Change search
ReferencesLink to record
Permanent link

Direct link
Diabetes and Hypertension Consistently Predict the Presence and Extent of Coronary Artery Calcification in Symptomatic Patients: A Systematic Review and Meta-Analysis
Umeå University, Faculty of Medicine, Department of Public Health and Clinical Medicine.
Umeå University, Faculty of Medicine, Department of Public Health and Clinical Medicine.
Umeå University, Faculty of Medicine, Department of Medical Biosciences.
Show others and affiliations
2016 (English)In: International Journal of Molecular Sciences, ISSN 1422-0067, E-ISSN 1422-0067Article in journal (Refereed) Accepted
Abstract [en]

Background: The relationship of conventional cardiovascular risk factors (age, gender, ethnicity, diabetes, dyslipidaemia, hypertension, obesity, exercise, and the number of risk factors) to coronary artery calcification (CAC) presence and extent has never before been assessed in a systematic review and meta-analysis.

Methods: We included only English language studies that assessed at least three conventional risk factors apart from age, gender, and ethnicity, but excluded studies in which all patients had another confirmed condition such as renal disease.

Results: In total, 10 studies, comprising 15,769 patients, were investigated in the systematic review and seven studies, comprising 12,682 patients, were included in the meta-analysis, which demonstrated the importance of diabetes and hypertension as predictors of CAC presence and extent, with age also predicting CAC presence. Male gender, dyslipidaemia, family history of coronary artery disease, obesity, and smoking were overall not predictive of either CAC presence or extent, despite dyslipidaemia being a key risk factor for coronary artery disease (CAD).

Conclusion: Diabetes and hypertension consistently predict the presence and extent of CAC in symptomatic patients.


Place, publisher, year, edition, pages
Basel, Switzerland: MDPI , 2016.
Keyword [en]
meta-analysis, systematic review, coronary calcification, risk factors
National Category
Cardiac and Cardiovascular Systems
Research subject
URN: urn:nbn:se:umu:diva-124923OAI: diva2:956604
Available from: 2016-08-30 Created: 2016-08-30 Last updated: 2016-08-31
In thesis
1. Insights into the relationship between coronary calcification and atherosclerosis risk factors
Open this publication in new window or tab >>Insights into the relationship between coronary calcification and atherosclerosis risk factors
2016 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]


Coronary artery disease (CAD) is the most common cause of death in Europe and North America and early detection of atherosclerosis is a clinical priority. Diagnosis of CAD remains conventional angiography, although recent technology has introduced non-invasive imaging of coronary arteries using computed tomographic coronary angiography (CTCA), which enables the detection and quantification of coronary artery calcification (CAC). CAC forms within the arterial wall and is usually found in or adjacent to atherosclerotic plaques and is consequently known as sub-clinical atherosclerosis.

 The conventional cardiovascular (CV) risk factors used to quantify the estimated 10-year coronary event risk comprise dyslipidaemia, hypertension, diabetes mellitus, obesity, smoking and family history of CAD. Nevertheless, their relationship with significant (>50%) stenosis, their interaction with the CAC score and their predictive ability for CAC presence and extent has not been fully determined in symptomatic patients.



 For Papers 1-4 we took patients from the Euro-CCAD cohort, an international study established in 2009 in Umeå, Sweden. The study data gave us the CAC score and the CV risk factor profile in 6309 patients, together with angiography results for a reduced cohort of 5515 patients. In Papers 1 and 2 we assessed the risk factors for significant stenosis, including CAC as a risk factor. Paper 2 carried out this analysis by geographical region: Europe vs USA and northern vs southern Europe. Paper 3 investigated the CV risk factors for CAC presence, stratified by age and gender, while Paper 4 assessed the CV risk factors for CAC extent, stratified by gender.

 In paper 5 we carried out a systematic review and meta-analysis of all studies of the risk factor predictors of CAC presence, extent and progression in symptomatic patients. From a total of 884 studies, we identified 10 which fitted our inclusion criteria, providing us with a total of 15,769 symptomatic patients. All 10 were entered in the systematic review and 7 were also eligible for the meta-analysis.



Paper 1:           Among risk factors alone, the most powerful predictors of significant coronary stenosis were male gender followed by diabetes, smoking, hypercholesterolaemia, hypertension, family history of CAD and age; only obesity was not predictive. When including the log transformed CAC score as a risk factor, this proved the most powerful predictor of >50% stenosis, but hypercholesterolaemia and hypertension lost their predictive ability. The conventional risk factors alone were 70% accurate in predicting significant stenosis, the log transformed CAC score alone was 82% accurate but the combination was 84% accurate and improved both sensitivity and specificity.

 Paper 2:           Despite some striking differences in profiles between Europe and the USA, the most important risk factors for >50% stenosis in both groups were male gender followed by diabetes. When the log CAC score was included as a risk factor, it became by far the most important predictor of >50% stenosis in both continents, followed by male gender. In the northern vs southern Europe comparison the result was similar, with the log CAC score being the most important predictor of >50% stenosis in both regions, followed by male gender.

 Paper 3:           Independent predictors of CAC presence in males and females were age, dyslipidaemia, hypertension, diabetes and smoking, with the addition of family history of CAD in males; obesity was not predictive in either gender. The most important predictors of CAC presence in males were dyslipidaemia and diabetes, while among females the most important predictors of CAC presence were diabetes followed by smoking. When analysed by age groups, in both males and females aged <70 years, diabetes, hypertension and dyslipidaemia were predictive, with diabetes being the strongest; in females aged <70 years, smoking was also predictive. Among those aged ≥70 years, the results are completely different, with only dyslipidaemia being predictive in males but smoking and diabetes were predictive in females.

 Paper 4:           In the total cohort, age, male gender, diabetes, obesity, family history of CAD and number of risk factors predicted an increasing CAC score, with the most important being male gender and diabetes. In males, hypertension and dyslipidaemia were also predictive, although diabetes was the most important predictor. Diabetes was similarly the most important risk factor in females, followed by age and number of risk factors. Among patients with CAC, hypertension, dyslipidaemia and diabetes predicted CAC extent in both males and females, with diabetes being the strongest predictor in males followed by dyslipidaemia, while diabetes was also the strongest predictor in females, followed by hypertension. Quantile regression confirmed the consistent predictive ability of diabetes.

 Paper 5:           In the systematic review, age was strongly predictive of both CAC presence and extent but not of CAC progression. The results for CAC presence were overwhelmed by data from one study of almost 10,000 patients, which found that white ethnicity, diabetes, hypertension and obesity were predictive of CAC presence but not male gender, dyslipidaemia, family history or smoking. With respect to CAC extent, only male gender and hypertension were clearly predictive, while in the one study of CAC progression, only diabetes and hypertension were predictive. In the meta-analysis, hypertension followed by male gender, diabetes and age were predictive of CAC presence, while for CAC extent mild-moderate CAC was predicted by hypertension alone, whereas severe CAC was predicted by hypertension followed by diabetes.



Our investigation of the Euro-CCAD cohort showed that the CAC score is far more predictive of significant stenosis than risk factors alone, followed by male gender and diabetes, and there was little benefit to risk factor assessment over and above the CAC score for >50% stenosis prediction. Regional variations made little difference to this result. Independent predictors of CAC presence were dyslipidaemia and diabetes in males and diabetes followed by smoking in females. The risk factor predictors alter at age 70. The most important risk factor predictors of CAC extent were male gender and diabetes; when analysed by gender, diabetes was the most important in both males and females. Our studies have consistently shown the strong predictive ability of male gender in the total cohort and diabetes in males and females and this is reflected in the meta-analysis, which also found hypertension to be independently predictive. Interestingly, dyslipidaemia does not appear to be a strong risk factor.


Place, publisher, year, edition, pages
Umeå: Umeå University, 2016. 84 p.
Umeå University medical dissertations, ISSN 0346-6612 ; 1830
Coronary artery calcification, risk factors, atherosclerosis, stenosis
National Category
Cardiac and Cardiovascular Systems Public Health, Global Health, Social Medicine and Epidemiology
Research subject
urn:nbn:se:umu:diva-124909 (URN)978-91-7601-535-3 (ISBN)
External cooperation:
Public defence
2016-09-22, Hörsal D, UnodT9, Norrlands universitetssjukhus, Umeå, 09:00 (English)
Available from: 2016-09-01 Created: 2016-08-30 Last updated: 2016-09-09Bibliographically approved

Open Access in DiVA

No full text

Search in DiVA

By author/editor
Nicoll, RachelIbrahimi, PranveraOlivecrona, GunillaHenein, Michael
By organisation
Department of Public Health and Clinical MedicineDepartment of Medical Biosciences
In the same journal
International Journal of Molecular Sciences
Cardiac and Cardiovascular Systems

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 10 hits
ReferencesLink to record
Permanent link

Direct link