| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Special Feature |
Department of Radiology, Loma Linda University Medical Center (E.S.), Loma Linda, California 92354; the Childrens Hospital Los Angeles, University of Southern California (K.A., X.L., V.G.), Los Angeles, California 90027; and the Department of Biostatistics, University of California Los Angeles (J.S.), Los Angeles, California 90024
Address all correspondence and requests for reprints to: Vicente Gilsanz, M.D., Ph.D., Department of Radiology, MS 81, 4650 Sunset Boulevard, Los Angeles, California 90027. E-mail: vgilsanz{at}chla.usc.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Of the 2348 subjects studied, 70% had osteoporosis, 30% had at least one vertebral fracture, and 9% had at least one hip fracture. Aortic calcifications were inversely related to bone density and directly related to fractures. After adjusting for age and potential confounders, measures for aortic calcification predicted 26.1% of the variance in bone density (P < 0.001). Compared with women without calcification, the odds ratios for vertebral and hip fractures in those with calcification were estimated to be 4.8 (95% confidence interval, 3.6–6.5) and 2.9 (95% confidence interval, 1.8–4.8), respectively. The subgroup analysis of 228 women longitudinally studied showed that the percentage of yearly increase in aortic calcification accounted for 47% of the variance in the percentage rate of bone loss (P < 0.001). Moreover, a strong graded association was observed between the progression of vascular calcification and bone loss for each quartile. Women in the highest quartile for gains in aortic calcification had four times greater yearly bone loss (5.3 vs.1.3% yearly; P < 0.001) than women of similar age in the lowest quartile. Smaller, but highly significant differences were also found between all other quartiles.
We conclude that aortic calcifications are a strong predictor for low bone density and fragility fractures.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
In contrast to projection techniques, computed tomography (CT) can independently quantify the mineral in arterial and skeletal tissues with precision and accuracy (21, 22, 23). In the current study, we used CT to examine the possible relations between calcified plaques in the aorta, a risk factor for atherosclerosis and cardiovascular disease (24), and the two morphological traits that characterize osteoporosis: low bone density and fragility fractures (17). To this end, we systematically reviewed the medical records and analyzed the CT digital data of a large number of women who underwent bone density examinations. Our aim was to test the hypotheses that there is a significant age-independent relation between atherosclerosis and osteoporosis and that increases in vascular calcification and bone loss progress in parallel.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Study population
From 1984–1998, 4261 patients had one or more CT bone density determinations at the Department of Radiology, Loma Linda University Medical Center, (Fig. 1
). From a review of the medical records, which included information about dietary intake and physical activity, we identified postmenopausal women 50 yr of age and older of European descent who were ambulatory, did not smoke, and did not drink more than one alcoholic beverage per month. Excluded from this study were women with bone disorders other than primary osteoporosis, with any rheumatological or endocrine conditions requiring medical treatment, or with any history of cancer in the past 5 yr. Also excluded were subjects who received therapeutic doses of sex steroids, bisphosphonates, calcitonin, corticosteroids, thiazide, and fluoride. Using these criteria, we selected 2348 women whose CT studies were analyzed to assess the possible associations between bone mineralization and aortic calcification.
|
Further review of the records revealed that 1049 study subjects returned for additional CT bone density determinations 9 months to 8.2 yr later. Data from this subset indicated that 821 were on therapy or suffered from a condition that could alter bone metabolism and that 228 women continued to be ambulatory, healthy, and medication-free (Fig. 2). Thereafter, the follow-up CT studies of these 228 women were examined to determine whether changes in bone density and changes in aortic calcification progress in parallel.
|
All studies were obtained with either a CT-T 9800 scanner (General Electric Co., Milwaukee, WI) or a Somatom Plus scanner (Siemens, Erlangen, Germany), and reference phantoms for calibration were scanned simultaneously with the patient. Scout radiographs were used for location of the scanning sites.
Aortic calcification determinations.
Using specially designed software, the extent of calcification in the aortic wall was quantified at the lowest thoracic vertebra (T12) and the first four lumbar vertebrae (L1–L4) from the same CT axial scans used to obtain bone measurements. To this end, the aorta was magnified x10, its boundaries were identified at four equidistant sites, and the number of pixels in the circle or oval was defined by the four quadratic points calculated. All pixels in the perimeter of the aorta were measured based on a fixed percentage difference between the material density of the wall and the measured density of fat immediately adjacent to the wall. Thereafter, the ratio of pixels with and without calcification was determined visually, and the results were expressed as degrees of calcified wall, from 0–360 (Fig. 2
). The coefficient of variation (CV) for repeated measurements of aortic wall calcification was estimated to be 1.1%. The reliability of CT in predicting atherosclerotic lesions has been validated by histopathology studies of the coronary arteries (22).
Bone density determinations. Measurements of cancellous bone density were acquired from the cross-sectional images at the levels of T12-L4. Cancellous bone density of the vertebral bodies was measured in unfractured vertebrae, as previously described (27). For the purposes of this study, the density of cancellous bone was defined as the mean value of the CT unit of measurement (milligrams per centimeters3) at the midportion of the vertebral bodies. The CV for measurements of cancellous bone density in the vertebral bodies has previously been calculated to be 3% (27). Current criteria define osteoporosis as a condition wherein the value for bone mineral density is more than 2.5 SD below the young adult mean value, which corresponds to CT measurements for cancellous bone density less than 110 mg/cm3 (28).
Fracture determinations. Vertebral and hip fractures were identified from the lateral and the anteroposterior CT scout radiographs, respectively. From the lateral CT spine radiograph, measurements of vertebral height were calculated at the anterior, middle, and posterior portions of the vertebral body. For the purposes of this study, wedge fractures were defined by a reduction of at least 20% in vertebral anterior height compared with posterior height; central collapse fractures by a loss of at least 20% height in the midportion of the vertebral body compared with either the anterior or posterior height; and crush fractures by a reduction in both the anterior and posterior height when compared with the closest adjacent intact superior vertebrae. The CVs for measurements of vertebral body height have previously been calculated to be 1.6% (27).
Women with hip fractures were identified from the CT anteroposterior scout radiographs, which depicted the multiple screws, the sliding hip screw, or the prostheses required for its treatment, and the diagnosis was confirmed through medical records. The number of fractures at sites other than the vertebrae and the proximal femurs was determined solely by reviewing the medical records.
Statistical analysis
Statistical analysis was carried out using SPSS for Windows (version 10, SPSS Inc., Chicago, IL) and BMDP (version 7.0, BMDP Statistical Software, Cork, Ireland) statistical programs. The data were analyzed by the Student’s t test for paired and unpaired samples, simple and multiple regression analyses, and Hotelling’s T2 (29). All results are expressed as mean ± SD. An 
level of 0.05 was accepted for significance of all statistical procedures. Values for aortic calcification represent the mean value of all five sites from T12-L4, whereas values for vertebral cancellous bone density represent only the mean value for unfractured vertebrae from T12-L4.
To examine differences in bone density between subjects with and without aortic calcification, regression models were fit for aortic calcification as the dependent variable and bone density as the independent variable using the General Linear Models procedure (30). The effects of adjustment for age, weight, height, body mass index (BMI), years post menopause, systolic and diastolic blood pressure, self-reported calcium intake, and physical activity index were also explored by including these as independent variables in the model. Differences among quartiles were assessed by ANOVA (31).
Differences in the percentage of vertebral and hip fractures among participants with and without aortic calcification were assessed using two-group regression analysis. To obtain better control of the effect of age on fracture prevalence, women were classified according to 5-yr age groupings (50–54, 55–59, 60–64, 65–69, 70–74, 75–79, 80–84, 85–89, and 90–94) when studying vertebral fractures and 10-yr age groupings (50–59, 60–69, 70–79, 80–89, and 90–99) when studying fractures of the proximal femur. The number of fractures in subjects with and without aortic calcification was compared using Student’s t tests for unpaired samples and the Hotelling’s T2 (30).
For the longitudinal analysis, we used a linear regression model to compare rates of progression in aortic calcification and bone loss (as percentage yearly change). Changes in aortic calcification were used as the dependent variable, and percentage of bone loss was used as the independent variable. To account for potential confounders and other covariates, baseline values for age, anthropometric parameters, blood pressure, self-reported calcium intake, self-reported physical activity, and vertebral and hip fractures, as well as percentage changes in these variables between baseline and follow-up studies were included in the model. Differences among quartiles were assessed by ANOVA (31).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
CT digital data indicated that the prevalence of aortic calcification, osteoporosis (defined as a value for cancellous bone density 2.5 SD below peak values), and vertebral and/or proximal femur fractures increased with age (Table 1). Overall, measures of aortic calcification correlated as follows: moderate correlation with age (r = 0.50; P < 0.001), years post menopause (r = 0.49; P < 0.001), and bone density (r = –0.35; P < 0.001); weak correlation with values for vertebral and hip fractures (r = 0.23 and 0.13, respectively; both P values < 0.001), physical exercise (r = 0.23; P < 0.001), and systolic blood pressure (r = 0.16; P < 0.001); and no correlation with BMI, diastolic blood pressure, or calcium intake.
|
. Women with aortic calcification were older, further from menopause, less physically active, and had higher systolic blood pressure, lower bone density, and a greater number of vertebral and hip fractures. Regression models showed that these differences did not remain significant once age was taken into account. In contrast, aortic calcification had a significantly independent association with bone density. After adjusting for age, measures of aortic calcification (AC) predicted 26.1% of the variance in bone density (BD): AC = –137.041 – 0.240*BD + 3.101*Age. Similar findings were observed when the entire population was divided into four equal strata (Table 3).
|
|
). Overall, the estimate of the relative risk for vertebral fractures in women with aortic calcification was 4.8 (95% confidence interval, 3.6–6.5) when compared with women without calcification. However, the estimate differed among age groups. For example, the odds ratio for vertebral fractures in women with calcification was estimated to be 2.2 at the ages of 50–54 yr and rose to 4.0 by 80–84 yr of age. Figure 3 also shows that in women with fractures, the number of vertebral fractures was significantly greater in subjects with aortic calcification than those without. This was true for younger age categories and when assessing the number of fractures in subjects with and without calcification as a vector of observations for all age categories [Hotelling’s T2 test (P = 0.03)]. Similarly, both the prevalence and the number of hip fractures were significantly greater in women with aortic calcification than those without (Fig. 4). The estimate of the relative risk for proximal femur fractures in women with aortic calcification was 2.9 (95% confidence interval, 1.8–4.8) when compared with women without calcification. Among subjects with fractures, bilateral hip fractures were more common in women with aortic calcification than those without [14.3 vs. 19.8% (P = 0.04)].
|
|
The longitudinal cohort consisted of 228 postmenopausal women, ages 65.2 ± 9.8 yr, who had repeat CT studies 2.1 ± 1.9 yr after the initial examination (median, 15.3 months; interquartile range, 14 months). Subjects with aortic calcification were significantly older, had lower bone density, and had a greater number of vertebral and hip fractures both at baseline and at the final CT measurement (all P values < 0.001).
At follow-up, there were no significant differences in body mass, blood pressure, self-reported calcium intake, self-reported physical activity, or number of hip fractures. In contrast, values for aortic calcification, bone density, and vertebral fractures changed significantly (all P values <0.001). On average, in subjects with baseline calcification, the yearly rate of progression in aortic calcification was 13.8%, and the yearly rate of bone loss 2.6%, but there was a wide distribution of values. Stepwise regression analysis indicated that the rate of change in aortic calcification accounted for a significant proportion of the variance in the yearly rate of bone loss and that once this independent variable entered into the equation, no other covariate increased the predictive power of the model. This was true whether rates of bone loss and aortic calcification gains were assessed as yearly percentage change or as yearly changes in absolute values; BD = 1.57 + 0.0806 AC (r2 = 0.471; P < 0.001) and BD = 163.83 + 0.117X (r2 = 0.223; P < 0.001), respectively (Fig. 5).
|
). Women in the highest quartile for gains in aortic calcification had four times greater yearly bone loss (5.3 vs. 1.3% yearly; P < 0.001) than women of similar age in the lowest quartile. Smaller, but highly significant differences were also found between all other quartile comparisons.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Our results, showing that women with vascular calcifications are predisposed to osteoporosis and fragility fractures, complement existing epidemiological data indicating that women with osteoporosis are at an increased risk of mortality that is unrelated to the occurrence of fractures but is due to stroke or coronary disease (32, 33). Some have suggested that low bone mass is an even stronger predictor of cardiovascular disease than other well-known risk factors, such as serum cholesterol and smoking (32, 33). Similarly, the results of the subgroup of subjects longitudinally studied, showing that women with a rapid progression of vascular calcification are at increased risk for osteoporosis, complement published data indicating that older women who lose bone rapidly have an increased rate of mortality due to cardiovascular disease (34).
The large sample size, the inclusion of a longitudinal arm, and the use of CT to obtain graded measures of skeletal and vascular calcification are strengths of the present study. Previous studies on the possible links between aortic calcification and bone density were limited by the use of projection techniques and simply ranked vascular calcification as present or absent, yielding discrepant results (1, 2, 3, 4, 5, 6, 7, 10, 11, 12, 13, 14, 35, 36). In addition, measurements of bone mineral content in the axial skeleton with dual-energy x-ray absorptiometry (DXA) are inaccurately high in the presence of calcified plaques in the adjacent aorta, which are incorporated in the region of interest (18, 19, 20). It is noteworthy that the previous cross-sectional study with the largest cohort of postmenopausal Caucasian women (2051 subjects older than 65 yr) found no independent association and attributed the relation between vascular calcifications and bone loss measured by DXA to the aging process (3). Similarly, the results of a smaller longitudinal study, also using DXA, found measures of bone loss in postmenopausal women to be unrelated to aortic calcification (15).
There are several limitations in this study. First, the retrospective design did not allow for a rigorous collection of all pertinent data. It should be stressed, however, that the retrospective approach did not influence the main outcome variables of the current study: objective quantitative measures of traits associated with osteoporosis and atherosclerosis. Second, the subjects were not recruited from the community at large, but were selected from patients who underwent bone density determinations. This selection bias likely explains the relatively high prevalence of osteoporosis and fractures in the subjects studied (37). Any bias, however, introduced by the method of selection should not have influenced the results. Third, we examined only women of European descent, and therefore our results may not apply to men or to other ethnic groups. Lastly, it is impossible to determine whether the association we found represents a causal relation between skeletal and vascular mineralization or a confounding effect due to sex steroids, calcium-regulated hormones, vitamin K, homocysteine, lipid oxidation products, or other yet-to-be-identified metabolic or hormonal factors (38).
Much of the evidence for a causal relationship comes from basic research and from studies in experimental animals, suggesting that arterial tissue is calcified in an organized, regulated process by mechanisms similar to those involved in the mineralization of bone. The mineral deposit in the arterial wall, hydroxyapatite, is the same mineral found in bone, and it is structurally arranged with trabeculae and lacunae visible in the calcific deposit (39). Molecular biological studies indicate that regulated osteogenesis may occur in some cells of the arterial wall (40). Indeed, cells with both osteoblastic and osteoclastic potential have been described in vascular tissues, and bone-related proteins have been identified in calcified arterial lesions (40, 41, 42). Moreover, genetically engineered animals, susceptible to the development of calcified atherosclerotic lesions, have low values for bone density, whereas those resistant to the development of atherosclerotic lesions have higher bone density values (43, 44). Recently, statins, long recognized to be effective in lowering cholesterol levels and reducing cardiovascular risk, have been shown to trigger bone formation in tissue cultures and animals studies (45). In humans, several observational studies have suggested that statins are associated with increased bone mineral density and a reduced fracture risk (46, 47).
Regardless of the mechanism responsible for the simultaneous mineralization of vascular and skeletal tissues, the findings of this study have significant clinical implications with regard to the identification of elderly women at risk for osteoporosis. We found that postmenopausal women 60 yr of age and younger with aortic calcification had more than double the rate of bone loss (3.4 vs. 1.4% yearly) as women without calcified plaques. Because approximately 15% of all abdominal radiographs and 50% of all abdominal CT studies in women 50–60 yr of age depict aortic calcification (2), this finding provides an imaging marker for the identification of a large population of women who have accelerated bone loss and are at greatest risk for osteoporosis. Future studies are needed to determine the specificity and sensitivity of aortic calcifications in predicting fracture risk.
In conclusion, the results of the current study disagree with the classic notion that, in human aging, the inevitable progressive deterioration of every system is organ and cell specific (48). At least with reference to the skeletal and vascular tissues, the deterioration occurs at the same time, and the rate of decline progresses in parallel. The findings of this study, showing that beyond the aging process there is a significant relation between atherosclerosis and osteoporosis, should encourage research on the mechanisms regulating mineral deposition in connective tissues. They also support the search for effective prevention and treatment strategies that may simultaneously modify the risk for these two common conditions. Lastly, knowledge that in young, postmenopausal women the presence of calcified plaques in the aorta is associated with accelerated bone loss should aid in the early identification of subjects at risk for osteoporosis and fractures.
|
Footnotes |
|---|
Abbreviations: BMI, Body mass index; CT, computed tomography; CV, coefficient of variation; DXA, dual-energy x-ray absorptiometry.
Received June 4, 2003.
Accepted February 18, 2004.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. N. Bakhireva, G. A. Laughlin, R. Bettencourt, and E. Barrett-Connor Does Osteoprotegerin or Receptor Activator of Nuclear Factor-{kappa}B Ligand Mediate the Association between Bone and Coronary Artery Calcification? J. Clin. Endocrinol. Metab., May 1, 2008; 93(5): 2009 - 2012. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Raggi and M. Kleerekoper Contribution of Bone and Mineral Abnormalities to Cardiovascular Disease in Patients with Chronic Kidney Disease Clin. J. Am. Soc. Nephrol., May 1, 2008; 3(3): 836 - 843. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Kearns, S. Khosla, and P. J. Kostenuik Receptor Activator of Nuclear Factor {kappa}B Ligand and Osteoprotegerin Regulation of Bone Remodeling in Health and Disease Endocr. Rev., April 1, 2008; 29(2): 155 - 192. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Reddy, J. P. Bilezikian, S. J. Smith, and L. Mosca Reduced Bone Mineral Density Is Associated with Breast Arterial Calcification J. Clin. Endocrinol. Metab., January 1, 2008; 93(1): 208 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Andress Bone and Mineral Guidelines for Patients with Chronic Kidney Disease: A Call for Revision Clin. J. Am. Soc. Nephrol., January 1, 2008; 3(1): 179 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Manolagas and M. Almeida Gone with the Wnts: {beta}-Catenin, T-Cell Factor, Forkhead Box O, and Oxidative Stress in Age-Dependent Diseases of Bone, Lipid, and Glucose Metabolism Mol. Endocrinol., November 1, 2007; 21(11): 2605 - 2614. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Almeida, L. Han, M. Martin-Millan, L. I. Plotkin, S. A. Stewart, P. K. Roberson, S. Kousteni, C. A. O'Brien, T. Bellido, A. M. Parfitt, et al. Skeletal Involution by Age-associated Oxidative Stress and Its Acceleration by Loss of Sex Steroids J. Biol. Chem., September 14, 2007; 282(37): 27285 - 27297. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Vogelmeier and R. Bals Chronic Obstructive Pulmonary Disease and Premature Aging Am. J. Respir. Crit. Care Med., June 15, 2007; 175(12): 1217 - 1218. [Full Text] [PDF]
|
|
|
|
|
|
R. Sabit, C. E. Bolton, P. H. Edwards, R. J. Pettit, W. D. Evans, C. M. McEniery, I. B. Wilkinson, J. R. Cockcroft, and D. J. Shale Arterial Stiffness and Osteoporosis in Chronic Obstructive Pulmonary Disease Am. J. Respir. Crit. Care Med., June 15, 2007; 175(12): 1259 - 1265. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Marz, U. Seelhorst, B. Wellnitz, B. Tiran, B. Obermayer-Pietsch, W. Renner, B. O. Boehm, E. Ritz, and M. M. Hoffmann Alanine to Serine Polymorphism at Position 986 of the Calcium-Sensing Receptor Associated with Coronary Heart Disease, Myocardial Infarction, All-Cause, and Cardiovascular Mortality J. Clin. Endocrinol. Metab., June 1, 2007; 92(6): 2363 - 2369. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Raggi, A. Bellasi, E. Ferramosca, G. A. Block, and P. Muntner Pulse Wave Velocity Is Inversely Related to Vertebral Bone Density in Hemodialysis Patients Hypertension, June 1, 2007; 49(6): 1278 - 1284. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. B. Lehtinen, K. P. Burdon, J. P. Lewis, C. D. Langefeld, J. T. Ziegler, S. S. Rich, T. C. Register, J. J. Carr, B. I. Freedman, and D. W. Bowden Association of {alpha}2-Heremans-Schmid Glycoprotein Polymorphisms with Subclinical Atherosclerosis J. Clin. Endocrinol. Metab., January 1, 2007; 92(1): 345 - 352. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Bezerra, G. D. Calomeni, V. F. Caparbo, E. S. Gebrim, M. S. Rocha, and R. M. R. Pereira Low bone density and low serum levels of soluble RANK ligand are associated with severe arterial calcification in patients with Takayasu arteritis Rheumatology, December 1, 2005; 44(12): 1503 - 1506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. C. W. Skinner Biominerals Mineralogical Magazine, October 1, 2005; 69(5): 621 - 641. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Hamerman Osteoporosis and atherosclerosis: biological linkages and the emergence of dual-purpose therapies QJM, July 1, 2005; 98(7): 467 - 484. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Ali, R. S. Weinstein, S. A. Stewart, A. M. Parfitt, S. C. Manolagas, and R. L. Jilka Rosiglitazone Causes Bone Loss in Mice by Suppressing Osteoblast Differentiation and Bone Formation Endocrinology, March 1, 2005; 146(3): 1226 - 1235. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Rubin and S. J. Silverberg Vascular Calcification and Osteoporosis--The Nature of the Nexus J. Clin. Endocrinol. Metab., September 1, 2004; 89(9): 4243 - 4245. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
, median age for subjects without/with aortic calcification, respectively.
