Authors: Saki Hayashida, Naoki Haruyama, Hiroyuki Hayashida, Akiko Fukui, Osamu Ohta, Ryota Yoshitomi, Masaru Nakayama
Categories: Article, Aortic valve calcification, Cardiovascular event, Chronic kidney disease, Propensity score matching, Nephrology, Risk factors
Source: Scientific Reports
Authors: Saki Hayashida, Naoki Haruyama, Hiroyuki Hayashida, Akiko Fukui, Osamu Ohta, Ryota Yoshitomi, Masaru Nakayama
The association between aortic valve calcification (AVC) and cardiovascular (CV) events across diverse populations including patients with chronic kidney disease (CKD) remains controversial. This study aimed to determine whether AVC is associated with CV events in patients with CKD. In this prospective study, 1,279 participants with CKD were enrolled. A Cox proportional hazard model was applied to determine the association between AVC and CV events. The participants were divided into the following groups according to the number of calcified aortic cusps (CACs): no CACs (n = 922), one CAC (n = 209), and two to three CACs (n = 148). During a median follow-up of 2.9 years, CV events occurred in 185 participants. In multivariable Cox analyses, the hazard ratios (95% confidence intervals) of one CAC and two to three CACs for CV events compared with no CACs were 1.94 (1.32, 2.83) and 2.21 (1.46, 3.33), respectively. In a propensity score-matched cohort, participants with AVC (n = 284) had a significantly higher risk of CV events than those without AVC (n = 284). In CKD, the presence of AVC appears to be an independent risk factor for CV events, and the assessment of AVC is useful in predicting the prognosis.
The online version contains supplementary material available at 10.1038/s41598-025-21517-1.
Patients with chronic kidney disease (CKD) have a higher risk of cardiovascular (CV) events and mortality than the general population^1^. CV complications have also emerged as the most important issue involved in the poor prognosis of patients with CKD, particularly in patients with severe kidney failure and in patients on dialysis^2,3^. Cardiac valve calcification and vascular calcification are partly associated with CV complications in CKD^4^. The Kidney Disease: Improving Global Outcome guideline suggests that patients with CKD G3a–G5D with known vascular or valvular calcification have the highest CV risk^5^.
The prevalence of aortic valve calcification (AVC) has been reported in various cohort studies. In population-based studies, the prevalence of AVC ranged from 6.2% to 18.4%^6–11^. An association between CKD and the prevalence of AVC has been also documented. The reported prevalence of AVC ranges from 14.1% to 75.0% in patients on dialysis ^12–17^. However, in patients with CKD not on dialysis, the prevalence of AVC ranges from 18.6 to 47.9%^18–22^. Therefore, AVC appears to be more prevalent in patients with CKD than in the general population.
The association between AVC and adverse outcomes has been investigated in various populations. In the general population, studies reported that AVC was associated with CV events^6^ and mortality^7^, while Hoffmann et al*.* reported no significant association between AVC and CV events and mortality^8^. Additionally, in a large, population-based study, patients with both AVC and coronary artery calcification had an increased risk of CV events and all-cause death, whereas those with AVC alone did not^23^. In patients on dialysis, a few studies showed significant associations between AVC and all-cause mortality and CV deaths^16,24^. In contrast, other studies showed no significant associations between AVC and CV events and mortality^12,17^. Furthermore, few studies have reported the association between cardiac valve calcification and adverse outcomes in patients with CKD not on dialysis. One study reported that mitral valve calcification was independently associated with an increased risk of all-cause mortality, whereas AVC was not^18^. In another study, there was no significant association between cardiac valve calcification and CV events or all-cause mortality in this population^25^. Therefore, whether AVC is associated with CV events across diverse populations including patients with CKD remains controversial. Therefore, this study aimed to determine whether AVC is associated with CV events in patients with CKD not on dialysis.
Between June 2009 and June 2024, 1,455 consecutive Japanese patients who were admitted to the NHO Kyushu Medical Center for the evaluation of and education regarding CKD were selected. Of these, we excluded 68 who showed acute-on-chronic kidney injury, seven who had no available data for blood samples, 10 who underwent aortic valve replacement, 12 who were not precisely evaluated for AVC (n = 11) or had no available data for the left atrial diameter (LAD) (n = 1) on echocardiography, one who had a bicuspid aortic valve, five who did not undergo echocardiography, and 14 who did not perform an ankle–brachial blood pressure index test. The remaining 1,338 patients were discharged from the hospital without initiating kidney replacement therapy and were subsequently followed up at the same hospital. Of these, 59 who were lost to follow-up within 6 months of discharge were also excluded. Therefore, data of 1,279 patients who were collected up to December 2024 were prospectively analysed. Figure 1 shows a flow chart of the participant enrollment process. After discharge, participants were prospectively followed every 1 to 6 months at our nephrology outpatient department (or other outpatient departments of our hospital). At each visit, we reviewed laboratory data and evaluated the presence or absence of CV events or death using the hospital medical records. For participants lost to follow-up, we intermittently checked serum creatinine, CV events, and survival status by sending written inquiries to other hospitals or clinics they attended. A summary of the inquiry form is provided in Supplementary Figure. The latest surveillance was conducted between September 2024 and December 2024. On the basis of these follow-up methods, we evaluated CV events or death from any cause.Fig. 1Flow chart of the enrollment process CKD chronic kidney disease, AVC aortic valve calcification, LAD left atrial diameter, ABPI ankle-brachial blood pressure index, KRT kidney replacement therapy.
The study was approved by the Ethics Committee of the NHO Kyushu Medical Center (approval 09–09), registered with the University Hospital Medical Information Network (UMIN000017519), and performed in accordance with the guidelines of the Declaration of Helsinki. Written informed consent was obtained from all of the participants.
The primary endpoints were fatal or nonfatal CV events in the absence of kidney replacement therapy. These CV events were defined as follows according to a previous report^26^. We included atherosclerotic CV events (ACVEs), such as ischemic heart disease (IHD) (those requiring percutaneous intervention or coronary artery bypass grafting, acute myocardial infarction, and myocardial ischemia identified using myocardial scintigraphy), non-hemorrhagic stroke (brain infarction), requirement for interventions to treat peripheral artery disease (PAD) (percutaneous transcatheter angioplasty, lower-limb amputation, endarterectomy of the femoral artery, and bypass surgery), dissecting aneurysm of the thoracic and/or abdominal aorta, rupture of a thoracic or abdominal aortic aneurysm, requirement for a bypass or stent placement in a thoracic or abdominal aortic aneurysm or iliac artery aneurysm, requirement for stent placement for treating internal carotid or vertebral artery stenosis, and vertebral artery dissection. Additionally, we included nonatherosclerotic CV events (NACVEs), such as hospitalization for treating congestive heart failure (CHF), hemorrhagic stroke (brain hemorrhage, subarachnoid hemorrhage, and non-traumatic acute subdural hematoma), cardiac valvular diseases (sudden onset of severe aortic regurgitation, valve replacement surgery, and transcatheter aortic valve implantation for treating aortic stenosis), and sudden death. Non-CV death was defined as death that occurred in the absence of a CV event. Follow-up was defined as the period between baseline and a first event in participants who experienced events, or as the time to the completion of the study or loss to follow-up in participants who were censored.
Blood samples to measure serum creatinine, C-reactive protein (CRP), high-density lipoprotein cholesterol, low-density lipoprotein (LDL) cholesterol, serum uric acid, corrected serum calcium, serum phosphorus, hemoglobin, serum albumin, intact parathyroid hormone, and 1,25-dihydroxyvitamin D concentrations were obtained from participants early in the morning following an overnight fast on the second day of admission. Daily proteinuria was also measured. The eGFR (mL/min/1.73 m^2^) was calculated using the following new Japanese eGFR = 194 × serum creatinine^−1.094^ × age^−0.287^ (× 0.739 if female)^27^. The definition of hyperuricemia was a serum uric acid level of ≥ 7.0 mg/dL in men or ≥ 6.0 mg/dL in women^28^.
All of the participants were interviewed regarding their medical history, including hypertension, diabetes mellitus, prior CVDs, and malignancy. Demographic information (age and sex), medication history, and a history of smoking at presentation were recorded for each participant. Prior CVDs were defined as a history of IHD, CHF, stroke, PAD, thoracic and/or abdominal aortic aneurysm, and/or aortic dissection. PAD was defined as having a low ankle–brachial blood pressure index (< 0.9) or having undergone treatment for lower limb ischemia. The participants were also categorized according to their cigarette smoking status as current or past smokers. Body mass index was calculated as body mass (kg) divided by height (m^2^). Blood pressure was measured on three separate occasions on day 2 of hospitalization, with the participants in a sitting position, and the mean of the three values obtained was recorded.
In the present study, percutaneous echocardiography was performed by cardiologists who were skilled in the procedure and blinded to the participants’ detailed information. Left ventricular mass (LVM) was calculated using M-mode data obtained from parasternal long-axis images, according to the following formula^29^: LVM = 1.04 ([IVSd + LVPWd + LVDd]^3^ − LVDd^3^) − 13.6, where IVSd and LVPWd are the thicknesses of the interventricular septum and the posterior wall of the LV during diastole, respectively, and LVDd is the diameter of the left ventricle (LV) during diastole. The LVM index (LVMI) was expressed as LVM per square meter of body surface area, and was calculated using the Du Bois formula^30^ as body mass^0.425^ × height^0.725^ × 0.007184. We also evaluated moderate or severe aortic stenosis, as described previously^31^, and moderate or severe aortic regurgitation according to the degree of turbulence in the LV outflow tract during diastole on colour flow Doppler. The presence or absence of AVC was determined visually and only determined qualitatively using echocardiography, and the number of calcified aortic cusps (CACs) was examined.
Continuous data are expressed as the mean ± standard deviation or median (interquartile range) depending on the data distribution, and categorical data are expressed as the number (%). The participants were divided into the following three categories according to the number of CACs: no CACs, one CAC, and two to three CACs. A logistic regression model was also applied to identify the factors associated with AVC. The odds ratio and 95% confidence interval (CI) for having AVC were calculated for each variable. Survival curves were constructed using the Kaplan–Meier method and evaluated using the log-rank test. Cox proportional hazards models were used to determine whether AVC was associated with CV events, and the hazard ratios (HRs) and 95% CIs were calculated for each variable. Non-CV death before CV events was considered to be a competing event. Therefore, a Fine–Gray proportional subdistribution hazard model was also performed by taking into account the competing risk of non-CV death in the association between AVC and CV events^32^, and the subdistribution HRs (95% CI) for CV events were calculated. We selected the following covariates for multivariable Cox traditional CV risk factors (age, sex, smoking, diabetes mellitus, systolic blood pressure, dyslipidemia, body mass index, and prior CVDs); non-traditional CV risk factors (hemoglobin, CRP, serum phosphorus, serum albumin, and 1,25-dihydroxyvitamin D concentrations, and the eGFR); the presence of malignancy, which affects mortality; use of statins that affect CV events; and cardiac parameters, such as LAD, LVEF, LVMI, and aortic regurgitation or stenosis, which might be associated with CV events. Subgroup analyses were performed according to sex, the presence or absence of categorical variables, and the status of continuous data (values below or above the median value). The effects of interactions between AVC and other variables on CV events were evaluated by adding interaction terms for the associations between AVC and other variables to the relevant model. Propensity matching score with a caliper of 0.2 was performed using 1 nearest neighbour matching without replacement to minimize differences in baseline characteristics between participants with and without AVC. Propensity scores were determined on the basis of a multivariable logistic regression model that estimated the probability of AVC. Differences in the prevalence between the two groups were evaluated using the chi-square test and Fisher’s exact test of groups containing less than five individuals in any given cell. The statistical significance of differences between the two groups was examined using the Wilcoxon signed-rank test for nonparametric data or the unpaired Student t-test for parametric data. Statistical analyses were performed using STATA version 15 (Stata Corp., College Station, TX, USA), and P < 0.05 was considered to indicate statistical significance.
The median age of the patients (831 men and 448 women) was 71 years (range, 20–96 years). Among the patients, 213 (16.7%), 156 (12.2%), 247 (19.3%), 376 (29.4%), and 287 (22.4%) patients were categorized as having stages G1–2, G3a, G3b, G4, and G5 CKD, respectively. The prevalence of AVC was 28% (357 patients). Table 1 shows the clinical characteristics of the participants with and without AVC. The prevalences of stages G1–2, G3a, G3b, G4, and G5 CKD according to AVC status AVC ( −) group, 20.7% (n = 191), 13.7% (n = 126), 18.8% (n = 173), 26.8% (n = 247), and 20.1% (n = 185), respectively; AVC ( +) group, 6.2% (n = 22), 8.4% (n = 30), 20.7% (n = 74), 36.1% (n = 129), and 28.6% (n = 102), respectively. Participants with AVC were older, more likely to be men, and more likely to have a history of smoking, hypertension, diabetes mellitus, malignancy, or prior CVDs. The number of participants who had renin–angiotensin–aldosterone system inhibitors, β-blockers, vitamin K antagonists, statins, or calcium-containing agents (calcium carbonate or calcium aspartate) administered was higher in those with AVC than in those without AVC. Lower concentrations of LDL cholesterol, hemoglobin, serum albumin, and 1,25-dihydroxyvitamin D, a lower eGFR, and higher intact parathyroid hormone concentrations were found in participants with AVC than in those without AVC. Regarding cardiac parameters, the LAD and LVMI were higher in participants with AVC than in those without AVC. Additionally, the prevalence of aortic regurgitation or stenosis was higher in participants with AVC than in those without AVC.Table 1Baseline clinical characteristics of patients and incidence rates of CV events, ACVEs, and NACVEs with and without AVC.VariablesAll patientsAVC (–)AVC ( +) *POne CACTwo to three CACsP for trend (n = 1,279)(n = 922)(n = 357)(n = 209)(n = 148)Age (years)71 (59, 79)67 (54, 76)79 (73, 83) < 0.0177 (71, 82)81 (75, 85) < 0.01Male, n (%)831 (65)572 (62)259 (73) < 0.01147 (70)112 (76) < 0.01Smoking, n (%)690 (54)473 (51)217 (61) < 0.01125 (60)92 (62) < 0.01Hypertension, n (%)1,047 (82)715 (78)332 (93) < 0.01188 (90)144 (97) < 0.01Diabetes mellitus, n (%)476 (37)308 (33)168 (47) < 0.0193 (45)75 (51) < 0.01Prior CVDs, n (%)469 (37)281 (30)188 (53) < 0.0199 (47)89 (60) < 0.01 IHD, n (%)185 (14)106 (12)79 (22) < 0.0139 (19)40 (27) < 0.01 CHF, n (%)34 (3)23 (2)11 (3)0.567 (3)4 (3)0.69 Hemorrhagic stroke, n (%)20 (2)13 (1)7 (2)0.484 (2)3 (2)0.49 Non-hemorrhagic stroke, n (%)152 (12)89 (10)63 (18) < 0.0130 (14)33 (22) < 0.01 PAD, n (%)196 (15)100 (11)96 (27) < 0.0149 (23)47 (32) < 0.01TAA/AAA, n (%)81 (6)37 (4)44 (12) < 0.0120 (10)24 (16) < 0.01 Aortic dissection, n (%)7 (1)2 (0.2)5 (1)0.023 (1)2 (1)0.02Causes of CKD Chronic glomerulonephritis, n (%)392 (31)327 (35)65 (18) < 0.0142 (20)23 (16) < 0.01Hypertensive nephrosclerosis, n (%)346 (27)216 (23)130 (36) < 0.0173 (35)57 (39) < 0.01Diabetic nephropathy, n (%)272 (21)176 (19)96 (27) < 0.0153 (25)43 (29) < 0.01Other defined causes, n (%)233 (18)177 (19)56 (16)0.1436 (17)20 (14)0.09Unknown causes, n (%)36 (3)26 (3)10 (3)0.995 (2)5 (3)0.85Dyslipidemia, n (%)926 (72)655 (71)271 (76)0.08165 (79)106 (72)0.30Malignancy, n (%)97 (8)60 (7)37 (10)0.0219 (9)18 (12)0.01SBP (mmHg)132 (120, 144)131 (119, 143)135 (124, 148) < 0.01135 (124, 148)135 (125, 148) < 0.01DBP (mmHg)72 (66, 80)73 (67, 81)71 (64, 77) < 0.0171 (66, 78)68 (61, 76) < 0.01Use of RAAS inhibitors~,~
n (%)766 (60)514 (56)252 (71) < 0.01144 (69)108 (73) < 0.01Use of β-blockers, n (%)246 (19)163 (18)83 (23)0.0250 (24)33 (22)0.05Use of statins, n (%)476 (37)308 (33)168 (47) < 0.01102 (49)66 (45) < 0.01Use of vitamin K antagonists, n (%)60 (5)34 (4)26 (7) < 0.0113 (6)13 (9) < 0.01Use of active vitamin D3, n (%)98 (8)62 (7)36 (10)0.0423 (11)13 (9)0.11Use of calcium-containing agents , n (%)13 (1)6 (1)7 (2)0.044 (2)3 (2)0.048Body mass index (kg/m^2^)22.9 (20.5, 25.4)22.9 (20.5, 25.5)23.2 (20.8, 25.3)0.6223.5 (21.2, 25.4)22.6 (20.2, 25.1)0.86CRP (mg/dL)0.09 (0.05, 0.20)0.09 (0.05, 0.19)0.10 (0.05, 0.23)0.060.10 (0.05, 0.23)0.10 (0.05, 0.25)0.09Daily proteinuria (g)1.12 (0.28, 2.99)1.07 (0.27, 2.84)1.36 (0.32, 3.23)0.111.36 (0.28, 3.31)1.36 (0.37, 3.18)0.12HDL cholesterol (mg/dL)46 (37, 58)47 (37, 59)45 (37, 57)0.0944 (36, 55)47 (37, 60)0.24LDL cholesterol (mg/dL)98 (78, 122)101 (80, 126)91 (75, 114) < 0.0195 (79, 116)85 (70, 113) < 0.01Hemoglobin (g/dL)10.9 (9.4, 12.6)11.2 (9.6, 13.0)10.4 (8.9, 11.7) < 0.0110.8 (9.3, 11.9)9.7 (8.4, 11.1) < 0.01eGFR (mL/min/1.73 m^2^)28.8 (16.0, 48.5)32.1 (17.4, 54.2)23.9 (14.1, 35.7) < 0.0124.6 (14.1, 39.4)23.4 (14.2, 33.3) < 0.01Serum albumin (g/dL)3.4 (3.0, 3.7)3.5 (3.1, 3.8)3.3 (2.9, 3.6) < 0.013.4 (2.9, 3.7)3.3 (2.9, 3.6) < 0.01Hyperuricemia, n (%)643 (50)459 (50)184 (52)0.57105 (50)79 (53)0.46Corrected serum calcium (mg/dL)9.3 (9.1, 9.7)9.3 (9.1, 9.7)9.4 (9.1, 9.7)0.389.3 (9.1, 9.8)9.4 (9.1, 9.7)0.45Serum phosphorus (mg/dL)3.7 (3.3, 4.2)3.7 (3.3, 4.2)3.7 (3.3, 4.2)0.303.7 (3.3, 4.2)3.9 (3.3, 4.2)0.18Intact PTH (pg/mL)62 (41, 111)58 (39, 102)73 (48, 122) < 0.0166 (47, 124)79 (50, 122) < 0.011,25-dihydroxyvitamin D (pg/mL)30.4 (20.7, 41.9)31.2 (21.6, 42.7)28.0 (19.3, 38.6) < 0.0127.6 (19.5, 39.2)28.1 (19.1, 38.1) < 0.01LAD (mm)39 (34, 44)38 (34, 43)42 (37, 46) < 0.0142 (38, 46)41 (37, 45) < 0.01LVEF (%)69 (64, 74)69 (64, 73)69 (65, 74)0.3269 (65, 74)69 (63, 74)0.45LVMI (g/m^2^)117 (93, 147)112 (88, 142)130 (105, 160) < 0.01128 (103, 151)132 (109, 172) < 0.01AR/AS, n (%)41 (3)19 (2)22 (6) < 0.014 (2)18 (12) < 0.01AR, n (%)27 (2)16 (2)11 (3)0.134 (2)7 (5)0.04AS, n (%)16 (1)4 (0.4)12 (3) < 0.010 (0)12 (8) < 0.01Atrial fibrillation, n (%)44 (3)28 (3)16 (4)0.209 (4)7 (5)0.21CV events, *n (%)*185 (14)97 (11)88 (25) − 45 (22)43 (29) − Incidence rates of CV events (/100 person-years)3.672.369.52 − 7.9212.11 − ACVEs, *n (%)*112 (9)62 (7)50 (14) − 28 (13)22 (15) − Incidence rates of ACVEs (/100 person-years)2.221.515.41 − 4.936.19 − NACVEs, *n (%)73 (6)35 (4)38 (11) − 17 (8)21 (14) − Incidence rates of NACVEs (/100 person-years)1.450.854.11 − 2.995.91 − Values are expressed as the number (percent), or median (interquartile range).AVC aortic valve calcification, CV cardiovascular, ACVEs atherosclerotic CV events, NACVEs nonatherosclerotic CV events, CAC calcified aortic cusp, CVD cardiovascular disease, IHD ischemic heart disease, CHF congestive heart failure, PAD peripheral artery disease, TAA thoracic aortic aneurysm, AAA abdominal aortic aneurysm, CKD chronic kidney disease, SBP systolic blood pressure, DBP diastolic blood pressure, RAAS, renin–angiotensin–aldosterone system, CRP, C-reactive protein, HDL high-density lipoprotein, LDL low-density lipoprotein, eGFR estimated glomerular filtration rate, PTH parathyroid hormone, LAD left atrial diameter, LVEF, left ventricular ejection fraction, LVMI, left ventricular mass index, AR aortic regurgitation, AS aortic stenosis.; AVC ( +) = one calcified aortic cusp + two to three calcified aortic cusps.; Trend analyses were performed across 3 groups (AVC (-), one calcified aortic cusp, and two to three calcified aortic cusps).; Calcium-containing calcium carbonate or calcium aspartate.
Univariable logistic regression analyses are shown in Table 2. Older age, male sex, higher systolic and lower diastolic blood pressure, and the presence of smoking, hypertension, diabetes mellitus, prior CVDs, and malignancy were significantly associated with AVC. The use of renin–angiotensin–aldosterone system inhibitors, β-blockers, statins, vitamin K antagonists, active vitamin D3, and calcium-containing agents was associated with AVC. Higher CRP concentrations, lower LDL cholesterol, hemoglobin, serum albumin, and 1,25-dihydroxyvitamin D concentrations, and a lower eGFR were related to AVC. A higher LAD, a higher LVMI, and the presence of aortic regurgitation or stenosis were significantly associated with AVC.Table 2ORs for having AVC and HRs for CV events of AVC in univariable analyses.VariablesLogistic regression analysesfor AVCCox analysesfor CV eventsOR (95% CI)PHR (95% CI)PAge (per 10-year increase)2.47 (2.15, 2.84) < 0.011.97 (1.70, 2.27) < 0.01Male1.62 (1.24, 2.11) < 0.011.74 (1.26, 2.42) < 0.01Smoking1.47 (1.15, 1.89) < 0.011.95 (1.44, 2.65) < 0.01Hypertension3.84 (2.49, 5.94) < 0.013.24 (1.96, 5.36) < 0.01Diabetes mellitus1.77 (1.38, 2.27) < 0.011.86 (1.39, 2.48) < 0.01Prior CVDs2.54 (1.97, 3.26) < 0.013.80 (2.82, 5.12) < 0.01Dyslipidemia1.28 (0.97, 1.70)0.081.55 (1.10, 2.20)0.01Malignancy1.66 (1.08, 2.55)0.021.25 (0.72, 2.15)0.43SBP (per 10-mmHg increase)1.15 (1.07, 1.23) < 0.011.11 (1.03, 1.21) < 0.01DBP (per 10-mmHg increase)0.76 (0.68, 0.85) < 0.010.69 (0.60, 0.79) < 0.01Use of RAAS inhibitors1.91 (1.47, 2.48) < 0.011.93 (1.40, 2.65) < 0.01Use of β-blockers1.41 (1.05, 1.90)0.022.09 (1.50, 2.89) < 0.01Use of statins1.77 (1.38, 2.27) < 0.011.59 (1.19, 2.13) < 0.01Use of vitamin K antagonists2.05 (1.21, 3.47) < 0.012.07 (1.22, 3.51) < 0.01Use of active vitamin D31.56 (1.01, 2.39)0.041.35 (0.82, 2.23)0.23Use of calcium-containing agents 3.05 (1.02, 9.15)0.0460.90 (0.13, 6.44)0.92Body mass index (per 1-kg/m^2^ increase)1.00 (0.97, 1.03)0.960.95 (0.91, 0.99)0.01CRP (per 1-mg/dL increase)1.20 (1.01, 1.43)0.041.24 (1.07, 1.43) < 0.01Daily proteinuria (per 1-g/day increase)1.01 (0.97, 1.06)0.501.04 (0.996, 1.10)0.07HDL cholesterol (per 1-mg/dL increase)0.99 (0.99, 1.00)0.110.99 (0.98, 1.00)0.11LDL cholesterol (per 10-mg/dL increase)0.93 (0.89, 0.96) < 0.010.97 (0.93, 1.01)0.09Hemoglobin (per 1-g/dL increase)0.84 (0.79, 0.89) < 0.010.81 (0.76, 0.87) < 0.01eGFR (per 1-mL/min/1.73 m^2^ increase)0.98 (0.97, 0.98) < 0.010.98 (0.97, 0.99) < 0.01Serum albumin (per 1-g/dL increase)0.67 (0.55, 0.81) < 0.010.69 (0.56, 0.85) < 0.01Hyperuricemia1.07 (0.84, 1.37)0.571.03 (0.77, 1.37)0.87Corrected serum calcium (per 1-mg/dL increase)1.16 (0.91, 1.47)0.221.26 (0.95, 1.68)0.11Serum phosphorus (per 1-mg/dL increase)1.13 (0.97, 1.32)0.121.09 (0.88, 1.35)0.42Intact PTH (per 10-pg/mL increase)1.01 (0.997, 1.02)0.151.01 (0.999, 1.03)0.071,25-dihydroxyvitamin D (per 1-pg/mL increase)0.99 (0.98, 0.997) < 0.010.99 (0.98, 0.997)0.01LAD (per 1-mm increase)1.07 (1.05, 1.09) < 0.011.07 (1.05, 1.09) < 0.01LVEF (per 1-% increase)1.01 (0.99, 1.02)0.360.97 (0.96, 0.99) < 0.01LVMI (per 10-g/m^2^ increase)1.08 (1.05, 1.11) < 0.011.10 (1.08, 1.13) < 0.01AR/AS3.12 (1.67, 5.84) < 0.013.70 (2.09, 6.53) < 0.01Atrial fibrillation1.50 (0.80, 2.80)0.212.64 (1.50, 4.65) < 0.01OR, odds ratio; AVC, aortic valve calcification; HR, hazard ratio; CV, cardiovascular; CI, confidence interval; CVD, cardiovascular disease; SBP, systolic blood pressure; DBP, diastolic blood pressure; RAAS, renin–angiotensin–aldosterone system; CRP, C-reactive protein; HDL, high-density lipoprotein; LDL, low-density lipoprotein; eGFR, estimated glomerular filtration rate; PTH, parathyroid hormone; LAD, left atrial diameter; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; AR, aortic regurgitation; AS, aortic stenosis.; Calcium-containing calcium carbonate or calcium aspartate.
During a median follow-up of 2.9 years, fatal or nonfatal CV events occurred in 185 participants. Kaplan–Meier analysis showed significantly higher prevalences of CV events in patients with a higher number of CACs (Fig. 2). ACVEs occurred in 112 participants, with IHD in 41, non-hemorrhagic stroke in 35, interventions for PAD in eight, interventions for thoracic or abdominal aortic aneurysms or iliac artery aneurysms in 13, and other ACVEs in 15. Seventy-three NACVEs occurred, with CHF in 49 participants, hemorrhagic stroke in 10, cardiac valvular diseases in three, and sudden death in 11. Incidence rates of CV events, ACVEs, and NACVEs were higher in participants with AVC than without AVC , as shown in Table 1. The causes of non-CV death were malignancy in 41 participants, infection in 26, uremia in 13, other defined causes in 15, and unknown causes in 11. In univariable Cox analyses, many variables were associated with CV events (Table 2) .Fig. 2Kaplan–Meier curves for the absence of CV events in participants stratified according to the number of their calcified aortic cusps, compared using the log-rank test. CV cardiovascular; AVC aortic valve calcification, CAC calcified aortic cusp.
Table 3 shows the HRs and subdistribution HRs for CV events of AVC. In multivariable Cox analyses, participants with one CAC and two to three CACs had a significantly higher risk of CV events than those without CACs. Similarly, significant associations of AVC with ACVEs and NACVEs were also found. Furthermore, when the Fine–Gray model with non-CV death (n = 106) as a competing risk was used, the association between AVC and each outcome was similar to the above-mentioned results (Cox models).Table 3HRs and SHRs for each outcome of AVC.CV eventsNo. of eventsAVC (-)One CACTwo to three CACsAVC ( +) 97454388HR (95% CI)PHR (95% CI)PHR (95% CI)PModel 1Reference3.20 (2.23, 4.59) < 0.014.85 (3.36, 7.00) < 0.013.83 (2.85, 5.16) < 0.01Model 2Reference1.89 (1.30, 2.74) < 0.012.24 (1.51, 3.33) < 0.012.03 (1.48, 2.79) < 0.01Model 3Reference1.95 (1.34, 2.84) < 0.012.15 (1.44, 3.22) < 0.012.03 (1.47, 2.81) < 0.01Model 4Reference1.94 (1.32, 2.83) < 0.012.21 (1.46, 3.33) < 0.012.04 (1.47, 2.83) < 0.01SHR (95% CI)PSHR (95% CI)PSHR (95% CI)PModel 1Reference2.89 (2.03, 4.12) < 0.014.14 (2.87, 5.96) < 0.013.39 (2.54, 4.53) < 0.01Model 2Reference1.84 (1.27, 2.67) < 0.012.05 (1.36, 3.10) < 0.011.93 (1.40, 2.67) < 0.01Model 3Reference1.91 (1.31, 2.78) < 0.011.99 (1.31, 3.02) < 0.011.94 (1.41, 2.69) < 0.01Model 4Reference1.90 (1.30, 2.78) < 0.012.04 (1.34, 3.11) < 0.011.96 (1.41, 2.71) < 0.01ACVEsNo. of events62282250HR (95% CI)PHR (95% CI)PHR (95% CI)PModel 1Reference3.08 (1.95, 4.86) < 0.013.70 (2.25, 6.09) < 0.013.33 (2.26, 4.88) < 0.01Model 2Reference1.89 (1.18, 3.04) < 0.011.78 (1.04, 3.05)0.031.85 (1.22, 2.80) < 0.01Model 3Reference1.90 (1.18, 3.07) < 0.011.69 (0.98, 2.92)0.061.81 (1.19, 2.76) < 0.01Model 4Reference1.89 (1.16, 3.07)0.011.77 (1.02, 3.09)0.041.84 (1.20, 2.82) < 0.01SHR (95% CI)PSHR (95% CI)PSHR (95% CI)PModel 1Reference2.78 (1.78, 4.34) < 0.013.18 (1.95, 5.19) < 0.012.94 (2.02, 4.27) < 0.01Model 2Reference1.83 (1.13, 2.96)0.011.64 (0.94, 2.87)0.081.75 (1.14, 2.68)0.01Model 3Reference1.86 (1.14, 3.02)0.011.57 (0.89, 2.76)0.121.73 (1.13, 2.65)0.01Model 4Reference1.86 (1.14, 3.02)0.011.58 (0.89, 2.83)0.121.74 (1.13, 2.68)0.01NACVEsNo. of events35172138HR (95% CI)PHR (95% CI)PHR (95% CI)PModel 1Reference3.13 (1.74, 5.63) < 0.016.15 (3.54, 10.7) < 0.014.29 (2.68, 6.85) < 0.01Model 2Reference1.76 (0.96, 3.22)0.072.76 (1.51, 5.04) < 0.012.16 (1.31, 3.56) < 0.01Model 3Reference1.85 (1.00, 3.40)0.0482.55 (1.39, 4.69) < 0.012.15 (1.29, 3.57) < 0.01Model 4Reference1.84 (0.99, 3.42)0.062.54 (1.34, 4.81) < 0.012.12 (1.26, 3.58) < 0.01SHR (95% CI)PSHR (95% CI)PSHR (95% CI)PModel 1Reference2.90 (1.63, 5.16) < 0.015.35 (3.12, 9.17) < 0.013.88 (2.47, 6.11) < 0.01Model 2Reference1.76 (0.97, 3.22)0.062.54 (1.39, 4.64) < 0.012.10 (1.28, 3.42) < 0.01Model 3Reference1.87 (1.03, 3.42)0.042.35 (1.27, 4.33) < 0.012.09 (1.28, 3.44) < 0.01Model 4Reference1.90 (1.01, 3.54)0.0452.41 (1.28, 4.56) < 0.012.12 (1.28, 3.52) < 0.01Model Crude.Model Adjusted for age, sex, diabetes mellitus, LDL cholesterol, smoking, systolic blood pressure, prior CVDs, and BMI.Model Adjusted for model 2 plus use of statins, malignancy, CRP, hemoglobin, eGFR, 1,25-dihydroxyvitamin D, serum phosphorus, and serum albumin.Model Adjusted for model 3 plus LVEF, LAD, LVMI, and AR/AS.HR, hazard ratio; SHR, subdistribution hazard ratio; AVC, aortic valve calcification; CAC, calcified aortic cusp; CV, cardiovascular; CI, confidence interval; ACVEs, atherosclerotic cardiovascular NACVEs, nonatherosclerotic cardiovascular events; LDL, low-density lipoprotein; CVD, cardiovascular disease; BMI, body mass index; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; LVEF, left ventricular ejection fraction; LAD, left atrial diameter; LVMI, left ventricular mass index; AR, aortic regurgitation; AS, aortic stenosis.; AVC ( +) = one calcified aortic cusp + two to three calcified aortic cusps.Non-CV death corresponds to a competing risk. The number of non-CV death events was 67 in the AVC ( −) group, 21 in the one CAC group, and 18 in the two to three CACs group.
The adjusted HRs and subdistribution HRs for CV events of the presence of AVC in subgroups stratified by demographic and clinical characteristics are shown in Fig. 3. In Cox models, AVC was independently associated with CV events in all subgroups, while the Fine–Gray model showed that all subgroups, with the exception of the higher eGFR and lower CRP groups, showed a significant association between AVC and CV events. In the Cox model, there was a significant interaction for CV events between AVC and age.Fig. 3Adjusted HRs and SHRs for CV events of the presence of AVC among subgroups stratified by clinical parameters. Adjusted for age, sex, diabetes mellitus, LDL cholesterol, smoking, systolic blood pressure, prior CVDs, BMI, use of statins, malignancy, CRP, hemoglobin, eGFR, 1,25-dihydroxyvitamin D, serum phosphorus, serum albumin, LVEF, LAD, LVMI, and AR/AS. HR hazard ratio, SHR subdistribution hazard ratio, CV cardiovascular, AVC aortic valve calcification, CI confidence interval, LDL low-density lipoprotein, CVD cardiovascular disease, BMI body mass index, CRP, C-reactive protein, eGFR estimated glomerular filtration rate, LVEF left ventricular ejection fraction, LAD left atrial diameter; LVMI left ventricular mass index, AR aortic regurgitation, AS aortic stenosis. Non-CV death corresponds to a competing risk. The number of non-CV death events was 67 in the AVC ( −) group and 39 in the AVC ( +) group.
Baseline clinical characteristics after propensity score matching are shown in Supplementary Table 1. In the propensity score-matched cohort, absolute standardized differences between the two groups were < 0.1 for all variables. In the propensity score–matched cohort, the prevalences of participants with stages G1–2, G3a, G3b, G4, and G5 CKD were 5.6% (n = 32), 9.3% (n = 53), 23.4% (n = 133), 35.2% (n = 200), and 26.4% (n = 150), respectively. The c-statistic from the logistic regression model used to estimate the propensity score was 0.820, and after matching, 711 participants were excluded. In particular, the number of participants with stages G1–3a CKD decreased from 369 to 85. Supplementary Table 2 shows the HRs and subdistribution HRs for CV events of AVC after propensity score matching. In the fully adjusted Cox model, the presence of AVC was significantly associated with adverse outcomes (e.g., CV events, ACVEs, and NACVEs). Additionally, the fully adjusted Fine–Gray model with non-CV death (n = 66) as a competing risk showed that AVC was associated with all adverse outcomes of CV events, ACVEs, and NACVEs. Additionally, in the fully adjusted Cox and Fine–Gray models, one CAC and two to three CACs were significantly associated with CV events. With regard to the associations between ACVEs and the number of CACs, the Cox model showed significant associations for both one CAC and two to three CACs, while the Fine–Gray model showed a significant association with one CAC alone. As for NACVEs, in both Cox and Fine–Gray models, NACVEs were associated with two to three CACs, but not with one CAC.
In this study, AVC was identified as an independent factor for CV events in participants with CKD not on dialysis. Additionally, when CV events were divided into two categories, namely ACVEs and NACVEs, significant associations between AVC and both of these events were found. In subgroup analyses, significant associations between AVC and CV events were observed, despite the presence or absence of prior CVDs. Another subgroup analysis stratified by eGFR levels demonstrated that in the Cox model, participants with both lower and higher eGFR levels had a significant association between AVC and CV events. With respect to the Fine–Gray model, there was a significant association between AVC and CV events in the lower eGFR group, while a weak association was found in the higher eGFR group, although the finding did not reach statistical significance. No significant interactions for CV events between AVC and eGFR levels were observed in either model. Therefore, the associations between AVC and CV events appeared not to be affected by kidney function. Furthermore, because the associations between AVC and CV events obtained from the propensity score-matched cohort were similar to those before propensity score matching, the effects of AVC on CV events were considered robust.
A previous population-based study showed that AVC was associated with a high incidence of atherosclerotic risk factors, suggesting that AVC should be considered as a manifestation of systemic atherosclerosis^10^. In studies on patients with CKD, a combination of AVC and mitral valve calcification was associated with carotid artery lesions and PAD, which are considered surrogate markers of subclinical atherosclerosis^15,33^. Additionally, in one study, patients with CKD and AVC had a higher risk of having coronary artery disease^20^. In the present study, participants with AVC were more likely to be men, and to have smoking habits, diabetes mellitus, and hypertension, which are considered risk factors for atherosclerosis^34^. Furthermore, the prevalence of atherosclerotic comorbidities (e.g. IHD, non-hemorrhagic stroke, PAD, thoracic and/or abdominal aortic aneurysm, and aortic dissection) was higher in participants with AVC than in those without AVC. However, there was no significant difference in the prevalence of nonatherosclerotic comorbidities, such as CHF and hemorrhagic stroke, between participants with and without AVC. There may be strong associations between malnutrition, inflammation, and atherosclerosis in patients with CKD^35^. Previous studies have shown that higher high-sensitivity CRP or CRP (inflammation marker) and/or lower serum albumin (malnutrition marker) concentrations are related to AVC or cardiac valve calcification in patients with CKD not on dialysis^21,25^ and in those with end-stage kidney disease^14,15,36^. In the present study, univariable logistic analyses showed that higher CRP and lower serum albumin concentrations were associated with the presence of AVC. These findings suggest that AVC and atherosclerosis share common pathological mechanism and/or risk factors, which could in part reflect the inflammatory and malnutritional status. Hypoalbuminemia^37^ and higher CRP concentrations^38^ are associated with CV events. Therefore, this close relationship between AVC and atherosclerosis associated with malnutrition and inflammation may contribute to the effect of AVC on CV events.
Histological studies have shown that endothelial disruption due to increased mechanical or decreased shear stress, subendothelial lipid accumulation with superimposed immune cell infiltration, and adjacent microcalcifications occur in aortic valve tissues in the early stages of calcific aortic valve disease. Patients with sclerotic changes (leaflet thickening and calcification) in aortic valves, even under conditions of normal or near normal valve hemodynamics, have an increased risk of CV events^39,40^. Additionally, in one study on older adults, aortic sclerosis (the presence of AVC without hemodynamic obstruction) was observed in 29% of patients, and it was associated with an approximately 50% increase in the risk of CV events^41^. In this context, the finding that the presence of AVC significantly increased the risk of CV events in the present study might be attributable to a close association between AVC and aortic sclerosis.
A previous report showed that the AVC score significantly increased as the severity of concomitant aortic regurgitation increased in patients with severe aortic stenosis, suggesting that AVC causes concomitant aortic regurgitation^42^. Another study showed a correlation between the degree of AVC and the progression of aortic stenosis^43^. Therefore, AVC may lead to aortic stenosis or regurgitation. Long-term asymptomatic aortic regurgitation causes progressive LV dilatation, dyspnoea, and peripheral edema formation, while aortic stenosis leads to LV remodelling, hypertrophy, and dysfunction. Thereafter, these valve dysfunctions may contribute to the development of angina, heart failure, pulmonary edema, syncope, and sudden death^4^. Approximately 10% to 15% of patients with aortic sclerosis progress to valve obstruction (stenosis), accompanied by an increase in leaflet calcification, over 2 to 5 years^44^. A previous study reported that de novo AVC after the initiation of hemodialysis was associated with subsequent CV events^45^. In the present study, among 1,279 participants, only 41 showed aortic regurgitation or stenosis, and even in 357 participants with AVC, these valvular dysfunctions were found only in 22 participants. Furthermore, the evaluation of AVC and aortic valve disease by echocardiography was performed only at the time of enrolling in the study. Therefore, reassessment for AVC, aortic regurgitation, and aortic stenosis using echocardiography during follow-up may be important to identify the new development of AVC, aortic regurgitation, and aortic stenosis, to determine changes in the degree of aortic regurgitation or the progression of stenosis, and to precisely clarify the associations between AVC, aortic regurgitation or stenosis, and CV events.
This study has several limitations. First, all of the participants were recruited at a single regional hospital. Therefore, the sample was fairly homogeneous and subject to selection bias. Second, we recruited consecutive patients who were admitted to the hospital, they were relatively old, they were all Japanese, and the number of male participants was approximately 1.9 times higher than that of female participants. Third, a large number of participants with stages G1–3a CKD were excluded after propensity score matching. As shown in Table 1 and the prevalences of each stage CKD according to AVC status among all participants, participants without AVC covered a broad range of kidney function, whereas those with AVC tended to have more advanced CKD; the prevalence of stages G1–3a CKD was 34.4% (n = 317) in the AVC ( −) group versus 14.6% (n = 52) in the AVC ( +) group. This imbalance inevitably led to the exclusion of many patients with mild CKD from the AVC ( −) group after matching. Finally, we did not evaluate concentrations of fetuin-A, matrix-Gla protein, reactive oxygen species, asymmetric dimethylarginine, or lipoprotein (a), all of which might play roles in aortic valve calcific degeneration associated with CKD ^4,46^.
In conclusion, this study investigated whether AVC is associated with CV events in patients with CKD not on dialysis, and we found the association between AVC and adverse outcomes was independent of CV risk factors and cardiac alterations. In CKD, the assessment of AVC is useful for investigating the risk factors for clinical outcomes.
Supplementary Information 1. Supplementary Information 2. Supplementary Information 3.