Association of Breast Arterial Calcification With Stroke and Angiographically Proven Coronary Artery Disease

A Meta-analysis

Xuezhi Jiang, MD, FACOG, NCMP; Maureen Clark, BS; Rupali K. Singh, MD; Alex Juhn, BA; Peter F. Schnatz, DO, FACOG, FACP, NCMP


Menopause. 2015;22(2):136-143. 

In This Article


It is unclear whether BAC, as a type of calcium deposition, is part of the larger spectrum of atherosclerosis or follows a distinct pathophysiological pathway. It has been reported that BAC prevalence significantly increases with age,[23] and it is well-known that age is also a significant risk factor for CAD or atherosclerotic heart disease. For this reason, many have doubted the association between BAC and CVD. However, the results of more recent prospective data have suggested a more convincing association.[7] Most atherosclerotic lesions contain calcium, suggesting that there may be some overlap between these two types of calcification processes.[24] Although more research on the pathophysiology of BAC is needed, there seems to be a clear association between BAC and atherosclerosis of the coronary arteries. Contrary to atherosclerotic calcifications, which are located in the intimal layer or on the inner arterial wall, BACs are depositions of calcium within the media layer or muscular wall of breast arteries or arterioles. These two types of arterial calcification may not be identical and hence may not share the same pathophysiological pathway.

An association between atherosclerosis and osteoporosis has been suggested, regardless of age, body mass index, and cardiovascular risk factors.[25,26] Several studies have documented that reduced bone mineral density independently correlated with acute stroke, angiographically documented CAD, and increased cardiovascular mortality.[27,28,29,30] Interestingly, it has also been suggested that osteoporosis and the presence of BAC are strongly correlated, independent of age and other cardiovascular risk factors such as menopause, diabetes mellitus, and hypertension.[31]

Although both types of calcification are linked to reduced bone mineral density and osteoporosis, further studies are needed to clarify the pathophysiological mechanisms behind this connection. In 1963, Trueta[32] proposed that the vascular endothelial cell itself is the osteoblast precursor. Endothelial cells in the intimal layer of the arterial wall are capable of differentiation into osteoblasts, following the same process of differentiation that occurs in bone-derived osteoblasts and consequently producing bone mineral in the arterial wall.[25] The results of other studies suggest that smooth muscle cells located in the medial layer of arteries can become chondrocytelike or osteoblast-like, mineralizing proteins and generating some form of calcification in the arterial wall.[33] Whether the relationship between osteoporosis and these two types of arterial calcification (intimal and medial) is causal has yet to be determined. This relationship seems to indicate that these two types of arterial calcification are closely associated and may even share a similar pathophysiological pathway.

The results of the current meta-analysis demonstrate a strong association between BAC and CAD/stroke. A 2012 meta-analysis of the association between BAC and CAD (diagnosed by angiogram; n = 927) reported an increased risk of CAD in women with BAC, with an OR (95% CI) of 1.59 (1.21-2.09)[34]; however, it was limited because it included only five studies with relatively small sample sizes. Our robust sample size adds to this finding and helps to establish an association between CAD and BAC. This raises a research question on whether BACs seen on routine screening mammography can help to identify those women at risk for CAD/stroke.

Although this meta-analysis demonstrates an association between BAC and CAD/stroke, it does not prove that BAC is a predictor of future CAD or stroke development. To our knowledge, only one prospective cohort study on this topic has been published. In 2011, Schnatz et al[7] followed a cohort of 1,454 women for 5 years. Among women who did not have a previous diagnosis of coronary heart disease (CHD), those who had BAC at baseline had a 3.5-fold increased risk of developing CHD compared with those who did not have BAC at baseline. We also reported that 58.3% of BAC-positive women, versus 13.3% of BAC-negative women, developed a stroke within 5 years of follow-up (P< 0.001). After controlling for other CHD risk factors (age, hypertension, hypercholesterolemia, diabetes mellitus, smoking, family history, and menopause), we still found a 2.2-fold increased risk of CHD in women with BAC. This study demonstrates the potential of BAC as an early predictor of CAD/stroke yet highlights the need for more prospective studies to validate this hypothesis.

The results of this prospective study,[7] combined with the current meta-analysis, lend more support to the hypothesis of an association between BACs and CVD (CAD and/or stroke). Some studies that yielded negative results had much smaller sample sizes; when these studies were analyzed in conjunction with similar studies, the positive association remained significant.

One of the limitations of our analysis was a lack of prospective cohort studies. All studies included in this metaanalysis were cross-sectional studies rather than cohort studies. Although these cross-sectional studies help to clarify the association between BACs and CAD/stroke, they lack the ability to predict whether BACs can be used to identify women at risk for the future development of CAD/stroke. Because we were not able to access the databases for the individual studies, detailed subgroup analyses were not performed. Because the data on risk factors for individual participants (age, hypertension, diabetes, hypercholesterolemia, etc) were likewise not available, these variables were not adjusted for while calculating the ORs. However, adjusted ORs were calculated for 4 of 10 CAD studies. The OR for all stroke studies were calculated and adjusted, except one that is noncalculable because of a zero incidence of events in the BAC-negative group. Compared with each individual crude OR, none of the adjusted ORs changed the direction of observed effect. The fact that angiography results in several studies were obtained directly from women rather than from imaging reports leads to the possibility of recall bias. In addition, the fact that six stroke studies are heterogeneous warrants careful interpretation while addressing the association of BAC with stroke.

These results also lead to research questions that need to be addressed in future studies: Should more consistent and reliable reporting guidelines on BACs seen on mammogram be advocated? Will routine reporting of BACs by radiologists help to define their possible use in the prevention and early detection of CAD/stroke?