Mitral annulus calcification and the risk of infective endocarditis: insights from DANCAVAS screening trials Authors: Lytfi Krasniqi, Axel C P Diederichsen, Jes S Lindholt, Camilla Engelsgaard, Martin Busk, Lars Frost, Jess Lambrechtsen, Marek Karon, Jordi Sanchez Dahl, Frank A Flachskampf Published: 2026 ABSTRACT AbstractAimsRisk stratification in infective endocarditis (IE) does not incorporate structural substrates involved in early disease initiation. The aim of this study was to examine the association between mitral annulus calcification (MAC) burden and the risk of IE.Methods and resultsThis substudy utilized the population-based DANCAVAS cohort undergoing non-contrast computed tomography (CT) with MAC quantification with follow-up until 19 August 2025. The primary outcome was incidence rates and 10-year risk of IE across three exploratory categorical groups: No MAC (0 arbitrary units (AU)), MAC 1–299 AU, and MAC ≥300 AU. A total of 14 793 participants were included with mean (SD) age, 66.9 (3.8) years, and 14 053 males (95.0%). Median MAC was 13 AU (IQR3.3–56.2) in the MAC 1–299 group and 680 AU (IQR447–1271) in the MAC ≥300 group. IE incidence rate per 10 000 person-years was 1.92 for no MAC, 2.48 for MAC 1–299, and 21.3 for MAC ≥300 (P < 0.001). The 10-year cumulative risk was, 0.20%(95%CI 0.12–0.33) for no MAC, 0.22% (95%CI 0.10–0.49) for MAC 1–299, and 1.85% (95%CI 0.93–3.69) for MAC ≥300 (P < 0.001). Each 100 AU increase in MAC was associated with a 3% increase in the risk of IE (HR 1.03, 95%CI 1.01–1.05, P = 0.011). MAC ≥300 AU remained associated with IE after adjustment for established risk factors (HR 7.05, 95%CI 2.85 to 17.4, P < 0.001).ConclusionGreater MAC burden was associated with a higher risk of IE. In exploratory categorical analyses, MAC ≥300 AU was associated with the highest observed risk, although these findings should be interpreted with caution given the limited number of events.Keywords: Mitral annulus calcification, Infective endocarditis, Non-contrast computed tomography Introduction The onset of infective endocarditis (IE) depends on several factors: The presence of a surface that bacteria can colonize, pathogens entering the bloodstream, and the effectiveness of the host’s immune response.1,2 Even with intensive treatment, IE continues to cause significant morbidity and mortality.1–3 Therefore, understanding how to categorize patients at risk of IE is crucial, as this could help identify those who might benefit from preventive strategies. We recently demonstrated that the degree of aortic valve calcification assessed by non-contrast computed tomography (CT), was associated with the risk of IE independent of well-established risk factors as cardiac implantable electronic device (CIED), diabetes mellitus, and a history of valvular heart disease.4 The risk of aortic valve calcification and IE was incremental, surpassing the risk seen in patients with a history of valvular heart disease, a group considered at intermediate risk of IE.5 Recently, it has been demonstrated that a bacterial pathogen such as Staphylococcus aureus IE can originate from a calcified mitral annulus, and it has been speculated that mitral annular calcification (MAC) may serve as a nidus for pathogens leading to IE.6,7 MAC is readily assessable by noncontrast CT. However, no validated CT thresholds have been established for stratifying IE risk according to MAC burden, and the thresholds evaluated in the present study should therefore be considered exploratory. Although it is plausible that the presence of calcium on the mitral valve increases the risk of IE, data describing the risk of IE in patients with MAC remain limited. The objective of this study was therefore to examine the association between MAC burden and the risk of IE. Methods Ethical approval This study received approval from the Danish Data Protection Agency and the Regional Scientific Ethical Committee of Southern Denmark (S-20140028, 2014). Informed consent, both written and verbal, was obtained from all participants at the time of enrolment. Data availability The data supporting the findings of this study are available from the corresponding author upon reasonable request, subject to approval under applicable Danish legal regulations, and a compliant data access agreement prior to release. Trial design and participants This observational subgroup analysis is based on the DANCAVAS trial (NCT03946956), a population-based cardiovascular screening study that enrolled approximately 15 000 participants (predominately male) aged 60–74 years between 2014 and 2019.8,9 The DANCAVAS I trial enrolled both sexes aged 65–74, but cost-effectiveness analysis led to recruiting only men from 2015 onward.10 Therefore, the DANCAVAS II trial included only men aged 60–64.9 Participants from the regions of Southern and Central Denmark were selected. There were no exclusion criteria.11 The examinations were conducted at five screening sites: Odense University Hospital, Svendborg Hospital, Vejle Hospital, Silkeborg Hospital, and Nykoebing Falster Hospital. This substudy included patients with available MAC measurements, excluding individuals with prosthetic heart valves. Data source Follow-up data for the DANCAVAS population were data linked at an individual level from several documented national Danish healthcare registries.12–18 In Denmark, all residents are assigned a unique identification number at birth or immigration, enabling cross-registry linkage of individual-level data.13 The Danish registries collect data prospectively and comprehensively, and as all Danish citizens receive healthcare funded by the Danish National Health Service due to the universal health care system based on the Beveridge model. Therefore, complete follow-up for all patients is secured. The following registries were utilized: the Central Personal Registration System (vital status, date of death, and cases of migration),13 and the Danish National Patient Registry (DNPR)(all hospital contacts, International Classification of Diseases Tenth Revision diagnostic codes, admission time and location, and surgeries).15 Interventions and exposure Patients who were screened and randomized to the intervention group underwent Non-Contrast CT scans, while the control group received standard care. Imaging protocols and settings have been previously described.8,19 MAC was scored using the Agatston method and expressed in arbitrary units (AU). MAC was defined as calcium deposition at the junction between the left atrium and left ventricle at the level of the mitral annulus, excluding calcium in the mitral leaflets, aortic valve, left main coronary artery, left circumflex artery, right coronary artery, and leads. Measurements performed on a SyngoVia workstation (Siemens Healthcare Solutions). MAC quantification was performed by readers blinded to subsequent outcome status. MAC was evaluated as a continuous exposure. In exploratory analyses, MAC burden was also categorized as no MAC (0 AU), MAC 1–299 AU, or MAC ≥300 AU. These thresholds were borrowed from previous AVC analyses4,20 and have not been validated for MAC or IE risk stratification. Follow-up and outcomes All patients were followed until death, emigration, mitral valve replacement or aortic valve replacement (implantation included), or end of study period (19 August 2025). The primary outcome was IE, and secondary outcomes included a composite endpoint of IE and all-cause death, bacterial species, and location of infection. Risk was assessed by comparing incidence rates, and 5- and 10-year risk across MAC groups. All possible cases of IE were identified using the DNPR15 with the following ICD-10 codes: I011 (acute rheumatic endocarditis), I33.x (acute or subacute endocarditis), I38.x (Endocarditis, valve unspecified), I39.x (endocarditis and heart valve disorders in diseases classified elsewhere), Z035D (observation for other suspected cardiovascular diseases or infective endocarditis). Patients with a record registered with the relevant ICD-10 code were subsequently reviewed independently by two physicians. In Denmark, diagnostic evaluation and management of suspected IE adhere to current ESC guidelines and the modified Duke framework, and final treatment decisions are routinely made through multidisciplinary team conferences dedicated to IE. Cases were classified as IE based on the conclusions from these evaluations. Patients who did not meet criteria for definite or possible IE at admission, with or without a firm alternative diagnosis, had IE diagnosis rejected. Prevalent valvular heart disease was identified using ICD-10 diagnostic codes, specifically I34 for nonrheumatic mitral valve disorders and I35 for nonrheumatic aortic valve disorders, consistent with the previous methodology.5 We report definitions of outcomes, ICD10 codes, and procedure codes in the Supplementary material online, eTable S1. Statistics Continuous data were presented as mean (standard deviation (SD)) if normally distributed and as median (interquartile range (IQR)) if non-normally distributed. Categorical data were presented as proportions. Inter-group comparisons were done with Student's t-tests or Wilcoxon rank-sum tests, Chi-squared test or Fisher’s exact test, as appropriate. The primary analysis examined the association between MAC modelled continuously per 100 AU and IE using Cox regression. Time-to-event was calculated as the time from screening to the date of event, which included Kaplan–Meier estimations and plots, log-rank tests, and Cox Proportional Hazards regression. Proportional hazard assumptions were assessed with Schoenfeld residuals and reported if not satisfied. Incidence rates were formed from the number of events divided by the person-time (with stptime in STATA). A multivariable Cox regression model was used to assess the association between MAC and IE, adjusting for predefined clinical variables (CIED, age, sex, diabetes mellitus, chronic kidney disease, and aortic valve calcification per 100 AU). Complete case analyses were performed due to the low proportion of missing data and the assumption that data were missing at random. Interactions were tested in the univariate Cox regression model by including interaction terms between covariates and the primary predictor (MAC 1–299 vs. ≥300). Competing risk analyses were performed using the Aalen-Johansen estimator and the Fine and Gray subdistribution hazard model. The MAC threshold was examined in all with > 0 MAC arbitrary units (AU) using Youden's index. As a sensitivity analysis, all tests were performed including only IE on the mitral valve or unspecified, and we also applied the widely used registry-based approach with the same ICD codes without validation of events, which has been established in previous high-impact cardiovascular studies.5,21,22 Missing data were reported in the tables with []. Point estimates were supplemented by 95% confidence intervals (CI) when appropriate, and a P-value of less than 0.05 was considered significant. Statistical analysis was performed with STATA/IC18 (StataCorp, College Station, TX 77845, USA). Role of the funding source The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Results From 25 September 2014, to 27 February 2019, a total of 14 987 individuals were randomized and underwent a screening with non-contrast CT in the DANCAVAS trial (Structured Graphical Abstract). We excluded patients with a prosthetic heart valve (mitral valve n = 9, aortic valve n = 94), or those with missing MAC measurement (n = 91), resulting in the final study population of 14 793 individuals (see Supplementary material online, eFigure S1). Baseline characteristics were comparable between the groups (Table 1). Patients were predominantly male (95.0%) with a mean (SD) age of 66.9 (3.8) years; 1968 (13.3%) had diabetes mellitus, and 259 (1.8%) had a CIED. Median (IQR) MAC was 13 AU (IQR 3.3–56.2) in the group with MAC 1–299 (n = 2948), and 680 AU (IQR 447–1271) in the group with MAC ≥300 (n = 498). Table 1Descriptive baseline characteristics of study populationTotalNo MACMAC 1–299MAC≥300Individuals, no. (%)14 79311.347 (76.7%)2948 (19.9%)498 (3.4%)Mean (±SD) or median [IQR]Mean (±SD) or median [IQR]Mean (±SD) or median (IQR)Mean (±SD) or median [IQR] Demographics Male sex14 053 (95.0%)10 766 (94.9%)2814 (95.5%)473 (95.0%) Age (years)66.9 (3.8)66.9 (3.8)68.5 (3.5)69.1 (3.4) Body mass index (kg/m2)27.9 (4.4) [12]27.9 (4.4) [11]28.4 (4.3) [1]29.3 (5.1) Body surface area (m2)2.1 (0.2) [12]2.1 (0.2) [11]2.1 (0.2) [1]2.1 (0.2) Waist size (cm)101.2 (12.4) [22]101.2 (12.4) [15]103.3 (12.3) [7]106.0 (14.5) Systolic blood pressure (mmHg)148.8 (18.7) [53]148.8 (18.7) [31]151.0 (18.7) [21]153.1 (19.6) [1] Diastolic blood pressure (mmHg)82.4 (9.8) [54]82.4 (9.8) [32]82.8 (10.0) [21]82.1 (10.4) [1] Blood samples Hemoglobulin (mmol/L)9.1 (0.7) [93]9.1 (0.7) [74]9.0 (0.7) [15]8.9 (0.9) [4] Creatinine (mikromol/L)87.9 (20.9) [92]87.9 (20.9) [72]87.1 (23.2) [16]92.4 (53.0) [4] C-reactive protein (mg/L)3.3 (7.2) [1556]3.3 (7.2) [1317]3.4 (6.8) [208]5.0 (13.0) [31] Haemoglobin A1c (mmol/mol)38.8 (7.3) [111]38.8 (7.3) [83]40.0 (7.3) [22]42.2 (10.3) [6] Calcification (Agatston Score) Mitral annulus calcification0 [0–0]0 [0-0)13 [3–56]680 [447–1271] Aortic valve calcification5 [0–78] [24]1 [0–47] [17]41 [5–172] [7]201 [37–650] Coronary artery calcification113 [9–478] [4]87 [5–383]214 [35–731]533 [161–1506] n (%) n (%) n (%) n (%) Comorbidities Valvular heart disease329 (2.2%)188 (1.7%)93 (3.2%)48 (9.6%) Diabetes mellitus1968 (13.3%)1331 (11.7%)500 (17.0%)137 (27.5%) Active or previous smoker9739 (65.8%) [59]7350 (64.8%) [50]2029 (68.8%) [9]360 (72.3%) Hypertension7429 (50.2%)5367 (47.3%)1696 (57.5%)366 (73.5%) Chronic kidney disease1437 (9.7%)1081 (9.5%)283 (9.6%)73 (14.7%) Stroke931 (6.3%)635 (5.6%)236 (8.0%)60 (12.0%) Atrial fibrillation2083 (14.1%)1464 (12.9%)525 (17.8%)94 (18.9%) Chronic obstructive pulmonary disease1289 (8.7%)931 (8.2%)286 (9.7%)72 (14.5%) Previous infective endocarditis7 (<0.1%)5 (<0.1%)1(<0.1%)1 (0.2%) Cardiac implantable electronic device259 (1.8%)189 (1.7%)58 (2.0%)12 (2.4%)Data are presented as mean (±standard deviation (SD)) or median [IQR 25–75] for continuous variables and n (%) for categorical variables, respectively.Missing data were reported in the tables with []. Descriptive baseline characteristics of study population Data are presented as mean (±standard deviation (SD)) or median [IQR 25–75] for continuous variables and n (%) for categorical variables, respectively. Missing data were reported in the tables with []. Median follow-up was 8.5 years (IQR 7.6–9.3 years), with a total follow-up of 121 567 patient-years. We identified 203 potential IE cases. Of these, 51 cases were confirmed as definite IE events. Seven cases of IE were prior to enrolment in the DANCAVAS study, with none experiencing an IE event postrandomization. Forty-four cases occurred after DANCAVAS enrolment, and 12 of these were in patients after being censored due to mitral valve replacement or aortic valve replacement after enrolment, resulting in 32 confirmed cases with IE (Table 2). Of the 32 patients with definite IE, nine underwent cardiac surgery, with histopathologic examination confirming IE in all nine. Patients with IE had higher MAC, increased body mass index and waist size, increased levels of creatinine, and were more likely to have a history of diabetes mellitus, atrial fibrillation, and CIED, but did not show significantly higher aortic valve calcification (mean AVC 45 vs. 33, P = 0.09) or chronic kidney disease (see Supplementary material online, eTable S2). Table 2Association between mitral annular calcification and infective endocarditisHazard ratio (95% CI) P-value Univariable analysis (number of subjects: 14.793)Mitral annulus calcification score pr. 100 AU1.03 (1.01–1.05) 0.011 Multivariable analysis (number of subjects: 14.677)Mitral annulus calcification ≥300a7.05 (2.85–17.4) <0.001 Mitral annulus calcification categorical 1–2991.09 (0.42–2.85)0.86 Cardiac implantable electronic device7.56 (2.61–21.9) <0.001 Age0.96 (0.87–1.06) 0.40 Male sex1.51 (0.20–11.2)0.69 Diabetes mellitus2.80 (1.32–5.95) 0.007 Chronic kidney disease1.30 (0.49–3.46)0.60 Aortic valve calcification score pr. 100 AU1.07 (1.03–1.10) <0.001 Incidence ratesRate pr. 10.000 years (95% CI) P-valuebNo mitral valve calcification1.92 (1.21–3.05)Mitral valve calcification 1–2992.48 (1.11–5.52) P < 0.001Mitral annulus calcification ≥30021.3 (10.7–42.6)Cardiac implantable electronic devicec20.4 (7.66–54.4) P = 0.002Distribution of endocarditis cases by bacterial typeTotalNo MACMAC 1–299MAC≥300 N (%) N = 32 N = 18 (0.16%) N = 6 (0.20%) N = 8 (1.61%) Streptococcus mitis 4 (12.5%)3 (16.7)01 (12.5%) Streptococcus bovis 2 (6.3%)1 (5.6%)01 (12.5%) Streptococcus pneumoniae 1 (3.1%)1 (5.6%)00Streptococcus agalactiae1 (3.1%)1 (5.6%)00 Staphyloccocus aureus 11 (34.4%)7 (38.9%)1 (16.7%)3 (37.5%) Staphyloccocus epidermidis 5 (15.6%)1 (5.6%)3 (50.0%)1 (12.5%) Enterococcus faecalis 4 (12.5%)2 (11.1%)02 (25.0%) Candida dubliniensis 1 (3.1%)1 (5.6%)00 Corynebacterium striatum 1 (3.1%)01 (16.7%)0 Pseudomonas aeruginosa 1 (3.1%)1 (5.6%)00 Providencia rettgeri 1 (3.1%)01 (16.7%)0Location of infectionTotalNo MACMAC 1–299MAC≥300Mitral valve5 (15.6%)2 (11.1%)03 (37.5%)Aortic valve19 (59.4%)12 (66.7%)5 (83.3%)2 (25.0%)Cardiac implantable electronic deviceb4 (12.5%)1 (5.6%)1 (16.7%)2 (25.0%)Unspecified4 (12.5%)3 (16.7%)01 (12.5%)No interaction between sex, diabetes mellitus, peripheral arterial disease, or cardiac implantable electronic device. P-values <0.05 are written in bold, indicating significant differences between groups.AU, arbitrary units; CI, Confidence interval. aNo MAC was the reference. bSubjects with an MAC 1–299 compared to subjects with an MAC ≥300. cAt the time of randomization all proportional hazards assumption for Cox regression models were met. Association between mitral annular calcification and infective endocarditis No interaction between sex, diabetes mellitus, peripheral arterial disease, or cardiac implantable electronic device. P-values <0.05 are written in bold, indicating significant differences between groups. AU, arbitrary units; CI, Confidence interval. aNo MAC was the reference. bSubjects with an MAC 1–299 compared to subjects with an MAC ≥300. cAt the time of randomization all proportional hazards assumption for Cox regression models were met. For each 100 AU increase in MAC was associated with a 3% higher risk of IE (HR 1.03, 95%CI 1.01–1.05, P = 0.011)(Table 2). Valvular heart disease was not significantly associated, with an HR of 1.72 (95%CI 0.23–12.6, P = 0.59) (see Supplementary material online, eTable S3). Risk of IE in MAC≥300 compared to No MAC was significantly increased (adjusted HR 7.05 (95%CI 2.85–17.4), P < 0.001), or MAC≥300 compared to MAC 1–299 (adjusted HR 7.36, 95%CI 2.46–22.0, P < 0.001). The incidence of IE was 0.16% (n = 18) in the No MAC group, 0.20% (n = 6) in the MAC 1–299 group, and 1.61% (n = 8) in the MAC ≥300 group (P < 0.001). IE incidence rate per 10 000 person-years was 1.9 (95%CI 1.21–3.05) for No MAC, 2.5 (95%CI 1.11–5.52) for MAC 1–299, and 21.3 (95%CI 10.7–42.6) for MAC ≥300 (P < 0.001) (Table 2). This association remained in a Fine–Gray competing risk model accounting for the competing risk of death and adjusted according to the same covariables (subHR 6.29, 95%CI 2.42–16.3, P < 0.001), and a 10-year risk of 1.6% (95%CI 0.8–3.07). As a sensitivity analysis restricted to patients with native mitral valves IE or unspecified only (n = 9 events), MAC remained significantly associated with IE (HR 4.19, 95%CI 1.80–9.71, P = 0.001). No IE events were observed among patients with CIED and No MAC. The 5- and 10-year IE rates were 0.08% (95%CI 0.04–0.16) and 0.20% (95%CI 0.12–0.33) for the No MAC group, 0.18% (95%CI 0.07–0.43) and 0.22% (95%CI 0.10–0.49) for the MAC 1–299 group, and 1.08% (95%CI 0.45–2.56) and 1.85% (95%CI 0.93–3.69) in the MAC ≥300 group (log rank test MAC 1–299 vs. ≥300, P < 0.001, Figure 1). The Kaplan–Meier 5 and 10 year all-cause mortality was 4.48% (95%CI 4.11–4.88) and 14.2% (95%CI 13.3–15.1) for the No MAC group; 6.61% (95%CI 5.77–7.57) and 19.2% (95%CI 17.4–21.2) for the MAC 1–299 group, and 15.3% (95%CI 12.4–18.7) and 35.7% (95%CI 30.5–41.4) for the MAC ≥300 group (log rank test MAC 1–299 vs. ≥300, P < 0.001, Figure 1). Kaplan–Meier curves for risk of infective endocarditis and all-cause mortality stratified by mitral annulus calcification score. Cumulative incidence of infective endocarditis (IE) among patients grouped by mitral annulus calcification (MAC) score. The blue line represents patients with no MAC, the red line represents patients with MAC scores less than 300 arbitrary units (AU), and the green line represents patients with MAC scores ≥300 AU. *Unadjusted hazard ratio calculated between patients with an MAC 1–299 and MAC ≥300. The Kaplan–Meier estimates of composite endpoint after 5 and 10 years were 4.52% (95%CI 4.15–4.92) and 14.3% (95%CI 13.4–15.2) for the No MAC group, respectively, 6.64% (95%CI 5.80–7.61) and 19.3% (95%CI 17.4–21.3) for the MAC 1–299 group, and 16.0% (95%CI 13.0–19.5) and 36.6% (95%CI 31.4–42.2) for the MAC ≥300 group (log rank test MAC 1–299 vs. ≥300, P < 0.001). The causes of IE were Streptococci in 8 cases (25%), Staphylococci in 16 (50%), enterococci in 4 (13%), and other species in 4 (13%). Among the eight IE events in participants with MAC ≥300 AU, three involved the mitral valve, whereas the remaining five involved the aortic valve, a CIED, or had an unspecified location. In a post hoc exploratory analysis, Youden’s index identified a cut-off of 321 AU, with an area under the ROC curve of 0.72. In exploratory subgroup analyses, the association between MAC ≥300 AU and IE was directionally consistent across most subgroups, with no statistically significant interactions except for sex and coronary artery calcification, although some estimates were imprecise because of the limited number of events (Figure 2). The association was also directionally consistent among patients with and without valvular heart disease. In a sensitivity analysis restricted to mitral valve or unspecified IE, the findings remained consistent. Among 188 patients with concomitant MAC ≥300 AU and AVC ≥300 AU, three developed IE. Forest plot of Cox regression analysis comparing mitral annulus calcification 1–299 vs. ≥300 for infective endocarditis, stratified by baseline characteristics. Hazard ratios from subgroup analysis comparing patients with mitral annulus calcification ≥300 to those with mitral annulus calcification 1–299, stratified by clinical characteristics. Abbreviations: MAC, mitral annulus calcification, eGFR, estimated glomerular filtration rate, HR, hazard ratio, AU, arbitrary unit. Proportional hazard assumptions were satisfied. In an additional sensitivity analysis using a widely applied registry-based approach without event validation,5,21,22 47 events were identified. The 10-year Kaplan–Meier estimate of IE was 2.02% (95%CI 1.05–3.85) in the MAC ≥300 group (log-rank test MAC 1–299 vs. ≥300, P < 0.001), with a consistent HR of 1.03 (95%CI 1.02–1.04). Discussion In this large population-based prospective study, MAC was associated with an increased risk of IE, after accounting for established risk factors such as CIED, diabetes mellitus, and aortic valve calcification. Although the absolute incidence of IE was low, individuals with MAC ≥300 demonstrated a higher risk than that observed for known intermediate risk factors for IE. To our knowledge, this is the first study to examine MAC quantified on non-contrast CT in relation to IE risk, offering hypothesis-generating insights that may inform future approaches to substrate-based risk stratification. Guidelines consistently emphasize the importance of identifying patients at risk of IE and implementing targeted preventive measures. These recommendations encompass a broad spectrum of practices, including maintaining proper hygiene (oral, skin, and wound care) and immediate curative antibiotic treatment for bacterial infections. For high-risk patients, recommendations include antibiotic prophylaxis before dental and certain invasive procedures.23,24 Despite these recommendations, it remains unclear which patients qualify as high risk, particularly intermediate risk. High-risk patients include those with previous IE, prosthetic valves, and complex congenital heart disease, while intermediate-risk patients include those with rheumatic heart disease, degenerative valve disease, bicuspid aortic valve disease, CIED, and hypertrophic cardiomyopathy.5,24 Incidence rates of IE range from 10 to 120 cases per 10 000 person-years for high-risk patients and up to 12 cases per 10 000 person-years for intermediate-risk patients. However, these classifications are derived primarily from retrospective data, which inherently carries certain limitations.24 The prevalence of MAC in our cohort was very much comparable to other population-based studies.25 Our findings provide novel insights, demonstrating that MAC ≥300 is associated with an incidence rate of 23 per 10 000 cases or a 1.1% 5-year risk of IE, similar to the rates observed in high-risk groups such as complex congenital heart disease21 and the 1.8% 7-year risk reported in patients with small ventricular septal defects.26 Although patients with higher MAC burden have increased mortality, potentially reducing the likelihood of observing IE events, the association with IE remained evident, suggesting that competing risk of death does not fully explain the observed relationship. Notably, no IE events were observed among patients with CIED and No MAC, suggesting that MAC may contribute substantially to the observed risk; however, given the low number of events, this finding should be interpreted with caution. Contemporary imaging literature indicates that MAC severity encompasses both calcium burden and morphological characteristics.6 The Agatston method quantified overall calcium burden but did not capture features such as extension into adjacent structures, which may be relevant to MAC as a local substrate for IE. Multimodality imaging data were unavailable, precluding application of the proposed grading system. Although our data were insufficient to fully explore potential additive effects of MAC and aortic valve calcification, the higher incidence observed for MAC ≥300 underscores its importance in IE risk stratification. Interestingly, few patients had concomitant extensive calcification of the aortic and mitral valves, emphasizing that while these two processes share some characteristics they are distinct and associated with different comorbidities.27 While aortic valve calcification progression is associated with atherosclerosis risk factors such as dyslipidaemia and hypertension,28 the risk of MAC is complex, with many potential mechanisms, ranging from conditions that increase LV systolic pressure, sequelae after chest irradiation therapy, to conditions that lead to hypercalcemia and metastatic calcification of the valve.29 Thus, MAC is common in many diseases that may increase the risk of IE, including mitral annular prolapse disease, hypertrophic cardiomyopathy, and chronic kidney disease. It is noteworthy that the association between MAC and IE remained consistent across most subgroups, but not in patients with chronic kidney disease, in whom MAC is more prevalent.30 This may be explained by the low number of events. The cumulative burden of traditional cardiovascular risk factors such as smoking or hypertension appears not to drive the association, as coronary artery calcification itself was not independently associated with IE and the increased risk was confined to patients with concomitant high coronary artery calcification (≥400 AU) and MAC, rather than elevated coronary artery calcification alone. We previously hypothesized that aortic valve calcification may serve as a nidus for bacterial colonization, leading to an increased risk of IE through surface alterations that facilitate bacterial adherence.4 This mechanism may also apply to MAC, given the similar calcific pathology. Calcium deposition in the mitral valve annulus is known to be associated with increased cardiovascular complications,30 including a potential role in IE,6,7 though insights into the mechanisms were limited. Moreover, established risk factors for IE, such as age-related valvular degeneration,31 male sex,32 or bicuspid aortic valve,33 are closely associated with increased calcification, further supporting this hypothesis. Haemodynamic changes caused by conditions like bicuspid aortic valve or aortic regurgitation may also lead to endothelial injury, as previously suggested.34 The most common bacterial pathogens were Staphylococcus aureus and Streptococcal species, corroborating previous findings,6,7 further supporting the theory of calcification as a nidus for these organisms. Our findings suggest that calcification may serve as a bacterial nidus for bacterial colonization, offering new mechanistic insights into IE risk stratification and potential preventive strategies, which warrant further investigation in a larger population. These findings offer novel insights into IE risk stratification and suggest that calcification mitral annulus may serve as a key marker for identifying high risk patients. The proposed mechanistic link provides important insights for potential prophylactic and therapeutic strategies, warranting further investigation in future research. Limitations This study is a prospective cohort study based on a large population-based, randomized, cardiovascular screening and intervention trial. However, several limitations persist. The most important is the very low number of validated IE events, including only 32 cases overall and 8 events among individuals with MAC ≥300 AU. Although not all IE cases involved the mitral valve, the findings remain clinically relevant, as IE is a complex systemic disease in which the diagnosis may be established according to Duke criteria even in the absence of definitive echocardiographic findings. This is particularly relevant in the setting of severe MAC, where vegetations may be difficult to identify in an everyday clinical setting, introducing uncertainty in determining the exact anatomical focus. In a post hoc assessment using Freedman’s method, the study had 80% power only to detect an HR of approximately 15.6 or greater given the 32 observed IE events and an exposure prevalence of 3.4%, so more modest associations cannot be excluded. Although the consistency of findings across analytical approaches provides some support for the observed associations, these results should be interpreted as hypothesis generating. The Danish clinical databases are regarded as high quality due to several synergistic elements and because treatment of IE is prolonged and centrally managed, it is unlikely that any cases of IE were not captured.15 Previous studies have demonstrated a general high level of confidence in the positive predicted values of cardiovascular diagnostic codes in Denmark and the DNPR and internationally recognized to be of very high quality.17,35 Furthermore, each possible IE event underwent a complete medical record review by two physicians independently, which enhances the validity and enables the unique ability to accurately identify and present the infection site. Although we do not report Duke criteria for individual cases, event adjudication followed Danish clinical practice, which applies current ESC guidelines and the modified Duke framework, ensuring high diagnostic accuracy while maintaining consistency with the registry-based methods that underpin current estimates of high and intermediate IE risk.5,21,22 Second, this trial almost exclusively included men, so the significant sex imbalance in the cohort may restrict the generalizability of the findings. External validation in an independent cohort with a more balanced sex distribution would be valuable to confirm the robustness and generalizability of the findings. Given the higher incidence of MAC in women, the findings may retain applicability to some extent in women. Third, the Kaplan-Meier method may overestimate the incidence of IE, as it does not account for competing risks such as mortality. While the methodology aligns with previous studies,5 we also analysed a combined endpoint of IE and all-cause mortality to address this limitation. Fourth, inter-reader reproducibility data for MAC quantification were not available. Fifth, mitral leaflet calcification was not quantified and may represent an additional substrate for IE. The present findings therefore pertain specifically to annular calcification. Conclusion MAC appears to be associated with an increased risk of IE beyond established conditions such as CIED, diabetes mellitus, and aortic valve calcification. The risk increased with higher MAC burden, and the risk among participants with MAC ≥300 AU approached that observed in high-risk populations. However, given the limited number of IE events, these findings should be interpreted with caution. MAC ≥300 may represent a potential marker for identifying patients at elevated risk of IE, although these results are hypothesis-generating and warrant confirmation in larger studies. Supplementary Material Acknowledgements Data management and statistical advice was provided and REDCap was hosted by OPEN, Open Patient data Explorative Network, Odense University Hospital, Region of Southern Denmark. During the preparation of this manuscript the authors occasionally used AI based tool (ChatGPT) for grammar and phrasing support. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article. Contributor Information Lytfi Krasniqi, Department of Cardiac, Thoracic and Vascular Surgery, Odense University Hospital, J. B. Winsløwsvej 4, Odense 5000, Denmark; Department of Clinical Research, University of Southern Denmark, Campusvej 55, Odense 5000, Denmark. Axel C P Diederichsen, Department of Clinical Research, University of Southern Denmark, Campusvej 55, Odense 5000, Denmark; Department of Cardiology, Odense University Hospital, J.B. Winsløws Vej 4, Odense 5000, Denmark. Jes S Lindholt, Department of Cardiac, Thoracic and Vascular Surgery, Odense University Hospital, J. B. Winsløwsvej 4, Odense 5000, Denmark; Department of Clinical Research, University of Southern Denmark, Campusvej 55, Odense 5000, Denmark; Elite Centre for Individualized Medicine in Arterial Disease (CIMA), Odense University Hospital, J.B. Winsløws Vej 4, Odense 5000, Denmark. Camilla Engelsgaard, Department of Clinical Research, University of Southern Denmark, Campusvej 55, Odense 5000, Denmark; Department of Cardiology, Odense University Hospital, J.B. Winsløws Vej 4, Odense 5000, Denmark. Martin Busk, Department of Cardiology, Vejle Hospital, Beriderbakken 4, Vejle 7100, Denmark. Lars Frost, Department of Cardiology, Diagnostic Centre, Regional Hospital Silkeborg, Falkevej 1D, Silkeborg 8600, Denmark. Jess Lambrechtsen, Department of Cardiology, Odense University Hospital, Baagøes Alle 15, Svendborg 5700, Denmark. Marek Karon, Department of Medicine, Nykoebing Falster Hospital, Fjordvej 15, Nykoebing Falster 4800, Denmark. Jordi Sanchez Dahl, Department of Clinical Research, University of Southern Denmark, Campusvej 55, Odense 5000, Denmark; Department of Cardiology, Odense University Hospital, J.B. Winsløws Vej 4, Odense 5000, Denmark; Department of Cardiovascular Medicine, Mayo Clinic, 200 1st St SW, Rochester, MN 55901, USA. Data availability Data from this study, including de-identified participant data and a comprehensive data dictionary, can be made available. Data will be shared whose proposed use has been approved from the appropriate national legal regulations, primarily for replication of our results. A signed data access agreement, compliant with regional legislation and data authority requirements, is required prior to data release. To request access, contact Lytfi.Krasniqi@rsyd.dk. Supplementary material Supplementary material is available at European Heart Journal Open online. Author contributions Lytfi Krasniqi (Conceptualization [equal], Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Writing—original draft, Writing—review & editing [lead]), Axel Cp Diederichsen (Conceptualization [equal], Data curation, Methodology, Resources, Visualization, Writing—review & editing [supporting]), Jes S. Lindholt (Funding acquisition, Methodology, Resources, Writing—review & editing [supporting]), Camilla Engelsgaard (Methodology, Writing—review & editing [supporting]), Martin Busk (Methodology, Writing—review & editing [supporting]), Lars Frost (Methodology, Writing—review & editing [supporting]), Jess Lambrechtsen (Methodology, Writing—review & editing [supporting]), Marek Karon (Methodology, Writing—review & editing [supporting]), and Jordi Sanchez Dahl (Conceptualization, Writing—original draft [equal], Data curation, Methodology, Validation, Writing—review & editing [supporting], Supervision [lead]) Funding No specific grant from any funding agency in the public, commercial, or not-for-profit sectors was received for this substudy of DANCAVAS. References Associated Data Supplementary Materials Data Availability Statement The data supporting the findings of this study are available from the corresponding author upon reasonable request, subject to approval under applicable Danish legal regulations, and a compliant data access agreement prior to release. Data from this study, including de-identified participant data and a comprehensive data dictionary, can be made available. Data will be shared whose proposed use has been approved from the appropriate national legal regulations, primarily for replication of our results. A signed data access agreement, compliant with regional legislation and data authority requirements, is required prior to data release. To request access, contact Lytfi.Krasniqi@rsyd.dk.