Vascular Medicine 2006; 11: 232–238

Renal artery calcium: relationship to systemic calcified atherosclerosis Matthew A Allisona,b, Dominic DiTomassoc, Michael H Criquia,b, Robert D Langera,d and C Michael Wrightb Abstract: In this study we tested the hypothesis that calcium due to atherosclerosis in the renal arteries would be significantly associated with calcium in multiple other vascular beds, independent of traditional risk factors for cardiovascular disease (CVD). Electron beam computed tomography was conducted in 1461 consecutive asymptomatic patients free of clinical CVD to determine the presence and extent of calcium in the renal arteries, coronary and non-coronary vasculature and the aortic and mitral annuli. The overall prevalence for calcium in either renal artery was 18.0%, with men having a significantly higher prevalence (20.2%) than women (15.0%) [p ⫽ 0.01]. Renal artery calcium (RAC) was significantly correlated with calcium located in the carotids, coronaries, thoracic aorta, abdominal aorta and iliac arteries and calcium in the mitral and aortic annuli (r range ⫽ 0.22–0.37). In a multivariable model containing the traditional CVD risk factors, the presence of calcium in the renal arteries was significantly associated with age, male sex and a diagnosis of hypertension. After adjustment for these variables, the presence of calcium in the thoracic or abdominal aorta was significantly associated with RAC (OR ⫽ 2.1 and 2.0, respectively; p ⬍ 0.01 for both). The sensitivity for prevalent RAC was highest in those individuals with any calcium in the abdominal aorta (94.5%). In conclusion, calcium related to atherosclerosis in the renal arteries is highly associated with atherosclerotic calcification in other vascular beds, especially the aorta, and the valvular annuli. These relationships are independent of traditional CVD risk factors. Key words: atherosclerosis; calcium; renal arteries; risk factors

Introduction Atherosclerosis is a systemic1 inflammatory process2,3 that is typically initiated at arterial branch points.4 Calcium is actively deposited in the lipid core of atherosclerotic plaques via an organized and regulated process5–7 similar to cortical bone formation.8,9 Calcified vascular lesions are readily detectable by contemporary technologies, including spiral and electron beam computed tomography (EBCT). Previous studies using EBCT have demonstrated significant

Departments of aFamily and Preventive Medicine, bMedicine and cBioengineering, University of California, San Diego, La Jolla, CA, USA; dGeisinger Health System, Danville, PA, USA Address for correspondence: Matthew A Allison, 3855 Health Sciences Drive, MC 0817, La Jolla, CA 92093-0817, USA. Tel: ⫹1 858 822 3585; Fax: ⫹1 858 822 3797; E-mail: [email protected] © 2006 SAGE Publications

associations between calcified atherosclerosis in the carotids, coronaries, thoracic aorta, abdominal aorta and iliac arteries,10 as well as the aortic and mitral annuli.11 These associations are present even with adjustment for cardiovascular risk factors. The accumulation of atherosclerosis in the renal arteries resulting in significant stenosis (⬎50%) is a known risk factor for secondary hypertension. The preponderance of studies on atherosclerosis in these arteries have utilized clinical populations with angiography as the diagnostic test.12–15 However, angiography is an insensitive measure of the extent of atherosclerotic burden. The present study was conducted in a large cohort of subjects free of clinical cardiovascular disease (CVD) with the aims of determining (1) the prevalence and extent of calcium related to atherosclerosis in the renal arteries (RAC) and (2) whether RAC was significantly associated with atherosclerotic calcification in other vascular areas. We hypothesized that calcium in the vascular 10.1177/1358863x06073449

Systemic atherosclerosis and renal artery calcium

beds and valvular annuli would be independently associated with RAC even after adjustment for traditional CVD risk factors. Methods Subjects From February 1, 2001 to June 29, 2001, 1461 consecutive patients presented for preventive medicine services at a university affiliated disease prevention center in San Diego, California, and were evaluated for the extent of calcified atherosclerosis in five different vascular beds: the carotid, coronary, thoracic aorta, abdominal aorta and iliac vessels. Most subjects were self-referred or referred on the advice of their primary care provider. Patients for this analysis were free of clinical CVD (angina, myocardial infarction, stroke, transient ischemic attack [TIA], coronary revascularization [CABG, PTCA or stent]) or carotid artery surgery. Participants completed a detailed health history questionnaire that collected information on history of hypertension, diabetes, high cholesterol, smoking, medications, family history of coronary heart disease, diet, exercise and prior surgeries. The study protocol was approved by the Human Research Protection Program at the University of California at San Diego, which granted a waiver of informed consent. Imaging Imaging was conducted using an Imatron C-150 scanner (GE, San Francisco, CA, USA). Images for each vascular bed were obtained from a single scan using a 100-ms scan time and preceding caudally from the base of the skull to the symphysis pubis. Each bed was obtained by a distinct scan of the segment in question employing the following slice thicknesses: 3 mm for the coronary bed, 6 mm through the neck, abdomen and pelvis, and 5 mm for the thorax. Cardiac tomographic imaging was electrocardiographically triggered at 40% or 65% of the R-R interval, depending on the subject’s heart rate. During separate breathholds at half-maximal inspiration, imaging of the heart, thorax and abdomen was conducted. Atherosclerotic calcification was defined as a plaque area ⱖ1 mm2 with a density of greater than or equal to 130 Hounsfield units (HU). Quantitative calcium scores were determined according to the method described by Agatston et al.16 In brief, the calcium score per lesion was calculated by multiplying the area of the contiguous pixels by the corresponding density number using the following scale for density (1 ⫽ 130–199 HU, 2 ⫽ 200–299 HU, 3 ⫽ 300–399 HU, 4 ⫽ ⱖ400 HU). The total coronary calcium score was then determined by adding up all of the lesion scores from all of the slices. Agatston calcium scores for vascular beds other than the coronaries were adjusted for slice thickness using the following formula: adjusted Vascular Medicine 2006; 11: 232–238

233

score ⫽ original score * slice thickness/3.0. Volume averaging was avoided by scoring each homogeneous slice thickness segment separately.10 The coronary calcium score consisted of calcified lesions in the left main, left anterior descending, left circumflex and right coronary arteries. Data from the left and right sides were combined to give the extent of calcium in the carotid and iliac beds. The thoracic aorta was defined as the segment from the aortic root to the diaphragm while the abdominal aorta was the segment from the diaphragm to the iliac bifurcation. The extent of calcium in the aortic and mitral annuli was determined by using chest CT images obtained for the coronary arteries. Mitral annular calcification was located at the junction between the left atria and left ventricle. Owing to their proximity, care was taken to discriminate between the left circumflex coronary artery and the mitral annulus. Calcium in the aortic annulus was defined as calcium detected just inferior to the origin of the coronary arteries and located at the point of attachment of the aortic valve.11 Using the Agatston method for calcium scoring and the same software used for the vascular beds and valvular annuli, image files of the abdomen were retrospectively interrogated for the presence and extent of calcium in both renal arteries. During scoring, calcium in these arteries was categorized as arising from the ostia or arterial segment proper. Calcium in the wall of the abdominal aorta was not included in the assessment for RAC by excluding any visible calcium in the extrapolated vertical plane of the aorta running through the renal artery. The total amount of calcium in the renal arteries was calculated as the sum of the ostial and arterial segments. The reader for renal artery calcium only viewed images that contained the renal arteries and this individual was blinded to the scores for the other beds. Laboratory Casual serum lipid and glucose measurements were obtained by fingerstick using the Cholestec LDX® system. All subjects had blood pressure measured while in the seated position and after resting for 5 minutes. Weight was assessed with the patient lightly clothed and without shoes. Body fat measurement was conducted using bioimpedence on the Omron™ HBF300. Body mass index was calculated as the weight (kg) divided by the height (m2). Statistical analysis Hypertension was defined as systolic or diastolic blood pressure ⬎140 mmHg or ⬎90 mmHg, respectively, or current use of anti-hypertensive medication combined with a physician diagnosis of this condition. Individuals with a total cholesterol to high-density lipoprotein (HDL) ratio greater than 5 or who reported using a medication for this condition were classified

234

MA Allison et al

as dyslipidemic. Diabetes was defined by current use of anti-glycemic medications or a serum glucose ⬎200 mg/dl. Smoking status was defined as current, former or never. Tests for group differences for those with and without any renal calcium were conducted using ANOVA or the Kruskal–Wallis test (as appropriate) for continuous variables and the chi-squared test for categorical variables. Since it is well-known that calcium distributions differ by age and gender, mean values by group were also adjusted for these factors using ANCOVA. Simple and partial correlations were calculated using the Spearman rank correlation procedure. The potential association between the traditional risk factors for CVD and the presence of RAC was explored by multivariable logistic regression first adjusted for age and sex, and then adjusted for all CVD risk factors (ie age, sex, BMI, smoking, diabetes, hypertension, dyslipidemia).17 The individual association between any RAC and calcification in each vascular bed and valvular annulus was tested using logistic regression. These associations were analyzed using sequential models that were (1) unadjusted, (2) adjusted for age and sex, and (3) adjusted for the CVD risk factors. Finally, to determine which vascular bed or valvular annulus was significantly and independently associated with the presence of RAC after adjusting for all CVD risk factors and the extent of calcium in the other beds or annuli, we constructed two multivariable logistic regression models: (1) one that contained all of the vascular beds and (2) one that contained both annuli. Both of these models included all of the CVD risk factor variables as covariates.

Table 1

For each vascular bed and annulus, the sensitivity, specificity, and positive and negative predictive values for the presence and absence of RAC were calculated using 2 ⫻ 2 tables. Receiver operator characteristic curves were constructed to determine the ‘ability’ of each vascular bed to discriminate the presence/ absence of any RAC. These results are expressed as an area under the curve (AUC). The level of significance for this study was 0.05 (two-tailed). All analyses were conducted using SPSS version 12 (Chicago, IL, USA). Results The characteristics of the study cohort stratified by the presence or absence of renal arterial calcium are presented in Table 1. The overall prevalence of calcium in either renal artery was 18.0% (263/1464), with men having a significantly higher prevalence (20.2%, 168/831) than women (15.0%, 95/633) [p ⫽ 0.01]. Similarly, the prevalence of calcium in the renal ostia was higher (15.6%) than that in the arterial segments proper (10.0%), and the prevalence in the left was higher than in the right for both of these segments (left os: 12.2%, left artery: 7.0%; right os: 10.7%, right artery: 5.5%). Owing to the high number of subjects with no RAC, the distributions of these variables were highly skewed. Accordingly, the median calcium score for RAC was 0 in all segments with a minimum of 0 and a maximum of 779 (interquartile range: 0–0). With adjustment for age and gender, the presence of RAC was associated with significantly higher prevalences of diabetes (4.0 vs 2.3%), dyslipidemia

Cohort characteristics stratified by the presence of renal artery calcium.

Variable

RAC ⬎ 0

RAC ⫽ 0

p-value

Agea Femalea BMIa Total body fata Former smokerb Current smokerb Diabetes mellitusb Dyslipidemiab Hypertensionb Family history of CHDb Carotid calcium scorec Coronary calcium scorec Thoracic aorta calcium scorec Abdominal aorta calcium scorec Iliac calcium scorec Aortic annulus calcium score Mitral annulus calcium scored

72.4 95 26.8 30.5 137 23 11 90 107 72 93 258 537 1605 1171 39 0

58.6 538 26.8 28.6 419 103 25 333 271 297 0 1.81 0 0 0 0 0

⬍0.01 ⬍0.01 0.79 ⬍0.01 ⬍0.01 0.93 0.05 0.04 ⬍0.01 0.37 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01

(8.8) (36.1) (4.4) (8.9) (52.1) (8.7) (4.2) (34.2) (40.7) (27.4) (0–389) (40–809) (177–2486) (894–6044) (158–3158) (0–186) (0–0)

(10.0) (44.8) (4.6) (8.4) (34.9) (8.6) (2.1) (27.7) (22.6) (24.7) (0–0) (0–51) (0–66) (0–359) (0–210) (0–0) (0–0)

RAC, renal artery calcium. (SD); bcount (%); cmedian (interquartile range); dalthough the median values are the same, the distribution of mitral annular calcium is significantly different for these two groups. aMean

Vascular Medicine 2006; 11: 232–238

Systemic atherosclerosis and renal artery calcium

(33.5 vs 28.5%), hypertension (36.8 vs 26.5%) and ever smoking (55.5 vs 48.8%) [p ⬍ 0.01 for all]. The distributions of the calcium scores in all vascular beds assessed, and the valvular annuli, were also significantly different in those with any RAC after this same adjustment. On the other hand, family history of coronary heart disease (CHD) was of borderline significance (29.1 vs 23.0%, p ⫽ 0.06), while BMI and percent body fat were essentially equivalent in those with and without RAC. The age and gender-adjusted Spearman rank correlations between the total RAC score and calcium scores in the vascular beds and valvular annuli were statistically significant (p ⬍ 0.01 for all), with the correlations ranging from 0.22 to 0.36. The highest and lowest correlations were between the RAC score and calcium in the thoracic aorta and the mitral annulus, respectively. Notably, while these correlations were nearly identical for correlations between calcium scores in the renal ostia and the beds/annuli (range: 0.22–0.37), the correlations were not as strong for calcium in the renal artery segments (range: 0.14–0.23). For the latter, however, all were statistically significant and the nature of the correlations by vascular area was unchanged. Furthermore, although age was significantly correlated with RAC score (r ⫽ 0.47, p ⬍ 0.01), after adjustment for age and gender, none of the following risk factor variables were so: HDL and LDL (low-density lipoprotein) cholesterols, total to HDL cholesterol ratio, BMI, percent body fat. In separate age- and sex-adjusted logistic regression models for each risk factor variable, ever smoking, hypertension, dyslipidemia and family history of CHD were significantly associated with the presence of any RAC ( p ⬍ 0.05 for all), while diabetes and higher BMI levels were not significant (Table 2). When all of the risk factors were combined in a single multivariable model, age, male sex and hypertension were independently associated with prevalent RAC. The odds

Table 2

for any RAC were highest for those with hypertension (OR ⫽ 1.53) while every 1-year increase in age was associated with 16% higher odds for any RAC and being male was associated with 29% higher odds. In this model, a family history of CHD was of borderline significance with an odds ratio of 1.19 ( p ⫽ 0.06). Using separate logistic regression models, the associations between the presence of calcium in the vascular beds and the valvular annuli with prevalent RAC are shown in Figure 1. With adjustment for age and sex, each of these vascular areas was significantly associated with the presence of RAC, with the highest magnitude of association being found for both the thoracic and abdominal aorta (OR ⫽ 2.2, p ⬍ 0.01). When the models were further adjusted for body mass index, smoking, hypertension, diabetes, dyslipidemia and family history of CHD, all beds/annuli, except the iliacs, remained significantly associated with any RAC. As before, the magnitude of association with prevalent RAC was highest for the thoracic (OR ⫽ 2.1, p ⬍ 0.01) and abdominal (OR ⫽ 2.0, p ⬍ 0.01) aortic segments. In a logistic model including all of the CVD risk factors and all of the vascular beds as independent variables, the presence of calcium in the thoracic aorta was most highly associated with the presence of any RAC (OR ⫽ 1.7, p ⬍ 0.01) followed by calcium in the abdominal aorta (OR ⫽ 1.6, p ⬍ 0.01), carotid arteries (OR ⫽ 1.5, p ⬍ 0.01) and coronary arteries (OR ⫽ 1.4, p ⫽ 0.02), while the iliac arteries were not significantly associated with prevalent RAC (OR ⫽ 0.88, p ⫽ 0.4). In a logistic model including all of the CVD risk factors and both valvular annuli, the odds for prevalent RAC were the same for both the aortic and mitral annuli (OR ⫽ 1.5, p ⬍ 0.01). When all of the vascular beds and both of the valvular annuli were combined into a single logistic regression model that was also adjusted for all of the CVD risk factors, the results were essentially unchanged from that pro-

Multivariable risk factor analysis for renal artery calcium. Separate modelsa

Combined modelb

Risk factor

Odds ratio

95% CI

Odds ratio

95% CI

Age (1-year increase) Sex (male vs female) Ever smoked (yes vs no) Hypertension (yes vs no) Dyslipidemia (yes vs no) Diabetes (yes vs no) BMI (1 unit increase) Family history of CHD (yes vs no)

1.16 1.31 1.17 1.60 1.21 1.35 1.01 1.23

1.13–1.18 1.12–1.54 1.00–1.38 1.30–1.96 1.02–1.44 0.88–2.08 0.97–1.05 1.03–1.47

1.16 1.29 1.12 1.53 1.11 1.29 1.00 1.19

1.14–1.18 1.08–1.54 0.95–1.33 1.23–1.89 0.92–1.33 0.83–2.00 0.96–1.04 0.99–1.44

BMI, body mass index; CHD, coronary heart disease. aAdjusted for age and gender. bAll variables in the same model. Vascular Medicine 2006; 11: 232–238

235

236

MA Allison et al

Figure 1 Odds of renal artery calcium by prevalence of calcium in each vascular bed. (White bars ⫽ unadjusted; light gray bars ⫽ adjusted for age and sex; dark gray bars ⫽ adjusted for age, sex, body mass index, dyslipidemia, hypertension, diabetes, smoking and family history of CHD. I ⫽ 95% confidence interval; p ⬍ 0.05 for all except iliacs.)

vided above with the exception that calcium in the mitral annulus was no longer significantly associated with prevalent RAC. Table 3 shows the sensitivity, specificity, positive predictive values and negative predictive values for any renal artery calcium by vascular bed and valvular annuli. The presence of calcium in the abdominal aorta had the highest sensitivity for any RAC (94.5%), while the absence of calcium in the mitral annulus gave the highest specificity (96.8%). A ‘positive’ test result for calcium in the mitral annulus gave the highest positive predictive value for any RAC (62.5%) while a ‘negative’ test result for calcium in the abdominal aorta gave the highest negative predictive value (97.7%). On receiver operator characteristic (ROC) analysis, the area under the curve (AUC) was highest in the thoracic aorta (0.773), indicating the highest ability to discriminate the presence and absence of calcium in the renal arteries. The AUC for the abdominal aorta was similar (0.722).

Table 3 ‘Screening test’ analysis for prevalent renal artery calcium. Vascular area

Sensitivity Specificity PPV

NPV

Carotids Coronaries Thoracic aorta Abdominal aorta Iliacs Aortic annulus Mitral annulus

74.9 90.9 88.2 94.5 85.6 58.9 24.7

93.3 95.8 96.4 97.7 93.9 90.5 85.4

76.4 45.8 67.8 51.1 49.0 85.8 96.8

41.0 26.9 37.3 29.6 26.8 47.7 62.5

PPV, positive predictive value; NPV, negative predictive value. Vascular Medicine 2006; 11: 232–238

Discussion The results of this large study of individuals free of clinical CVD demonstrate that the presence of calcium in the renal arteries is significantly associated with calcium due to atherosclerosis in any of five distinct vascular beds or in either valvular annulus, and that these associations are independent of traditional CVD risk factors. With concurrent adjustment for calcium in the other beds or annuli, all except the iliac arteries remained significantly and independently associated with RAC, with the strongest association being with calcium in the thoracic or abdominal aorta. Furthermore, the correlations between RAC and calcium in the vascular beds and annuli were highly significant. Of the CVD risk factors, age, sex, and hypertension were significantly associated with the presence of any RAC after multivariable adjustment, while the correlations between RAC and lipids, and body mass index were all very modest and non-significant. Previous studies suggest that calcification of atherosclerotic plaques occurs via a process similar to endochondral bone formation of skeletal bone.8 Histologically, these calcium deposits appear first in stage 4 lesions and therefore represent more advanced plaques.18 The preponderance of clinical research on calcified atherosclerosis has focused on the coronary arteries. These studies have shown that the amount of calcium in the coronary arteries is highly correlated with the extent of atherosclerotic plaque burden and, to a lesser extent, the degree of stenosis in these vessels.19 The existing literature on calcium in the renal arteries is very limited and we are aware of only two reports on the subject. In a clinical population, Siegel and colleagues reported a positive, but not statistically significant, association between the extent of RAC

Systemic atherosclerosis and renal artery calcium

determined by computed tomography and the presence of renal artery stenosis on angiography20; a finding similar to those for calcium in the coronary arteries.19 In a study of 96 Caucasian individuals with diabetes, Freedman et al found significant correlations between RAC and calcium due to atherosclerosis in the carotid, coronary and iliac arteries as well as the infrarenal aorta, which were of similar magnitude to those found in our study.21 Given this information, the results of our study provide the first estimate of the prevalence of RAC, as well as its association with CVD risk factors and atherosclerosis in multiple vascular beds in a large cohort free of clinical disease. Importantly, the independent association between RAC and calcium in other vascular beds may in part be due to the shared exposure between the atherosclerosis already present and the influence of traditional CVD risk factors. The overall prevalence of renal artery calcium in our study was 18%. Freedman et al reported a RAC prevalence of 65% in their cohort of diabetic patients.21 In the Cardiovascular Health Study, 7% of this elderly population was found to have significant luminal stenosis by renal duplex sonography.22 Beyond these studies, the prevalence of significant renovascular atherosclerotic disease is primarily based on the percent with significant luminal stenosis and has been reported to range from 10% to 25%.12–14,23,24 Notably, these studies used different methodologies for measuring luminal stenosis (ie angiography, ultrasonography, autopsy), employed different cut-points for significant stenoses (50–75%) and were conducted in clinical cohorts of subjects typically with either significant coronary, cerebrovascular or end-stage renal disease. Therefore, generalization of such studies to less selected populations is problematic. However, the differences in prevalences and the finding of a non-significant association between RAC and luminal stenosis reported by Siegel et al20 suggest that the presence of calcium in the renal arteries is not a sensitive measure for the presence of hemodynamically significant disease. The risk factors that we found to be significantly associated with RAC are similar to those that have been previously reported to be significant for calcium in other vascular beds. In general, increasing age and male gender are the strongest correlates of calcium in all vascular beds.10,25,26 In addition to these two variables, hypertension was significantly associated with RAC in our cohort. This risk factor has been found to be highly associated with atherosclerotic calcification in other vascular beds to include the carotids, coronaries, thoracic aorta and abdominal aorta, as well as the mitral and aortic annuli.10,11,26 Cholesterol levels (both HDL and LDL) do not appear to be strongly associated with vascular calcium10,27; a finding that was reproduced here. Atherosclerotic renal artery stenosis is the leading cause of ischemic renal disease28 and patients with the Vascular Medicine 2006; 11: 232–238

237

most severe anatomic atherosclerotic renal vascular disease have worse hypertension and increased mortality.29 Furthermore, higher levels of serum creatinine are associated with higher levels of subclinical atherosclerosis,30 suggesting a continuous relationship between atherosclerosis and renal function. Early detection of the atherosclerotic process in the renal arteries may therefore allow for the early implementation of prevention strategies aimed at preserving as much kidney function as possible. Based on the results of our study, treatment of hypertension should be a principal focus to do so as this was the only modifiable CVD risk factor that was significantly associated with RAC. However, given the scarcity of information on renal artery calcium, further research is warranted to explore its pathophysiologic relationships with renal dysfunction and the potential clinical utility of this measure. Limitations Subjects for this study were self-selected. Therefore, the results of this study may not be generalizable to general/community-based populations. Misclassification of diabetes status in this study was possible due to the definition used for this variable (ie self-report or medication for diabetes). Calcium in the intima or media of the arterial wall is not distinguished by EBCT. Notably, medial calcification occurs primarily in those with diabetes mellitus31 or chronic kidney disease32 and is typically located in the lower extremities of these patients. As the prevalence of diabetes is quite low in our study, we believe misclassification based on calcium location is low.

Conclusions The results of this study indicate that calcium in the renal arteries is highly associated with atherosclerotic calcification in other vascular beds and in the valvular annuli. These relationships are largely independent of traditional CVD risk factors.

Acknowledgements This work was supported by a grant (Allison) and a summer research fellowship (DiTomasso) from the American Heart Association.

References 1

2

Khoury Z, Schwartz R, Gottlieb S, Chenzbraun A, Stern S, Keren A. Relation of coronary artery disease to atherosclerotic disease in the aorta, carotid, and femoral arteries evaluated by ultrasound. Am J Cardiol 1997; 80: 1429–33. Fazio S, Linton MF. The inflamed plaque: cytokine production and cellular cholesterol balance in the vessel wall. Am J Cardiol 2001; 88: 12E–15E.

238

3 4

5 6

7

8

9

10

11

12

13

14

15

16

MA Allison et al

Libby P. Vascular biology of atherosclerosis: overview and state of the art. Am J Cardiol 2003; 91: 3A–6A. DeBakey ME, Lawrie GM, Glaeser DH. Patterns of atherosclerosis and their surgical significance. Ann Surg 1985; 201: 115–31. Bostrom K. Insights into the mechanism of vascular calcification. Am J Cardiol 2001; 88: 20E–22E. Dhore CR, Cleutjens JP, Lutgens E et al. Differential expression of bone matrix regulatory proteins in human atherosclerotic plaques. Arterioscler Thromb Vasc Biol 2001; 21: 1998–2003. Engelse MA, Neele JM, Bronckers AL, Pannekoek H, de Vries CJ. Vascular calcification: expression patterns of the osteoblast-specific gene core binding factor alpha-1 and the protective factor matrix gla protein in human atherogenesis. Cardiovasc Res 2001; 52: 281–89. Demer LL, Tintut Y. Mineral exploration: search for the mechanism of vascular calcification and beyond: the 2003 Jeffrey M. Hoeg Award lecture. Arterioscler Thromb Vasc Biol 2003; 23: 1739–43. Mody N, Tintut Y, Radcliff K, Demer LL. Vascular calcification and its relation to bone calcification: possible underlying mechanisms. J Nucl Cardiol 2003; 10: 177–83. Allison MA, Criqui MH, Wright CM. Patterns and risk factors for systemic calcified atherosclerosis. Arterioscler Thromb Vasc Biol 2004; 24: 331–36. Allison MA, Cheung P, Criqui MH, Langer RD, Wright CM. Mitral and aortic annular calcification are highly associated with systemic calcified atherosclerosis. Circulation 2006; 113: 861–66. Gross CM, Kramer J, Waigand J, Luft FC, Dietz R. Relation between arteriosclerosis in the coronary and renal arteries. Am J Cardiol 1997; 80: 1478–81. Harding MB, Smith LR, Himmelstein SI et al. Renal artery stenosis: prevalence and associated risk factors in patients undergoing routine cardiac catheterization. J Am Soc Nephrol 1992; 2: 1608–16. Jean WJ, al-Bitar I, Zwicke DL, Port SC, Schmidt DH, Bajwa TK. High incidence of renal artery stenosis in patients with coronary artery disease. Cathet Cardiovasc Diagn 1994; 32: 8–10. Olin JW, Melia M, Young JR, Graor RA, Risius B. Prevalence of atherosclerotic renal artery stenosis in patients with atherosclerosis elsewhere. Am J Med 1990; 88: 46 N–51 N. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M Jr, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990; 15: 827–32.

Vascular Medicine 2006; 11: 232–238

17 Hosmer DW Jr, Lemeshow S. Applied logistic regression, 2nd edition. New York: John Wiley & Sons, Inc, 2000. 18 Stary HC. The development of calcium deposits in atherosclerotic lesions and their persistence after lipid regression. Am J Cardiol 2001; 88: 16E–19E. 19 Sangiorgi G, Rumberger JA, Severson A et al. Arterial calcification and not lumen stenosis is highly correlated with atherosclerotic plaque burden in humans: a histologic study of 723 coronary artery segments using nondecalcifying methodology. J Am Coll Cardiol 1998; 31: 126–33. 20 Siegel CL, Ellis JH, Korobkin M, Dunnick NR. CT-detected renal arterial calcification: correlation with renal artery stenosis on angiography. AJR Am J Roentgenol 1994; 163: 867–72. 21 Freedman BI, Hsu FC, Langefeld CD et al. Renal artery calcified plaque associations with subclinical renal and cardiovascular disease. Kidney Int 2004; 65: 2262–67. 22 Hansen KJ, Edwards MS, Craven TE et al. Prevalence of renovascular disease in the elderly: a population-based study. J Vasc Surg 2002; 36: 443–51. 23 Appel RG, Bleyer AJ, Reavis S, Hansen KJ. Renovascular disease in older patients beginning renal replacement therapy. Kidney Int 1995; 48: 171–76. 24 Kuroda S, Nishida N, Uzu T et al. Prevalence of renal artery stenosis in autopsy patients with stroke. Stroke 2000; 31: 61–65. 25 Allison MA, Wright CM. Age and gender are the strongest clinical correlates of prevalent coronary calcification (R1). Int J Cardiol 2005; 98: 325–30. 26 Wong ND, Sciammarella M, Arad Y et al. Relation of thoracic aortic and aortic valve calcium to coronary artery calcium and risk assessment. Am J Cardiol 2003; 92: 951–55. 27 Allison MA, Wright CM. A comparison of HDL and LDL cholesterol for prevalent coronary calcification. Int J Cardiol 2004; 95: 55–60. 28 Greco BA, Breyer JA. Atherosclerotic ischemic renal disease. Am J Kidney Dis 1997; 29: 167–87. 29 Wright JR, Shurrab AE, Cheung C et al. A prospective study of the determinants of renal functional outcome and mortality in atherosclerotic renovascular disease. Am J Kidney Dis 2002; 39: 1153–61. 30 Bleyer AJ, Shemanski LR, Burke GL, Hansen KJ, Appel RG. Tobacco, hypertension, and vascular disease: risk factors for renal functional decline in an older population. Kidney Int 2000; 57: 2072–79. 31 Edmonds ME. Medial arterial calcification and diabetes mellitus. Z Kardiol 2000; 89: 101–104. 32 Raggi P. Cardiovascular calcification in end stage renal disease. Contrib Nephrol 2005; 149: 272–78.