Presence and significance of aortic valve calcification detected by cardiac CT - a population-based study from the Danish Cardiovascular Screening Trial (DANCAVAS)

Lida Khurrami

Research output: ThesisPh.D. thesis

9 Downloads (Pure)


Aortic Valve Stenosis (AS) is a growing challenge in developed countries with an increasing prevalence in the elderly population. Already in the prolonged and asymptomatic initial phase of AS, irreversible cardiac changes as left atrial (LA) dilation and left ventricle (LV) remodelling, have been observed. Symptomatic AS is associated with significant mortality and morbidity. Thus timely diagnosis and, if possible, prevention or treatment is crucial. Non-contrast cardiac computed tomography (NCCT) offers precise and reproducible detection of aortic valve calcification (AVC), quantified in Agatston units (AU). The AVC score is directly correlated with the severity of AS and overall mortality in patients with established valvular disease. However, less is known about the impact of an elevated AVC score in the general population. This thesis draws on three papers, based on the data from a pilot study on approximately 2000 Danish males and females and the population-based Danish Cardiovascular Screen Trial (DANCAVAS I and II), where 15,000 random individuals (aged 60-74 years) underwent a cardiovascular (CV) screening including NCCT. The thesis evaluates the prevalence of AVC in the study population and the association of AVC to CV risk factors, transthoracic echocardiographic (TTE) characteristics, and outcomes as aortic valve replacement (AVR) and all-cause mortality.
Paper I. Method: A random sample of 2060 females and males aged 65-74 years was invited to the DANCAVAS pilot study. Participants with AVC score ≥ 300 AU were invited for a TTE together with age-matched controls with AVC<300. Aims: To determine the presence and extent of AVC quantified by NCCT, determine the association between traditional CV risk factors and AVC score, and evaluate the association between AVC and cardiac size and function assessed by TTE. Results: Of 2060 invited individuals, 664 males and 636 females participated, mean age 68.9 (±2.7). Among those, 455 (68.5%) of males and 358 (56.3%) of females had AVC scores >0 AU. In a multiple regression analysis, age, sex, prior CV disease, smoking, and hypertension were associated with AVC scores. Individuals with AVC ≥300 AU had a higher peak and mean aortic valve gradient, smaller indexed aortic valve area (AVAi), greater left ventricular mass, and larger left atrial volume.
Paper II. Method: All male participants from the DANCAVAS trial with AVC and coronary artery calcifications (CAC) scores were included. CV risk factors were self-reported, measured, or both, and further explored using descriptive and regression analyses for association with AVC. Aims: To assess the association between CV risk factors, including CAC and biomarkers and AVC, in a large population of men aged 60-74 years. Results: 14,073 participants were included. The AVC scores ranged from 0 to 9067 AU with a median AVC value of 6 AU (IQR 0-82) and 215 (1.5%) of participants having an AVC score ≥ 1200. In the regression analyses, all CV risk factors were associated with AVC. However, after adjusting for CAC≥400, only age, hypertension, obesity, known CV disease, and serum phosphate remained significantly associated, while smoking, diabetes, hyperlipidaemia, eGFR, and serum calcium were not.
Paper III. Method: All DANCAVAS participants with AVC≥ 300 AU were invited for a supplementary examination, including cardiac auscultation and an echocardiogram. The final study population consisted of 831 men. Using Danish national regulatory registries, patients were followed for either all-cause mortality or AVR. Descriptive statistics, ROC analysis, and cox regression analysis were performed. Aim: To describe the association between AVC score and presence of cardiac murmur, echocardiographic signs of LV concentric and LA remodelling, AS severity, and outcomes as aortic valve replacement and all-cause mortality, in a randomly selected population of elderly men with AVC ≥300 AU. Results: 1104 (11%) had AVC≥300 AU and met the inclusion criteria for the study. Of those, 831 had sufficient TTE. Systolic murmur and visual finding of aortic valve sclerosis on the echocardiogram were closely associated with the presence of AS and AVC. There was a clear linear association between AVC and AVAi, and ROC analyses suggested a good association between AVC and AS with area under the curve of 0.86 for mild AS, 0.93 for moderate AS, and 0.96 for severe AS. At five years follow up, 63 patients died, and 18 underwent AVR. AVC was not an independent predictor of mortality, but AVC>1200 was associated with AVR.
Conclusion: In a random population sample of individuals aged 60–75 years, AVC was common with a prevalence of nearly 60%, mostly with AVC values < 300 AU. AVC was significantly associated with most known CV risk factors, concentric remodelling, and LA dilatation in a mixed population of males and females aged 65-74 years. Further exploration of these risk factors in a larger population of 60–74-year-old males demonstrated that AVC was associated with all modifiable CV risk factors, while in an adjusted model including CAC score ≥400, only age hypertension, obesity, phosphate, and known CV disease remained significant. In a selected population of 10,471 Danish males, 65–74-years of age, 11% had AVC values ≥ 300 AU. AVC was associated with aortic valve area, aortic valve flow velocity, and LA dilatation but not with LV remodelling, and neither was it an independent predictor of all-cause mortality in this population.
Original languageEnglish
Awarding Institution
  • University of Southern Denmark
  • Diederichsen, Axel, Principal supervisor
  • Møller, Jacob Eifer, Co-supervisor
  • Lindholt, Jes Sanddal, Co-supervisor
  • Sandgaard, Niels, Co-supervisor
Publication statusPublished - 13. Jan 2022

Note re. dissertation

Print copy of the full thesis is restricted to reference use in the Library.


Dive into the research topics of 'Presence and significance of aortic valve calcification detected by cardiac CT - a population-based study from the Danish Cardiovascular Screening Trial (DANCAVAS)'. Together they form a unique fingerprint.

Cite this