NEW EVIDENCE CONNECTING CARDIOVASCULAR
DISEASE AND OSTEOPOROSIS
NIAMS-NHLBI Working Group
September 14-15, 1999
Several lines of evidence suggest
that there are connections between cardiovascular disease and osteoporosis.
Epidemiological studies suggest an association between low bone mass and
mortality from coronary artery disease and stroke. Insights into the molecular
mechanisms of vascular calcification suggest parallels with bone formation. New
information about the actions of statins, a class of cholesterol-lowering
drugs, suggest that they may have effects in both bone and vascular tissue. A
Working Group was convened by the National Heart, Lung, and Blood Institute and
the National Institute of Arthritis and Musculoskeletal and Skin Diseases to
bring together researchers from the bone and cardiovascular fields to review
recent observations bearing on possible links between cardiovascular disease
and osteoporosis and identify future research directions.
Several studies have reported an association between low bone mass and
cardiovascular disease. More recently, an association was found between low
bone mass and the risk of mortality from cardiovascular disease later in life.
These results would suggest that strategies to prevent osteoporosis might
prevent the development of cardiovascular disease as well. However, an
association is not consistently observed between low bone mass and
physiological markers of cardiovascular disease risk and progression. Further,
it is unclear how the calcification of atherosclerotic lesions, which appears
to parallel bone formation in some respects, is related to cardiovascular
mortality. A number of technical problems confound both the assessment of bone
mineral density and its correlation with vascular calcification.
Work with animal models
suggests several biological interactions and functional parallels between bone
and vascular tissue. Both types of tissue are responsive to hormones such as
sex steroids, and both cardiovascular disease and osteoporosis appear to
develop in part through the effects of inflammatory mediators. Cells with both
osteoblastic and osteoclastic potential may reside in vascular tissue. The
induction of obesity and diabetes in low density lipoprotein receptor-deficient
mice by a high-fat diet results in aortic calcification, characterized by
adventitial expression of genes typical of osteoblastic differentiation.
Oxidized lipids, long recognized as a factor in atherogenesis, may have
opposite effects on cells in the vascular and bone environments, accounting for
the paradoxical coincidence of vascular mineralization with bone loss. However,
it is clear that a variety of conditions can produce vascular calcification,
and that some types of calcification are distinct from that observed in
atherosclerotic lesions. For example, mice lacking osteoprotegerin frequently
exhibit aortic calcification in addition to osteopenia. Mice lacking matrix Gla
protein exhibit marked arterial calcification, suggesting that matrix Gla
protein acts as an inhibitor of arterial mineralization. The ectopic
calcification of aortic valves is enhanced in mice deficient in
osteoprotegerin, suggesting that osteoclast-like cells may resorb calcium
mineral in the artery wall.
statins have been shown to induce new bone formation in cell cultures and
animal models. The mechanism(s) responsible for this effect on bone are
unknown. The statins are recognized as very effective agents in lowering
cholesterol levels and reducing cardiovascular risk. Their primary effect is
inhibition of a key enzyme in cholesterol synthesis, but this inhibition may
have complex down-stream effects. For example, a neuroprotective effect of
statins in stroke appears to reflect the inhibition of lipid modification of
the GTPase Rho, which in turn leads to up-regulation of endothelial nitric
oxide synthase. Disruption of sterol metabolism may also affect protein
trafficking, membrane invagination, exocytosis, and fluidity.
Attempts to detect the skeletal effects of statins in
existing clinical cohorts have yielded only suggestive trends, falling short of
statistical significance. However, since the statins are targeted to the liver
for maximal effects on cholesterol synthesis, they may not reach bone in
effective concentrations. Interestingly, amino-bisphosphonates, which are
widely used to block bone resorption, also may act in part by inhibiting lipid
modification of signaling proteins such as Rho. There is a clear distinction,
however, between the anti-resorptive effect of bisphosphonates and the anabolic
effect of statins. The anti-resorptive action of bisphosphonates, along with
the observation of apparently resorptive cells in some instances of vascular
calcification, raises the possibility that bisphosphonates could actually
exacerbate cardiovascular disease.
Much of the evidence available at this time is
suggestive, rather than definitive. Further efforts across a broad range of
scientific and medical disciplines will be necessary to refine these
indications into clear conclusions. The goals stated below are illustrative of
those identified in the course of the meeting, but are not intended to be
comprehensive or exclusive of other relevant research areas.
relationship between cardiovascular disease, osteoporosis, and related
pathologies. It would be valuable to know whether osteoporosis and
atherosclerosis progress in parallel, and whether they share genetic or
environmental risk factors. Studies are needed to determine whether the
association between bone mass and cardiovascular disease indicated in women is
also observed in men, and in specific racial and ethnic groups which are known
to differ either in bone density and vitamin D metabolism or cardiovascular
risk. Also, studies to determine the mechanisms of bioprosthetic mineralization
(e.g valve implants) and vascular mineralization in dialysis patients are
needed. In many instances, analyses may be most efficiently performed by
utilizing existing clinical cohorts in which bone or vascular parameters
already have been or will be determined. Reproducible, non-destructive serial
imaging methods, such as quantitative CT scanning, need to be validated.
Characterize the relationship between lipid
metabolism and skeletal health. Clinical studies are needed to determine
whether lipid accumulation contributes to osteoporosis. Studies on the
relationship between marrow fat and body fat, marrow fat and osteogenesis, and
marrow fat and cardiovascular events may be valuable.
Determine significance of vascular calcification in
progression and outcome of cardiovascular disease. It is important to
determine whether vascular calcification is harmful or beneficial both in terms
of plaque instability and congestive heart failure. Research on the types,
distribution and anatomy of vascular calcification is needed. The
identification of serum markers for the incidence and progression of vascular
calcification, corresponding to those widely used to assess bone turnover,
would be a valuable adjunct to radiological methods. It is important to
determine whether vascular calcification alters serum markers currently
attributed to bone turnover.
Determine similarities and
differences in the mechanisms of vascular calcification and bone
mineralization. It is important to determine the cellular and molecular
biology of atherosclerotic calcification, including the roles of oxidized
lipids, cell populations, and gene products that parallel aspects of
osteogenesis. It is important to determine the effects of normal and oxidized
lipids on bone turnover. The role, if any, of cellular (e.g., osteoclastic)
resorptive mechanisms in vascular calcification should be determined. It is
important to determine the contribution of inflammatory mediators and cytokines
to cardiovascular disease and bone remodeling.
Determine the mechanisms underlying the phenotypes
of animal models in which pathological vascular calcification or
osteoporosis/osteogenesis is observed. One resource-efficient approach is
to test for cardiovascular phenotypes in animals with known bone phenotypes and
vice versa. Continued use of mouse genetics and genetically-modified mice such
as the osteoprotegerin and matrix Gla protein deficient mice could potentially
yield important new information. The observations on arterial calcification
raise the possibility of examining naturally occurring mutations of these genes
with these apparent phenotypes to determine the relationship between these gene
products in physiologic and pathologic calcification. New resources such as
other animal models, particularly larger species (e.g., swine and rabbit)
suitable for physiologic studies and those with features closely resembling
human disease would be helpful. Also, an organ culture model for vascular
calcification should be developed.
Determine the effects of statins and
bisphosphonates on the skeleton and the cardiovascular system. Side-by-side
comparisons of vascular and bone cell responses to these commonly used
therapeutic agents are needed. In vivo models are needed to explore indirect
and down-stream effects, such as the alteration of signaling molecule
maturation. It is important to determine whether bisphosphonates exacerbate or
reduce vascular calcification.
commonly used therapeutic and preventive treatments for osteoporosis have an
impact on cardiovascular disease progression and conversely, whether treatments
for atherosclerosis have an impact on skeletal health. Since many
therapeutic agents, such as estrogen, calcium supplements,
1,25-dihydroxyvitamin D, warfarin, and antioxidants potentially affect both
bone and cardiovascular function, it is essential to exclude damaging effects,
and identify the beneficial side effects. Determine whether other
lipid-lowering agents have effects on the skeleton. Small scale clinical
studies should be utilized to make these determinations.
Explore refinements of statin administration to
optimize effects on bone. Alternative approaches to clinical use of statins
should be considered, including transdermal application, utilization of statins
that are not extracted so efficiently by the liver, and possibly the use of
high doses of those statins with better biodistribution beyond the liver.
Parallel determination of the effects of these alternative approaches to statin
use on cholesterol synthesis and cardiovascular disease should be included.
John S. Adams, M.D.
Cedars-Sinai Medical Center
Los Angeles, CA
Warren S Browner, M.D., M.P.H.
Professor of Medicine, Epidemiology
& Biostatistics & Anesthesia
Veterans Affairs Medical Center 18
San Francisco, CA
Steven R. Cummings, M.D.
Professor of Medicine and Epidemiology
Univ. of California-San Francisco
San Francisco, CA
Linda Demer, M.D., Ph.D.
Professor and Chief, Division of Cardiology
University of California, Los
Los Angeles, CA
Colin R. Dunstan, Ph.D.
Thousand Oaks, CA
M. Giachelli, Ph.D.
Associate Professor, Department of Bioengineering
Univ. of Washington School of Medicine
Judith A. Hsia, M.D.
Professor, Division of Cardiology
Washington University Medical Center
Karsenty, M.D., Ph.D.
Associate Professor, Department of Molecular and
Baylor College of Medicine
Lewis H. Kuller, M.D., Dr.P.H.
Professor and Chairperson,
Department of Epidemiology
University of Pittsburgh
James K. Liao, M.D.
Assistant Professor, Cardiovascular
Brigham and Women's Hospital
R. Mundy, M.B., B.S.
J.C. and Irene H. Heyser Memorial Professor of
Bone and Mineral Metabolism
University of Texas Health Science Center
San Antonio, TX
Susan M. Ott, M.D.
Division of Metabolism Endocrinology, and Nutrition
Clay F. Semenkovich, M.D.
Associate Professor, Division of Atherosclerosis, Nutrition, and Lipid
St. Louis, MO
Stephen I. Katz, M.D.,
Claude Lenfant, M.D.
Deborah Applebaum-Bowden, Ph.D.
Joan A. McGowan,
Director, Musculoskeletal Diseases Branch
William Sharrock, Ph.D.
Director, Bone Biology Program