Department of Medicine College of Physicians & Surgeons Columbia University New York, New York 10032
Address all correspondence and requests for reprints to: Shonni J. Silverberg, M.D., Columbia University, College of Physicians & Surgeons, Department of Medicine, PH8-864, 630 West 168th Street, New York, New York 10032. E-mail: sjs5{at}columbia.edu.
The association between cardiovascular disease and calcium deposition in the vasculature was first described in the 19th century (1). We know now that vascular calcification is associated with congestive heart failure, hypertension, myocardial ischemia, and coronary insufficiency. The extent of valve calcification is the best predictor of vascular disease progression in patients with aortic stenosis (2). Similarly, the risk for coronary heart disease for both men and women increases with aortic arch calcification, as does the risk of stroke, particularly in postmenopausal women (3). The process of vascular calcification is heterogeneous, occurring in different parts of the vessel and causing damage in varying ways. Atherosclerotic calcification occurs in the intima, potentially leading to luminal obstruction. Medial artery calcification, on the other hand, often seen in patients with diabetes and renal failure, occurs in the tunica media, and can lead to abnormal vascular compliance. Diffuse calcium phosphate deposition can also occur when the physiological calcium phosphate solubility threshold is exceeded.
Vascular calcification was in the past considered a passive process, a degenerative consequence of aging. It is now understood that calcium deposition in the vasculature is an active and regulated process akin to bone formation (4, 5). Calcified atherosclerotic arteries contain tissue that is histomorphologically indistinguishable from bone. A subpopulation of artery wall cells (calcifying vascular cells) has the ability to undergo osteoblastic differentiation. These cells may express various bone matrix proteins and skeletal regulatory factors that have been demonstrated in human calcified plaque, including osteocalcin, bone sialoprotein, osteonectin, collagen I, alkaline phosphatase, Msx-2, and Cbfa-1. Arterial calcification could therefore be said to recapitulate osteogenesis.
Vascular calcification has been found to coexist with bone loss in numerous epidemiological studies, suggesting a relationship between osteoporosis and atherosclerosis. Low bone mineral content at menopause has been shown to be a risk factor for increased cardiovascular mortality later in life (6). In the Framingham study, women with the lowest bone mass as determined by metacarpal cortical area on hand x-ray, now an outdated technique, had the highest incidence of coronary artery disease over a 30-yr course (7). Low bone density has also been associated with aortic calcification, as detected by conventional x-ray (8, 9, 10), and with surrogate markers of subclinical vascular disease (11, 12). However, whether the association between osteoporosis and vascular calcification is causal has not been definitively demonstrated. Vascular calcification in and of itself could affect bone metabolism; arterial stenosis could decrease peripheral blood supply, suppressing bone cell function, or atherosclerotic disease might limit physical activity, leading to bone loss. Moreover, both osteoporosis and aortic calcification increase with age (13) and, in some studies, an association between bone mineral density (BMD) and aortic calcification was lost once age was adjusted for (14, 15). Furthermore, the association has been observed almost only in women (7, 10, 11, 12), raising further uncertainty whether the two disease processes are truly linked.
A shift in bone growth, as it were, from the skeleton to the vasculature has been proposed to explain the association of bone loss and vascular calcification. Several factors make this simplistic view unlikely. First, the calcium homeostatic system is exquisitely regulated. Second, the imbalance in sheer quantity of calcium lost from the skeleton and that deposited in the vasculature mandates a more complex explanation. Estrogen deficiency has been proposed as a common mediator in the emergence of cardiovascular disease and bone loss. The menopausal decrease in estrogen could be important; Hak et al. (8) showed that the relation between bone loss and progressive aortic calcification was strongest in women during menopause as compared with postmenopausal women. However, in vitro evidence showing that estradiol promotes calcification of vascular smooth muscle cells and the bulk of recent evidence supporting adverse cardiovascular effects of hormone replacement therapy make a protective role for estrogen less likely. A perturbation in calcium homeostasis could be responsible; excess vitamin D intake can cause not only skeletal abnormalities but can lead to arterial calcification (16). Higher levels of circulating calcium, even within the normal range, have been shown to be associated with carotid plaque thickening, an early marker of cardiovascular disease (17). There is also evidence that oxidized lipids play a role as the inflammatory stimulus to vascular calcification. Epidemiological studies support a relationship between hyperlipidemia and vascular calcification, and elevated low-density lipoprotein has been associated with low radial BMD (18, 19). This is supported by in vitro evidence that lipids stimulate osteoblastic differentiation in the artery while depressing osteoblastic and stimulating osteoclastic differentiation in bone (20, 21). Additionally, high-density lipoprotein may exert its cardioprotective effect by inhibiting the osteogenic activity of calcifying vascular cells (22). Finally, the receptor activator of nuclear factor-B ligand/osteoprotegerin (OPG) system has recently been considered a possible link between bone and vascular disease. In OPG-deficient mice, severe osteoporosis and medial calcification develop (23). Receptor activator of nuclear factor-
B has been shown to promote aortic valve calcification, whereas OPG inhibits it (24). Coronary artery disease and cardiovascular mortality have been associated with elevated OPG levels (25, 26); the higher OPG levels might be a compensatory mechanism to counteract osteoporosis or vascular calcification. Given the pace of current research in this area, the molecular link between vascular calcification and bone loss may soon be elucidated.
In this issue of JCEM, Schulz et al. (27) further investigate the clinical association between osteoporosis and vascular calcification. In a retrospective study of 2348 women, they studied the relationship between aortic calcification and BMD, both measured by computed tomography (CT), and vertebral and hip fractures. The authors also followed 228 of the women longitudinally. After adjusting for age, aortic calcification predicted 26% of the variance in BMD. Women with calcification were 4.8 times more likely to have spine fractures and 2.9 times more likely to have hip fractures, compared with women without aortic calcification. In the small longitudinal cohort, the increase in aortic calcification accounted for 47% of the variance in bone loss, and those in the highest quartile of aortic calcification had the greatest yearly bone loss.
The use of CT technology for both aortic and spine measurements in this study makes it an important addition to the existing literature. All previous studies used conventional radiographs to measure aortic calcification, a less precise technique that does not permit the graded quantification available with CT. For bone density measurement, other studies employed posterior-anterior dual x-ray absorptiometry (DXA) assessment, in which calcium deposits in the aorta could be mistakenly interpreted as a falsely high BMD. CT measurement of BMD also allows for three-dimensional volumetric assessment, so that variations in bone size are less important than in two-dimensional areal DXA.
In interpreting these data it is important to acknowledge, as the authors do, the possibility of ascertainment bias. The women included were undergoing a skeletal evaluation, and indeed 70% were classified as osteoporotic. Furthermore, the prevalence of aortic calcification in this population (76% of subjects) is substantially greater than has been generally reported. For example, in a study of 60,393 healthy women, aortic calcification was detected on posterior-anterior chest radiographs in 1.5% of women at ages 4554 yr (vs. 32.4% of women at ages 5054 yr in this study); 5.7% of women at ages 5564 yr (vs. 57.3% of women at ages 5559 yr and 67.4% of women at ages 6064 yr in this study) and 15.9% of women at ages 6589 yr (>70% in this study) (3). Although CT imaging is a more sensitive technique for detection of aortic calcification than chest x-ray, the marked discrepancy between the prevalence of aortic calcification in this cohort and that of other healthy populations raises the possibility that the results of this study may not be applicable to all women.
The longitudinal portion of this study offers some of its most tantalizing results, with the suggestion that the presence of aortic calcification predicts increased subsequent bone loss. Unfortunately, only a very small number of subjects (228; <10% of the cross-sectional cohort) met criteria for this component of the study. Furthermore, follow-up studies were performed 9 months to 8.2 yr after baseline scans were obtained. Recent data suggest that the rate of arterial calcification is not linear. For instance, the rate of coronary calcification increases with baseline calcium score, and the curve describing that progression is sigmoidal (5).
The above-described limitations notwithstanding, this study adds compelling data to the literature on the association of vascular calcification and bone loss. Further studies are needed to investigate this association in men and in women with bone density across a wider spectrum. Larger, prospective studies using standardized interscan time periods will be important to confirm the longitudinal findings of Schulz et al. (27). Such clinical observations will complement the growing body of data emerging from the laboratory on the intersection of vascular calcification and bone biology.
Footnotes
Abbreviations: BMD, Bone mineral density; CT, computed tomograph; DXA, dual x-ray absorptiometry; OPG, osteoprotegerin.
Received July 8, 2004.
Accepted July 14, 2004.
References