m/s compared to those with ePWV of <10 m/s (hazard ratio [HR] = 12.8, 95% CI 9.9–16.6, p < 0.001). Similarly, event-free survival was lower for nonfatal outcomes in the higher ePWV group (HR = 1.50, 95% CI 1.0–2.1, p = 0.008). In multivariable analysis for mortality, male sex and ePWV remained significant predictors (HR 1.75, 95% CI 1.18–2.61, p = 0.006; HR 1.81, 95% CI 1.08–3.04, p = 0.030). No independent predictors were found for nonfatal outcomes. However, for the composite endpoint, both male sex (HR = 1.54, 95% CI 1.14–2.08, p = 0.006) and ePWV (HR = 1.94, 95% CI 1.50–2.51, p < 0.001) were associated with increased risk.ConclusionePWV was an independent predictor of overall mortality as well as composite fatal and nonfatal outcomes. These findings suggest that this simple measure of arterial stiffness may be useful in clinical practice, particularly in rural settings where direct measurements are not available. Introduction Arterial stiffness, typically measured by pulse wave velocity (PWV), is a recognized biomarker of cardiovascular (CV) risk and mortality. Direct measurement of the carotid-femoral PWV (cfPWV) is the gold standard for evaluating arterial stiffness due to its strong predictive value for CV outcomes; however, cfPWV measurement requires specialized equipment and technical expertise, which limits its practicality, especially in rural or resource-limited areas [1, 2]. To overcome these limitations, estimated PWV (ePWV) has been developed and validated as an alternative, estimating arterial stiffness using age and mean arterial pressure (MAP). Studies indicate that ePWV provides a comparable prognostic value for all-cause mortality, CV mortality, and composite CV outcomes and has demonstrated utility in populations with varying CV risk profiles, including those with hypertension and chronic kidney disease (CKD) [3–8]. Recently, a large systematic review and meta-analysis including over 380,000 participants confirmed that elevated ePWV is strongly associated with increased risk of CV events, CV mortality, and all-cause mortality, further supporting its prognostic value in diverse populations [9]. In rural populations, where access to advanced diagnostic equipment is limited, the ability to estimate arterial stiffness using ePWV represents a significant advancement in CV risk assessment. Studies have demonstrated that ePWV retains independent predictive value for both CV and all-cause mortality, even when adjusted for traditional risk scores such as SCORE and the Framingham Risk Score [1–4, 7–11]. This is particularly relevant in identifying individuals at elevated risk who might otherwise be misclassified by traditional models alone. Furthermore, ePWV’s role in predicting composite outcomes, including fatal and nonfatal events such as myocardial infarction and stroke, underscores its potential for guiding preventive strategies and treatment in underserved populations, such as the rural population [6, 12]. This study investigates the predictive value of ePWV for all-cause mortality and fatal and nonfatal CV outcomes in a rural population, where direct measurement of PWV is not feasible. By leveraging a cohort of individuals from the ENAH study, we aim to explain the role of ePWV as an independent predictor of adverse events. Furthermore, we explore its utility in risk stratification for composite outcomes, contributing to a growing body of evidence on its clinical relevance, especially in rural settings with limited access to advanced diagnostic tools [13]. Methods Study Design and Population This study is part of the scientific project Endemic Nephropathy in Croatia – Epidemiology, Diagnostics, and Etiopathogenesis (ENAH – Ministry of Science, Education, and Sports 108-0000000329) [11]. ENAH project is a prospective, observational cohort study which initially included 3,305 adult participants, of which 2,232 were followed up (participation rate 67.5%). Baseline data were collected during field examination conducted in 2005, 2008, 2010, and 2015. The goal was to include as many participants as possible from the target population, using a “door-to-door” strategy to collect epidemiological data as well as blood and urine samples for further laboratory analysis. The participants were followed up for an average of 136.0 months (100.0–163.8), min-max (0.5–201.0), over the period from 2005 to 2022. The analysis of follow-up was performed on 1,175 participants from the ENAH project database who had baseline values of ePWV. Figure 1 in the online supplementary materials (for all online suppl. material, see https://doi.org/10.1159/000552567) shows the analysis process and the formation of participant groups included in this study. To assess the potential impact of loss to follow-up, a sensitivity analysis comparing baseline characteristics of participants included in the analysis and those lost to follow-up was also performed (online suppl. Table S9). Data Collection and Follow-Up Baseline data were collected through questionnaire capturing demographic, clinical, and lifestyle information, including personal and family medical history. Participants underwent clinical examinations, where blood pressure, anthropometric measurements, and biochemical analyses of blood and morning spot urine samples were obtained. Blood pressure was measured using an automated oscillometric device (Omron M6) on the nondominant arm. Measurements were performed in the seated position after at least 5 min of rest, using an appropriately sized cuff, in accordance with the European Society of Hypertension and European Society of Cardiology (ESH/ESC) recommendations. Three consecutive measurements were obtained, and the average of the second and third readings was used for the analyses. Hypertension was defined as systolic blood pressure (SBP) of ≥140 mm Hg and/or diastolic blood pressure (DBP) of ≥90 mm Hg, or current use of antihypertensive medication. Participants were advised to avoid smoking, caffeine intake, and physical activity for at least 30 min prior to measurement. Waist circumference was measured using a standardized protocol. Visceral obesity was defined according to waist circumference thresholds of ≥102 cm in men and ≥88 cm in women, consistent with commonly used cardiometabolic risk criteria. After completing the questionnaire and undergoing measurements, fasting blood samples and first and second morning urine samples were taken from the participants. Samples were labelled with a unique identification code to eliminate laboratory and research bias. Laboratory processing was conducted to determine all traditional CV risk factors, also albumin-to-creatinine ratio was measured from the morning urine sample. All analyses were done in one laboratory located in the University Hospital Center Zagreb. The validated ePWV equation used in the study was ePWV = 9.587 − 0.402 × age + 4.560 × 10−3 × age2 − 2.621 × 10−5 × age2 × MAP + 3.176 × 10−3 × age × MAP − 1.832 × 10−2 × MAP [4]. Mean arterial pressure was calculated as DBP + 0.4 (SBP − DBP). Cardiovascular risk scores were calculated using standard algorithms. The Framingham Risk Score was calculated based on age, sex, systolic blood pressure, lipid levels, smoking status, and diabetes, while the Heart SCORE was calculated using age, sex, systolic blood pressure, smoking status, and total cholesterol [10, 11]. The equation developed for individuals with CV risk factors was used, as the study population had a relatively high prevalence of cardiometabolic risk factors, including hypertension, diabetes, and obesity [4]. Data on both fatal and nonfatal events were collected from the Croatian Institute of Public Health records and general practitioners’ archives. Fatal outcomes were recorded according to International Classification of Diseases, Tenth Revision (ICD-10) codes, while nonfatal outcomes included myocardial infarction, atrial fibrillation, heart failure, stroke, transient ischemic attack, and initiation of chronic haemodialysis. Outcome Measures and Statistical Analysis The primary outcome of this study was all-cause mortality, while the secondary outcome included a composite of nonfatal CV, cerebrovascular, and renal events. Nonfatal outcomes were defined as myocardial infarction, atrial fibrillation, heart failure, stroke, transient ischemic attack, and initiation of chronic hemodialysis. Cardiovascular mortality was classified based on the International Classification of Diseases, Tenth Revision (ICD-10), using codes I00–I99 to identify deaths attributed to diseases of the CV system. Survival analysis was performed using Kaplan-Meier (KM) curves, with hazard ratios (HRs) calculated to assess associations between ePWV and both fatal and nonfatal events. A stepwise Cox proportional hazards regression analysis was conducted to evaluate predictors of the outcomes. Initially, univariate analysis identified variables significantly associated with the outcome of interest (p < 0.05). Variables identified in the univariate analysis, as well as those known to influence the outcome based on prior evidence, were included in subsequent multivariate regression models. To address potential multicollinearity, variables with high intercorrelation (e.g., body mass index [BMI] and weight, SBP and MAP, estimated glomerular filtration rate [eGFR] and CKD) were either excluded or analyzed in separate models. Four models were constructed: model 1 included the primary variables of interest; model 2 additionally incorporated the Framingham Risk Score and SCORE; model 3 substituted SBP with MAP; and model 4 replaced eGFR with CKD. These models were designed to comprehensively evaluate the predictors while minimizing confounding effects and collinearity. Data were presented as percentages, while continuous variables were expressed as mean ± standard deviation or as median with interquartile range. Normality was assessed using the D’Agostino-Pearson test and by visual inspection. Categorical variables were compared using the χ2 test or Fisher’s exact test when appropriate. For continuous variables, Student’s t test (independent or paired) was used for normally distributed data, and the Mann-Whitney U or Wilcoxon test for non-normal distributions. Comparisons across more than two groups were performed using ANOVA with Tukey-Kramer post hoc test for normally distributed variables, or the Kruskal-Wallis test with post hoc Mann-Whitney U test for non-normally distributed variables. Statistical analyses were performed using IBM SPSS version 23 and MedCalc version 23.0.6. Generative AI was not used in the preparation of this manuscript. Results A total of 2,232 participants were enrolled in this study, with follow-up data collected from 2005 to 2022. Participants were followed for an average of 11.3 years, with a range of 8.3–13.65 years. Of the 1,175 participants with known survival status, 912 were alive at the end of the follow-up period, while 263 were deceased. Figure 1 in online supplementary materials shows the study flowchart. The causes of death categorized by ICD-10 and stratified by ePWV and gender are presented in online supplementary Tables S7 and S8in the supplementary materials. Individuals with ePWV of >10 m/s had a significantly higher proportion of deaths attributed to diseases of the CV system (47.9%) compared to those with ePWV of <10 m/s (18.9%). On the contrary, deaths due to cancers were more prevalent in the ePWV of <10 m/s group (35.1%) compared to the >10 m/s group (22.1%). The overall p value for differences in causes of death between the two groups was 0.029. Females had a significantly higher proportion of deaths from diseases of the CV system (55.1%) compared to males (29.3%). In contrast, cancers accounted for a higher percentage of deaths among males (33.6%) than females (16.2%). Unknown causes of death were more frequent among males (11.2%) than females (6.8%). The overall p value for gender differences in causes of death was 0.009. The characteristics of participants divided by sex are described in Table 1. This study included a total of 1,175 participants, with 35.4% men and 64.6% women. Men had a significantly higher body weight and greater height, but there were no statistically significant differences in BMI (27.1 kg/m2 for men vs. 27.2 kg/m2 for women, p = 0.819). Visceral obesity was present in 56.1% of participants overall, with a higher prevalence in women compared to men (67.7% vs. 34.4%, p < 0.001). Smoking was more common among men compared to women (22.8% vs. 19.9%, p < 0.001). In terms of biochemical parameters, women had a significantly higher HDL cholesterol level and a lower serum creatinine level. The ePWV values did not differ significantly between men and women, with an average ePWV of 9.0 m/s across the study group. Table 1.Anthropometric and clinical characteristics of the study population divided by genderParticipants, NAll, N = 1175Men, N = 416Women, N = 759 p valueGender, % (N)​35.4 (416)64.6 (759)​Age, years51.0 (39.0–67.0)51.0 (39.0–67.0)52.0 (39.0–67.0)0.657Body weight, kg76.0 (67.0–88.0)84.0 (75.0–93.0)72.5 (64.0–83.0) <0.001 Body height, m167.0 (160.0–175.0)175.0 (170.0–180.0)164.0 (159.0–168.0) <0.001 BMI, kg/m227.2 (24.0–31.0)27.1 (24.5–30.5)27.2 (23.7–31.3)0.819BMI categories, % (N) <2532.5 (361)30.0 (119)33.9 (242)0.058 25–3037.0 (410)41.6 (165)34.4 (245)​ >3030.5 (339)28.5 (113)31.7 (226)​Waist circumference, cm95.0 (86.0–104.0)97.0 (90.0–104.0)93.0 (85.0–103.0) <0.001 Visceral obesity, % (N)56.1 (413)34.4 (88)67.7 (325) <0.001 Smoking status, % (N) Current24.9 (287)33.8 (139)19.9 (148) <0.001 Ex11.4 (131)19.5 (80)6.9 (51)​ Never63.7 (736)46.7 (192)73.2 (544)​SBP, mm Hg135.0 (120.0–153.0)138.8 (122.8–153.5)135.0 (119.0–153.9) 0.033 DBP, mm Hg81.5 (75.0–90.0)81.5 (75.0–90.9)81.5 (74.1–90.0)0.198MAP, mm Hg104.0 (93.0–115.0)104.7 (95.0–115.0)103.7 (92.3–114.5)0.072Arterial hypertension, % (N)63.1 (741)64.2 (267)62.5 (474)0.600Heart rate, bpm76.0 (70.0–85.0)75.5 (68.8–84.5)76.5 (70.5–85.5)0.212ePWV, m/s9.0 (7.0–12.0)9.1 (7.5–11.5)8.9 (7.1–11.7)0.560ePWV >10 m/s, % (N)39.1 (459)37.7 (157)39.8 (302)0.531Diabetes, % (N)13.3 (156)12.3 (51)13.8 (105)0.502eGFR, mL/min/1.73 m286.4 (74.0–99.0)88.3 (77.2–100.2)85.0 (72.9–98.5) 0.013 Serum creatinine, μmol/L76.0 (69.0–87.0)88.0 (80.0–95.0)71.0 (65.0–78.0) <0.001 Fasting blood glucose, mmol/L5.2 (4.8–5.8)5.2 (4.8–5.7)5.2 (4.7–5.8)0.150Total cholesterol, mmol/L5.7 (4.9–6.4)5.6 (4.9–6.4)5.7 (4.9–6.5)0.513LDL cholesterol, mmol/L3.5 (2.7–4.1)3.5 (2.8–4.1)3.4 (2.7–4.1)0.904HDL cholesterol, mmol/L1.5 (1.3–1.8)1.4 (1.2–1.7)1.6 (1.3–1.8) <0.001 Triglycerides, mmol/L1.3 (0.9–1.9)1.3 (0.9–1.9)1.3 (0.9–1.8)0.712ACR, mg/g5.5 (3.7–10.0)4.8 (3.0–9.0)5.9 (4.1–10.2) <0.001 BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; eGFR, estimated glomerular filtration rate; ePWV, estimated pulse wave velocity; LDL, low-density lipoprotein; HDL, high-density lipoprotein; ACR, albumin-to-creatinine ratio. p < 0.05. Anthropometric and clinical characteristics of the study population divided by gender BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; eGFR, estimated glomerular filtration rate; ePWV, estimated pulse wave velocity; LDL, low-density lipoprotein; HDL, high-density lipoprotein; ACR, albumin-to-creatinine ratio. p < 0.05. Table 2 describes the anthropometric and clinical characteristics of participants according to ePWV values. This analysis showed significant differences between the two groups, divided by ePWV thresholds of >10 m/s and <10 m/s. Participants with ePWV of >10 m/s were significantly older than those with ePWV of <10 m/s, had higher BMI and waist circumference, as well as significantly higher systolic, diastolic, and mean arterial pressures compared to those with ePWV of <10 m/s. Among those with ePWV of >10 m/s, there was a higher prevalence of arterial hypertension, diabetes, and significantly lower eGFR. These participants also had statistically significantly higher levels of fasting blood glucose, total cholesterol, LDL cholesterol, triglycerides, and albumin-to-creatinine ratio. Table 2.Anthropometric and clinical characteristics of subjects according to ePWV value​ePWV >10 m/s, N = 459ePWV <10 m/s, N = 716 p valueGender, % (N) Male34.2 (157)36.2 (259)0.531 Female65.8 (302)63.8 (457)​Age, years69.0 (63.0–74.0)41.5 (33.0–49.0) <0.001 Body weight, kg77.5 (69.0–88.0)76.0 (65.0–88.0)0.095Body height, m165.0 (158.0–171.0)169.0 (163.0–176.0) <0.001 BMI, kg/m228.7 (25.5–32.0)26.1 (23.4–30.1) <0.001 BMI categories, % (N) <2521.5 (90)39.2 (271) <0.001 25–3039.1 (164)35.6 (246)​ >3039.4 (165)25.2 (174)​Waist circumference, cm100.0 (90.3–107.0)92.0 (83.0–101.0) <0.001 Visceral obesity, % (N)69.5 (194)47.9 (219) <0.001 Smoking status, % (N) Current10.4 (47)34.1 (240) <0.001 Ex12.0 (54)11.0 (77)​ Never77.6 (350)54.9 (386)​SBP, mm Hg155.0 (142.0–170.0)125.0 (114.5–137.8) <0.001 DBP, mm Hg88.0 (80.0–98.0)79.5 (71.5–85.5) <0.001 MAP, mm Hg115.8 (105.3–126.0)97.0 (90.0–106.5) <0.001 Arterial hypertension, % (N)92.6 (425)44.1 (316) <0.001 Heart rate, bpm77.0 (70.0–85.0)75.8 (70.0–86.0)0.764ePWV, m/s12.1 (11.1–13.5)7.6 (6.7–8.7) <0.001 Diabetes, % (N)20.5 (94)8.7 (62) <0.001 eGFR, mL/min/1.73 m273.2 (61.0–82.8)94.0 (84.2–104.8) <0.001 Serum creatinine, μmol/L79.0 (70.0–93.0)75.0 (68.0–84.5) <0.001 Fasting blood glucose, mmol/L5.7 (5.1–6.3)5.0 (4.6–5.4) <0.001 Total cholesterol, mmol/L6.0 (5.2–6.3)5.4 (4.7–6.2) <0.001 LDL, mmol/L3.8 (3.0–4.4)3.2 (2.6–3.9) <0.001 HDL, mmol/L1.5 (1.3–1.8)1.4 (1.3–1.7) 0.031 Triglycerides, mmol/L1.5 (1.1–2.0)1.2 (0.8–1.8) <0.001 ACR, mg/g7.5 (5.0–18.4)4.7 (3.3–7.2) <0.001 ePWV, estimated pulse wave velocity; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; eGFR, estimated glomerular filtration rate; ePWV, estimated pulse wave velocity; LDL, low-density lipoprotein; HDL, high-density lipoprotein; ACR, albumin-to-creatinine ratio. p < 0.05. Anthropometric and clinical characteristics of subjects according to ePWV value ePWV, estimated pulse wave velocity; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; eGFR, estimated glomerular filtration rate; ePWV, estimated pulse wave velocity; LDL, low-density lipoprotein; HDL, high-density lipoprotein; ACR, albumin-to-creatinine ratio. p < 0.05. Table 3 presents the outcomes among participants divided by gender. Over an average follow-up period of 11.3 years, the overall mortality rate was 22.4% (263 participants). Men had a higher mortality rate compared to women (27.9% vs. 19.4%, p < 0.001). Men had a significantly higher rate of atrial fibrillation compared to women (7.7% vs. 4.5%, p = 0.031). The overall rate of heart failure was 5.3%. Men showed a slightly higher rate of heart failure compared to women (7.0% vs. 4.3%), but this difference was not statistically significant (p = 0.074). There were no differences between men and women in other nonfatal outcomes. The Table 4 shows the distribution of causes of death among men and women, with significant gender differences in CV and malignant causes. CV diseases were the leading cause, accounting for 48.1% of all deaths, with a significantly higher proportion among women compared to men (59.4 vs. 33.0%, p < 0.001). Malignant diseases accounted for 26.1% of deaths, with a higher incidence in men than women (37.9 vs. 17.4%, p < 0.001), while other causes showed no significant gender difference (p = 0.371). Table 3.Follow-up outcomes depending on genderFollow-up outcomesAll, N = 1,175Men, N = 416Women, N = 759 p valueDeath, % (N)22.4 (263)27.9 (116)19.4 (147) 0.001 Myocardial infarction, % (N)3.4 (40)4.3 (18)2.9 (22)0.261Atrial fibrillation, % (N)5.6 (66)7.7 (32)4.5 (34) 0.031 Heart failure, % (N)5.3 (62)7.0 (29)4.3 (33)0.074Stroke, % (N)3.5 (41)4.6 (19)2.9 (22)0.185TIA, % (N)2.7 (32)2.6 (11)2.8 (21)0.949Haemodialysis, % (N)1.0 (12)1.4 (6)0.8 (6)0.448TIA, transient ischemic attack. p < 0.05. Follow-up outcomes depending on gender TIA, transient ischemic attack. p < 0.05. Table 4.Causes of death: categories – CV, malignant, and otherCause of deathAll, N = 241Men, N = 103Women, N = 138 p valueCV diseases, % (N)48.1 (116)33.0 (34)59.4 (82) <0.001 Malignant diseases, % (N)26.1 (63)37.9 (39)17.4 (24) <0.001 Other, % (N)25.7 (62)29.1 (30)23.2 (32)0.371CV, cardiovascular. p < 0.05. Causes of death: categories – CV, malignant, and other CV, cardiovascular. p < 0.05. Table 5 presents the outcomes among participants divided by ePWV values. The overall mortality rate was significantly higher in those with ePWV of >10 m/s compared to those with ePWV of <10 m/s (49.2% vs. 5.2%, p < 0.001). The rates of myocardial infarction and initiation of haemodialysis were higher in participants with ePWV of >10 m/s compared to those with ePWV of <10 m/s (p = 0.007 and p = 0.023, respectively). Table 6 presents the causes of death divided by ePWV values greater than and less than 10 m/s, revealing significant differences for CV deaths. Among those with ePWV of >10 m/s, CV diseases were the primary cause of death, accounting for 48.2%, compared to 18.9% for those with ePWV of <10 m/s (p = 0.002). Malignant diseases were more common in the ePWV of <10 m/s group (35.1% vs. 22.1%, p = 0.127). Similarly, other causes were slightly more prevalent in the ePWV of <10 m/s group (45.9% vs. 30.1%, p = 0.082). Table 5.Follow-up outcomes depending on ePWV valueFollow-up outcomesePWV >10 m/s, N = 459ePWV <10 m/s, N = 716 p valueDeath, % (N)49.2 (226)5.2 (37) <0.001 Myocardial infarction, % (N)1.5 (7)4.6 (33) 0.007 Atrial fibrillation, % (N)5.7 (26)5.6 (40)0.942Heart failure, % (N)6.5 (30)4.5 (32)0.158Stroke, % (N)4.8 (22)2.7 (19)0.074TIA, % (N)2.8 (13)2.7 (19)0.999Haemodialysis, % (N)2.0 (9)0.4 (3) 0.023 ePWV, estimated pulse wave velocity; TIA, transient ischemic attack. p < 0.05. Follow-up outcomes depending on ePWV value ePWV, estimated pulse wave velocity; TIA, transient ischemic attack. p < 0.05. Table 6.Causes of death and ePWV values: categories – CV, malignant, and otherCause of deathePWV >10 m/s, N = 226ePWV <10 m/s, N = 37 p valueCV diseases, % (N)48.2 (109)18.9 (7) 0.002 Malignant diseases, % (N)22.1 (50)35.1 (13)0.127Other, % (N)30.1 (68)45.9 (17)0.082CV, cardiovascular. p < 0.05. Causes of death and ePWV values: categories – CV, malignant, and other CV, cardiovascular. p < 0.05. KM curves showed worse survival in participants with ePWV of >10 m/s compared to those with ePWV of <10 m/s (HR = 12.8, 95% CI 9.9–16.6, p < 0.001) (Fig. 1a). A similar difference was seen for nonfatal outcomes, with lower event-free survival in the higher ePWV group (HR = 1.50, 95% CI 1.0–2.1, p = 0.008) (Fig. 1b). KM survival analysis for fatal outcome (a) and for nonfatal outcomes (b) depending on the ePWV of >10 m/s or ≤10 m/s. In multivariable models for mortality, male sex and ePWV remained associated with higher risk (HR 1.75, 95% CI 1.18–2.61, p = 0.006; HR 1.81, 95% CI 1.08–3.04, p = 0.030). No variables reached significance for nonfatal outcomes. For the composite endpoint, both male sex (HR = 1.54, 95% CI 1.14–2.08, p = 0.006) and ePWV (HR = 1.94, 95% CI 1.50–2.51, p < 0.001) were linked with increased risk. In contrast, higher SBP and DBP were associated with lower risk (HR = 0.98, 95% CI 0.97–0.99, p = 0.005; HR = 0.98, 95% CI 0.97–0.99, p = 0.020). Detailed results of the univariable and multivariable regression analyses are presented in online supplementary Tables S1–S6, while additional data on participant flow, sensitivity analyses, and causes of death stratified by sex and ePWV are provided in online supplementary Tables S7–S9 and Figure S1. Discussion This study examined the predictive role of ePWV for overall mortality and both fatal and nonfatal CV, cerebrovascular and renal outcomes in a remote rural area in Croatia. Our findings confirm that elevated ePWV values (>10 m/s) are associated with a significantly higher risk of overall mortality, and CV, cerebrovascular and renal nonfatal outcomes supporting the existing evidence of arterial stiffness as a strong, independent predictor of adverse CV events [1–4]. These findings highlight the advantage of ePWV as a practical and accessible tool for CV risk assessment in resource-limited settings. Our results, along with those of other studies, clearly emphasize the clinical importance of ePWV as a surrogate marker of arterial stiffness, providing a practical solution for improving healthcare access in vulnerable populations, such as those in remote rural areas. Our results are consistent with findings from the Reference Values for Arterial Stiffness Collaboration and the study by Blacher et al. [14], which emphasized the importance of arterial stiffness in CV risk, especially among high-risk populations such as those with hypertension or diabetes [14, 15]. The results of our study reinforce the independent predictive value of ePWV for both fatal and composite fatal and nonfatal outcomes in this cohort. Even after adjusting for traditional risk factors such as age and blood pressure, ePWV remained a significant predictor, as demonstrated in our multivariate models. Our KM survival analyses further underscore the clinical relevance of ePWV, showing significantly lower survival rates in participants with ePWV of >10 m/s compared to those with ePWV of <10 m/s. Moreover, the association of ePWV with mortality and adverse outcomes persisted even when traditional risk scores, such as the Framingham Risk Score, were included in the analysis. This underscores the added value of ePWV in identifying high-risk individuals who might otherwise be underestimated by conventional risk assessments. Given the challenges of implementing direct PWV measurements in rural settings, the ability of ePWV to provide reliable predictions using readily available clinical parameters enhances its practicality for broader clinical use. These results advocate for the inclusion of ePWV in clinical practice guidelines, particularly for populations with limited access to advanced CV diagnostics. Thus, ePWV serves as a bridge to addressing healthcare disparities, offering a feasible approach to improving risk stratification and outcomes in underserved areas. This study highlighted the predictive strength of arterial stiffness markers, including PWV, for all-cause mortality, supporting our findings that ePWV serves as a reliable and accessible tool for stratifying CV risk, even in resource-limited settings. The predictive value of ePWV extends beyond traditional risk models. In the SPRINT trial, ePWV improved survival predictions and served as an indicator of treatment efficacy, particularly in hypertensive individuals [15]. Such findings emphasize the incremental value of ePWV over standard risk scores like SCORE and Framingham Risk Score [6, 7]. Additionally, studies by Greve et al. [4, 16] demonstrated that both measured and estimated PWV can similarly predict major CV events, underscoring ePWV’s practical value in settings where cfPWV is less feasible. Ben-Shlomo et al. [3] highlighted the benefit of PWV for CV event prediction, particularly in patients over 50 years old, further supporting the use of ePWV as a viable tool for CV risk stratification in various settings, including those with limited resources. Also, Huang et al. [17] reported that ePWV is an independent risk factor for all-cause and cardio-cerebrovascular disease mortality in stroke patients. Cruickshank et al. [18] observed that in diabetic patients, PWV displaced SBP as a mortality predictor, indicating that arterial stiffness is a more integrated marker of vascular health, especially in populations with complex comorbidities. Our study’s findings on the predictive value of ePWV for mortality are strongly supported by the results presented in the NHANES-based study [17]. Specifically, both studies found that an ePWV threshold of 10 m/s or greater is associated with a markedly higher mortality risk, supporting the use of this cut-point as a clinically relevant marker. Additionally, the NHANES study’s finding that each 1 m/s increase in ePWV raises mortality risk aligns with our results, further validating ePWV as an accessible and effective predictor of adverse outcomes across different populations [17]. The association of higher ePWV with increased mortality has been consistently observed in diverse populations, including the U.S. cohort analysed by Heffernan et al. [19, 20] and the Kailuan study in China, where ePWV of >10 m/s correlated with elevated risks of CV events and mortality, similar to our results [5]. In both the Health and Retirement Study and our study, ePWV is demonstrated to be a significant predictor of mortality [20]. The Health and Retirement Study found that ePWV, after adjusting for age, blood pressure, and other health-related factors, was associated with an increased likelihood of death over a 10- to 12-year period [19]. Similarly, our rural population study showed that participants with an ePWV of >10 m/s had significantly lower survival rates, with ePWV maintaining its predictive value for mortality even after adjusting for traditional CV risk factors. The increasing prevalence of CKD and associated CV complications, especially in underserved rural populations, highlights the importance of reliable and accessible diagnostic tools like ePWV. Townsend et al. [21] demonstrated that elevated PWV is a significant predictor of CKD progression and mortality, suggesting that arterial stiffness directly contributes to adverse renal outcomes and survival rates, independent of blood pressure control. Similarly, the findings by Côté et al. [22] emphasize that while ePWV equations may require improvement in specific subpopulations such as those with diabetes or CKD, they remain valuable for identifying individuals at increased risk. In rural regions, where direct PWV measurement is often impractical, ePWV could serve as a critical adjunct in early risk stratification and intervention planning, potentially reducing the burden of CKD and related CV diseases. Our findings in a rural Croatian population are consistent with the results of the recent meta-analysis by Li et al. [9], which demonstrated that higher ePWV is robustly associated with increased risks of CV events and mortality. While the meta-analysis included heterogeneous populations worldwide, our study reinforces its prognostic significance in the specific setting of resource-limited rural areas. Recent findings further reinforce the prognostic value of ePWV in populations with CKD and comorbidities such as atherosclerotic heart disease. Cui et al. [23] demonstrated that elevated ePWV is significantly associated with increased in-hospital and 1-year mortality in critically ill patients with CKD and atherosclerotic heart disease, even after adjusting for confounding factors. These results align with our findings, highlighting the role of ePWV as a practical tool for identifying high-risk individuals in rural populations where direct PWV measurement may not be feasible. Our study has certain limitations that need to be considered. One of the main limitations of our study is that we did not directly measure PWV. cfPWV is considered the gold standard for assessing arterial stiffness. However, our goal was not to compare the reliability of methods but to analyze the association of ePWV with fatal and nonfatal outcomes in a rural population. Using ePWV allowed us to explore this association in a resource-limited setting where direct PWV measurements are often unavailable. Furthermore, our cohort consisted exclusively of a rural population from the continental part of Croatia, and this might restrict the applicability of its findings to urban populations with different lifestyle factors and healthcare access. One limitation of this study is also the loss to follow-up and missing baseline ePWV data. However, sensitivity analysis showed comparable baseline characteristics between included participants and those lost to follow-up, suggesting limited impact on the main findings. Despite these limitations, the study has notable strengths. First, it included a large sample size and has a long follow-up period that allowed us to capture a substantial number of fatal and nonfatal outcomes, ensuring comprehensive outcome assessment. Second, our study highlights the practical utility of ePWV as a cost-effective and accessible tool for CV risk assessment in resource-limited settings such as remote rural areas. Conclusion While ePWV should not be considered a replacement for the gold standard cfPWV and should not be interpreted interchangeably, it serves as a complementary method, particularly valuable in settings where direct PWV measurement is not feasible. Our findings demonstrate that ePWV is a strong predictor of both fatal and nonfatal outcomes, underscoring its potential as a reliable and practical tool for CV risk assessment in resource-limited settings. These results confirm that ePWV can be a useful and dependable alternative for assessing arterial stiffness and predicting outcomes when access to advanced diagnostic methods is unavailable. Acknowledgments We are grateful to all the collaborators of this research project for their contributions to the fieldwork and laboratory analyses. Special thanks go to the farmers who willingly participated in the study, providing essential data for our research. We also extend our gratitude to Pero Ivanko from the Croatian Institute of Public Health (HZJZ) for providing critical data that supported this study. Statement of Ethics All participants signed a written informed consent and completed an extensive questionnaire containing personal and family history data, and were clinically examined. This study was performed in line with the principles of the Declaration of Helsinki. This study was conducted as part of the scientific project “Endemic Nephropathy in Croatia – Epidemiology, Diagnostics, and Etiopathogenesis” (ENAH, Ministry of Science, Education and Sports, grant No. 108-0000000329). The project and all its follow-up phases were approved by the following Ethical Committees: Ethics Committee of the University of Zagreb School of Medicine (approval issued 12 September 2003 – ref. no. 04-1213/03/2S/R12, approval issued 17 February 2006); Ethics Committee of the Croatian Institute of Public Health (approvals issued 25 February 2005, 20 October 2008 – ref. no. 001-0679/1-08/2, 16 April 2010 – ref. no. 001-270/1-10, 18 April 2016 – ref. no. 80-7701/1-6, and 8 February 2024 – Class: 030-02/24-15/2; Reg. no.: 117-15-24-2); Ethics Committee of the University Hospital “Dr. Josip Benčević” Slavonski Brod (approval issued 12 July 2004); and Ethics Committee of the University Hospital Centre Zagreb (29 September 2021, Class: 8. 1-21/260-2; Reg. No. 02/21 AG). Conflict of Interest Statement The authors have no conflicts of interest to declare. Funding Sources This work was supported by the Ministry of Science, Education and Sports of the Republic of Croatia through the project Endemic Nephropathy in Croatia – Epidemiology, Diagnostics and Etiopathogenesis (Grant No. 108-0000000329). The sponsor had no role in the design of the study; collection, analysis, or interpretation of data; writing of the manuscript; or the decision to publish the results. Author Contributions M.D. and V.D.: conception and design of work, collecting data, statistical analysis, interpretation of data. V.P. and D.R.: drafting of work. A.J., J.K., I.P., J.J., L.G., M.A.B., and K.Đ.: conception of work, collecting data, interpretation of data for work. M.F.: collection and analysis of data. K.C.: writing a draft, revision. B.J.: conception of work, draft writing, revision, approving final version. All authors have read and agreed to the published version of the manuscript.