Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Jan 1.
Published in final edited form as: Bone. 2007 Aug 22;42(1):43–52. doi: 10.1016/j.bone.2007.08.023

Calcified Atherosclerotic Plaque and Bone Mineral Density in Type 2 Diabetes

The Diabetes Heart Study

J Jeffrey Carr 1,2, Thomas C Register 3, Fang-Chi Hsu 2, Kurt Lohman 2, Leon Lenchik 1, Donald W Bowden 4,5,6, Carl D Langefeld 2, Jianzhou Xu 2, Stephen S Rich 2, Lynne E Wagenknecht 2, Barry I Freedman 4
PMCID: PMC2239236  NIHMSID: NIHMS38469  PMID: 17964237

Abstract

The purpose of the present study was to determine the relationships between atherosclerotic calcified plaque (CP) and bone mineral density (BMD) in subjects with type 2 diabetes mellitus (DM2). CP in the coronary arteries, carotid bifurcation, and abdominal aorta was measured using computed tomography (CT) in 1,023 diabetic subjects from 453 families. Trabecular volumetric BMD in thoracic (T-vBMD) and lumbar (L-vBMD) spine was measured with quantitative CT (QCT), while areal BMD (aBMD) in the lumbar spine and hip was measured by dual X-ray absorptiometry (DXA). Correlation coefficients were computed to assess the magnitude of associations and generalized estimating equations (GEE1) were used to make statistical inferences while accounting for familial correlation. Subjects were 53.8% female, 85% European American (EA) and 15% African American (AA). After adjustment for age, significant inverse associations between CP and vBMD persisted in EA men (correlations between -0.11 and -0.16, all p < 0.05 with the exception of carotid CP vs. T-vBMD, p=0.076) and in AA women, excluding aortic CP, (correlations between -0.16 and -0.25, all p < 0.05). Estrogen use in AA but not EA women was consistently associated with an inverse relation between CP and vBMD. Significant inverse relationships between CP and vBMD were observed in EA men and AA women with DM2 after adjusting for age and other covariates. QCT determined vBMD was more strongly related to CP than aBMD by DXA. The relation between CP and BMD in diabetes is influenced by age, sex, and ethnicity, with further effect modification by hormone replacement therapy.

Keywords: atherosclerosis, type 2 diabetes mellitus, osteoporosis, trabecular bone mineral density, calcium, computed tomography

Introduction

In 1964, Anderson, Barnett and Nordin first suggested that aortic calcification and low bone mineral content were related disorders(1). This prescient observation was initially met with skepticism. The common and often co-exiting processes of osteoporosis and atherosclerosis were previously considered to be independent degenerative phenomena associated with aging. Cross-sectional studies in post-menopausal women have documented the coincident occurrence of low bone mineral density (BMD) and atherosclerosis, and low bone mass has been linked to cardiovascular disease (CVD) incidence (1bc), as well as morbidity and mortality. In the placebo arm of the Multiple Outcomes of Raloxifene Evaluation (MORE) trial, post-menopausal women with low bone mass and no vertebral fractures had 8.3 incident CVD events per 1000 patient years, compared to 15.1 such events in women with prevalent vertebral fractures(2). An incremental increase in CVD events was detected for each degree of decline in BMD(3). Similarly, a 1.31-fold increase in risk of stroke was found for each standard deviation decline in BMD in the Study of Osteoporotic Fractures(4).

Longitudinal studies also demonstrate relationships between bone mass and aortic calcified plaque. Twenty five year follow-up data from the Framingham Heart study demonstrated that in women, but not men, the percentage decline in cortical bone in the second metacarpal was significantly associated with the increase in aortic calcified plaque assessed from a lateral lumbar radiograph(5). In 2348 healthy postmenopausal women, Schulz et al. reported that aortic calcified plaque assessed by computed tomography (CT) was inversely related to BMD and directly related to fracture risk(6). A significant percentage (26.1%) of the variance observed in BMD was accounted for by aortic calcified plaque after adjustment for age and other confounders. In a subgroup analysis of women with longitudinal data, the increase in aortic calcium explained 47% of the observed decline in BMD. Women in the highest quartile for gain in aortic calcified plaque had four times the annual bone loss of those in the lowest quartile(6). The inverse relationship between these variables was consistently observed regardless of the initial BMD or vascular calcium score. In another large study of 2,662 generally healthy postmenopausal women, aortic calcification was associated with lower BMD, accelerated bone loss in the proximal femur, and future risk of fracture in the hip (7).

Diabetes may influence the relationships between CP and BMD. Prior studies have shown that older diabetic subjects have higher dual X-ray absorptiometry (DXA)-measured bone mass than non-diabetic subjects. Despite this observation, the fracture risk in diabetic subjects was 1.64 times higher after adjustment for hip BMD and other fracture risk factors(8-11). Our prior work with the Diabetes Heart Study (DHS) families indicates that BMD (areal BMD by DXA as well as volumetric BMD of vertebral trabecular bone by Quantitative CT [QCT]) is not independently associated with the presence of diabetes after adjustment for body mass index (BMI) and other covariates(12). Available data suggests that DM2 does not lead to lower BMD by either DXA or QCT in the spine, yet diabetics have increased CVD risk and a higher burden of CP.

In the present study we evaluated whether the inverse relationship often observed between CP and BMD is also present in subjects with DM2, specifically with respect to BMD in various skeletal compartments (e.g. vertebral trabecular bone, areal measures of BMD in the hip and spine) in diabetic DHS participants.

Methods

Study Population

The DHS recruited a bi-racial cohort of families containing ≥ two siblings with DM2. Probands were identified through community advertising and local clinics in northwestern North Carolina. Diabetic index cases diagnosed after the age of 34 years, in the absence of historical evidence of diabetic ketoacidosis and having one additional type 2 diabetic sibling were recruited. This report is limited to observations in type 2 diabetic participants, defined as above and receiving clinical treatment for hyperglycemia (oral agents and/or insulin) or having a fasting blood glucose > 126 mg/dl. The study was approved by the Institutional Review Board at the Wake Forest University School of Medicine and all participants provided written informed consent.

The examination included interviews for medical history, current medications and health behaviors. We did not adjust CP for use of anti-resorptive agents (i.e., bisphosphonates, calcitonin, selective estrogen receptor modulators) since fewer than 2% of DHS participants took these agents. We also did not adjust for glucocorticoid usage, taken by 6% of DHS participants, since they did not impact BMD in these analyses (data not shown). Measurements of body size, resting blood pressure, 12-lead electrocardiogram, fasting blood draw and urine collection were conducted in the General Clinical Research Center. History of CVD was provided by participant self-report. We excluded subjects with elevated serum creatinine concentrations (≥ 1.6 mg/dl in men, ≥ 1.4 mg/dl in women) due to potential effects of kidney failure on BMD and CP. Stratified analyses were performed in those reporting prior coronary artery bypass surgery or coronary artery stents (however, these procedures did not significantly alter associations).

Vascular Imaging

Calcified atherosclerotic plaque was measured in the coronary arteries using single and multidetector CT systems incorporating a standardized protocol based on those currently implemented in the National Heart Lung and Blood Institute’s (NHLBI) Multi-Ethnic Study of Atherosclerosis (MESA) studies(13). From the initiation of CT exam in 1999, three versions of the General Electric CT system have been used (CTi, LightSpeed QXi & Pro16 system, General Electric Medical Systems, Waukesha, WI) capable of 520 msec, 520 msec and 244 msec temporal resolutions, respectively, when operating in the cardiac mode. The robustness of the coronary calcium score using various CT systems with differing temporal resolutions in this range has been established(13-15). Other technical factors were: 120 KVp, 106 mAs (weight < 220 lbs) or 133 mAs (weight ≥ 220 lbs) and 2.5 mm slice collimation (3 mm for the CTi system) and reconstruction with the standard kernel. Images were obtained during suspended respiration and with ECG gating at 50% of the RR interval. A calibration standard (Image Analysis, Columbia, KY) was placed underneath each participant during the scans. In addition to daily calibrations, twice a month calibration scans for measurement of calcium hydroxyapatite were performed and recorded to document stability of the CT systems measurement of CT numbers and calcium concentrations to document the absence of temporal drift. Calcified plaque measured in each of the epicardial coronary vessels was summed to create the total calcified plaque burden. Two cardiac scans were performed sequentially and the average of the two measurements was used in analysis. For the carotid exam, an un-enhanced CT scan was performed through the neck from C2-3 to the C6-7 disc levels covering the right and left carotid bifurcations using a helical acquisition with 2.5-mm or 3-mm slice collimation depending on CT scanner capability, a 120 kv, 280 mA, 0.8 second gantry rotation, 360 degree scan reconstruction and standard reconstruction kernel. The display field of view was 18 cm resulting in pixel dimensions of 0.35 mm × 0.35 mm. To measure calcified plaque in the abdominal aorta, an un-enhanced scan of the abdomen was performed. The technical factors for this series were 120 kV, 250 mA, 0.8 second gantry rotation helical mode (7.5 mm/s), 2.5 mm slice thickness and standard reconstruction kernel. The display field of view was 35 cm, resulting in a pixel dimension of 0.68 by 0.68 mm.

CT exams in these vascular beds were analyzed by experienced analysts producing measures of calcified plaque the Agatston score corrected for slice thickness on a GE Advantage Windows Workstation using the SmartScores software package (General Electric Medical Systems, Waukesha, WI). The reproducibility of coronary and carotid calcified plaque scores obtained from the duplicate scans and inter and intra-observer variability were all > 0. 96. As previously reported by our group, helical CT scans yield comparable vascular calcium measurements as electron beam CT (EBCT)(15).

Bone Imaging

DXA scans of posterior-anterior spine and proximal femur obtained using a fan-beam scanner (Delphi A, Hologic, Waltham, MA, USA) to determine BMD(16). Quantitative computed tomography (QCT) trabecular BMD of the thoracic and lumbar vertebrae were measured using the same CT images obtained for measuring calcified plaque in the coronary and abdominal aorta using our previously published methods(17). Volumetric BMD (vBMD) was measured in the thoracic spine (T8-T11) and lumbar spine (T12-L3) using QCT-5000 software (Image Analysis, Columbia, KY). The software imports the CT data creating a three dimensional volume. The operator selected the appropriate mid-vertebral slice for BMD analysis by angling the analysis plane as appropriate for each vertebral level. The program calculated the BMD in mg/cm3 on a level-by-level basis. After all the images were measured, these values were then averaged to provide a mean BMD for that entire location, either thoracic or lumbar spine.

Statistical Methods

The sample means, standard deviations and medians were computed for the continuous characteristics, BMD and CP. The proportions were calculated for the discrete demographic characteristics. To compare the study population characteristics between ethnicity and sex groups, the generalized estimating equation (GEE1) procedure (18) was used to account for the correlated data structure inherent in a study using siblings. Each characteristic was treated as the dependent variable. To test for the main effects, ethnicity and sex were treated as the covariates; to test for the interaction between these two variables, additional interaction term was fitted in the model. Sex-stratified analysis was performed to test for the main effect of ethnicity; ethnicity-stratified analysis was performed to test for the main effect of sex. Spearman correlation coefficients were computed to estimate the magnitude of the association between BMD (using QCT in the thoracic and lumbar spine, DXA in the hip and spine) and vascular calcium. Partial correlation coefficients were calculated to adjust for the potential common effects of age, ethnicity, sex, BMI, smoking and prevalent CVD. Criteria used to define prevalent CVD included Q-wave abnormalities on examination, self-reported evidence for clinical CVD (angina, myocardial infarction, stroke), or prior vascular procedures (i.e. coronary artery bypass graft, percutaneous transluminal angioplasty, endarterectomy). The partial correlation coefficients were obtained via computing the Spearman’s rank correlation coefficient between the residuals from models regressing vascular calcium and BMD, respectively, onto the covariates. Analyses were performed in the entire sample, separately by sex and ethnicity, and in post-menopausal women (defined as age ≥ 55 years) by estrogen use. Simple correlation coefficient tests were deemed invalid due to the correlated data structure and were re-evaluated using the GEE1 procedure. In order to better approximate the distributional assumptions of conditional normality and homogeneity of variance, the natural log of CP values plus a constant of one were analyzed. Standard regression diagnostics for collinearity and influence were computed for each model. All statistical analyses were considered significant at p<0.05. SAS software (Cary, NC) was used for the statistical analyses.

Results

The characteristics of the sex- and ethnicity-stratified study population are presented in Table 1. There are 1,023 DHS participants with DM2 from a total of 453 families. Overall, participants had a mean age of 62.4 years, mean BMI 32.7 kg/m2, mean diabetes duration 11.7 years and sub-optimal control of fasting blood sugar, HbA1c, systolic blood pressure, triglycerides and LDL cholesterol.

Table 1.

1a: Characteristics of Study Population by Ethnicity and Sex

1b:Tests for Differences in Population Characteristics by Ethnicity and Sex1

African America (AA) European American (EA)
Women Men Women Men
Variable N Mean SD Median N Mean SD Median N Mean SD Median N Mean SD Median
Age, yrs 110 59.6 9.1 58.5 53 59.9 9.1 60.9 440 62.5 9.5 62.4 420 63.4 8.9 64
DM Duration, yrs 110 11.6 7.9 9.5 53 12.3 8.6 10.2 431 11.3 7.1 9.7 410 12.0 7.7 10.2
HbA1c, yrs 110 8.8 2.7 8.1 53 8.4 2 8 440 7.6 1.8 7.2 419 7.5 1.4 7.2
BMI, kg/m2 110 35.2 7.6 34.4 53 31.2 6.4 30.8 439 33.7 7.6 32.8 417 31.3 5.9 30.5
SBP, mm Hg 110 146.6 24.1 143.5 53 142.5 19.4 143 439 140.7 19.2 139 417 138.8 19.1 137.5
DBP, mm Hg 110 76.5 11.3 75.8 53 76.1 10.7 78 439 70.5 9.9 69.5 417 72.6 9.9 72.5
Total chol, mg/dl 108 196 45 192 52 178 33 174 438 196 46 190 416 176 40 173
HDLc, mg/dl 108 52 16 50.5 52 45 12 44.5 438 46 13 44 415 38 10 37
LDLc, mg/dl 106 116 41 112.5 49 107 29 105 404 107 35 104 391 100 32 97
non-HDLc, mg/dl 108 144 44 142 52 133 36 128.5 438 150 45 143.5 415 137 39 132
Triglycerides, mg/dl 108 141 81 120.5 52 135.8 119 100 438 220.3 150 187 416 194 124 165
Calcified Plaque
  Coronary 105 282 631 12.5 49 509 1016 72 413 371 630 68 392 1194 1434 659.8
  Carotid 102 67 185 3.8 50 69 181 6 411 88 169 16.5 389 151 212 69.5
  Abdominal Aorta 97 1687 2720 491 48 1757 3953 418.5 366 3062 3809 1394 361 5303 5772 3663
Bone
  QCT Spine T-vBMD 103 180 51 174 50 160 40 163 406 146 45 140.9 392 135 37 134.9
  QCT Spine L-vBMD 95 163 47 156.6 48 152 37 156.5 362 131 43 129.5 359 122 35 119.2
  DXA Spine L-aBMD 86 1.07 0.19 1.08 45 1.10 0.17 1.10 356 1.02 0.18 1.01 345 1.09 0.17 1.08
  DXA-Total Hip aBMD 85 0.99 0.17 0.99 43 1.07 0.18 1.06 359 0.95 0.17 0.93 344 1.02 0.16 1.02
Smoking (%)
  Current 18 16.5% 22 41.5% 66 15% 73 17.4%
  Past 34 31.2% 24 45.3% 124 28.2% 245 58.3%
Medications (%)
  Lipid Lowering 49 44.5% 18 34.0% 201 46.0% 222 53.5%
  Estrogen 16 15.4% - 111 26.2% -
  Anti-hypertensive 69 82.1% 36 80.0% 261 82.1% 248 77.0%
  Thiazide Diuretic 30 28.3% 15 28.8% 115 27.3% 60 15.0%
Main Effects AA EA Men Women
Variable race sex Interaction Δsex Δsex Δrace Δrace
Age, yrs 0.0001 0.14 0.95 0.62 0.16 0.01 0.0014
DM Duration, yrs 0.79 0.19 0.99 0.65 0.22 0.89 0.81
HbA1c, yrs 0.0001 0.07 0.14 0.14 0.22 0.0001 0.0001
BMI, kg/m2 0.09 0.0001 0.10 0.0003 0.0001 0.70 0.03
SBP, mm Hg 0.0001 0.15 0.20 0.16 0.38 0.15 0.0001
DBP, mm Hg 0.0001 0.0018 0.0364 0.50 0.0002 0.02 0.0001
Total chol, mg/dl 0.83 0.0001 0.72 0.0112 0.0001 0.65 0.98
HDLc, mg/dl 0.0001 0.0001 0.70 0.0029 0.0001 0.0001 0.0001
LDLc, mg/dl 0.0019 0.0019 0.62 0.13 0.0060 0.13 0.01
non-HDLc, mg/dl 0.13 0.0001 0.79 0.13 0.0001 0.47 0.20
Triglycerides, mg/dl 0.0001 0.0014 0.24 0.89 0.0012 0.0013 0.0001
Calcified Plaque
  Coronary 0.0006 0.0001 0.0053 0.0472 0.0001 0.0013 0.0574
  Carotid 0.0053 0.0001 0.09 0.95 0.0001 0.0061 0.17
  Abdominal Aorta 0.0001 0.0001 0.0203 0.78 0.0001 0.0001 0.0003
Bone
  QCT Spine T-vBMD 0.0001 0.0001 0.16 0.0082 0.0001 0.0001 0.0001
  QCT Spine L-vBMD 0.0001 0.0001 0.53 0.07 0.0005 0.0001 0.0001
 DXA Spine L-aBMD 0.0340 0.0001 0.14 0.75 0.0001 0.93 0.01
 DXA-Total Hip aBMD 0.0037 0.0001 0.72 0.0323 0.0001 0.0365 0.04
Smoking (%)
  Current 0.0127 0.0137 0.0037 0.0002 0.33 0.0001 0.98
  Past 0.40 0.0001 0.09 0.06 0.0001 0.06 0.68
Medications (%)
  Lipid Lowering 0.08 0.35 0.10 0.24 0.14 0.0182 0.63
  Estrogen - - - - 0.0081
  Anti-hypertensive 0.76 0.11 0.73 0.77 0.11 0.65 0.99
  Thiazide Diuretic 0.20 0.0001 0.06 0.97 0.0001 0.0244 0.95
Open in a new tab

DM - diabetes mellitus; HbA1c - glycosylated hemoglobin; BMI - body mass index; SBP - systolic blood pressure; DBP - diastolic blood pressure; chol and c - cholesterol.

1

Please see text for statistical methods

Systolic blood pressure, diastolic blood pressure, hemoglobin A1c, LDL-cholesterol, HDL-cholesterol and spine vBMD were all higher in AAs compared to EAs, despite similar durations of DM2 and anti-hypertensive medication treatment (Table 1). AA men had the highest prevalence of current smokers, EA men had the highest level of past smoking. Serum triglyceride concentrations were higher in EA women than EA men, and higher in EA than in AA counterparts.

Calcified atherosclerotic plaque, reported as the total calcium score using Agatston Units, averaged 705 for the coronary arteries, 111 for the carotid bifurcations, and 3765 for the infra-renal abdominal aorta. When stratified by ethnicity and sex, EA men had a significantly higher burden of CP in all three vascular territories, EA women had the second highest burden of CP (after EA men) in the carotid bifurcation and abdominal aorta, while AA men had the second highest burden of coronary CP (Table 1). Overall, AA women had the lowest burden of CP in the three vascular territories. The differences in CP plaque were significant in all ethnicity sex comparisons with a significant interaction for both coronary and aortic CP. Carotid CP was not significantly different in AA by sex or between AA & EA women and no interaction was present (Table 1b).

BMD measured by DXA and QCT demonstrated significant differences by race and sex and no significant race by sex interactions were observed (Table 1b). For the vBMD measured by QCT, AA women had the highest trabecular vBMD in the spine, followed by AA men, EA women and then EA men. For the DXA measures of the spine and hip aBMD, AA men had the highest aBMD, followed by EA men, AA women, and then EA women.

Correlations between CP in the coronary, carotid and abdominal aorta territories and BMD using QCT (thoracic and lumbar spine) and DXA (hip and spine) are presented in Table 2 as unadjusted (model 1), adjusted for age (model 2), and adjusted for age, sex, ethnicity, BMI, smoking and prevalent CVD (model 3). Inverse relations were present between CP and BMD across multiple vascular beds in all ethnic and sex groups with the QCT measure of vBMD and to a limited extent with the DXA measures of aBMD. After adjustment for age, significant inverse correlations between CP and vBMD in multiple sites were observed primarily in EA men and AA women, although there was a tendency for the relation in EA women as well.

Table 2.

Spearman Correlation Coefficients for Arterial Calcified Plaque versus Bone Mineral Density by Ethnicity and Sex

African America European American
Women Men Women Men
Artery Measure Bone Measure Model N R* GEE p-value N R* GEE p-value N R* GEE p-value N R* GEE p-value
Coronary QCT Spine T-vBMD 1 108 -0.306 0.0001 50 -0.224 0.2260 437 -0.246 0.0001 407 -0.220 0.0001
2 108 -0.192 0.0400 50 0.008 0.6158 437 -0.059 0.2973 407 -0.126 0.0271
3 96 -0.193 0.1065 42 -0.031 0.7891 401 -0.024 0.7597 390 -0.156 0.0189
QCT Spine L-vBMD 1 100 -0.364 0.0001 48 -0.333 0.0425 392 -0.204 0.0005 374 -0.246 0.0001
2 100 -0.250 0.0116 48 -0.078 0.8834 392 0.000 0.7236 374 -0.110 0.0268
3 88 -0.262 0.0138 40 -0.077 0.6091 358 0.012 0.8514 358 -0.113 0.0608
DXA Spine L-aBMD 1 90 -0.292 0.0013 44 -0.048 0.6891 376 0.008 0.9802 352 0.134 0.0333
2 90 -0.264 0.0054 44 0.001 0.9112 376 0.124 0.0515 352 0.139 0.0112
3 79 -0.222 0.1020 39 0.088 0.8782 345 0.107 0.0839 337 0.082 0.1156
DXA Hip aBMD 1 89 -0.119 0.2192 42 -0.355 0.0396 379 -0.151 0.0359 351 0.043 0.9588
2 89 0.000 0.7732 42 -0.231 0.1870 379 0.012 0.5753 351 0.101 0.1406
3 78 -0.074 0.9363 37 -0.100 0.8075 347 0.043 0.2741 336 0.013 0.7369
Carotid Bifurcation QCT Spine T-vBMD 1 105 -0.350 0.0001 51 -0.161 0.0944 436 -0.223 0.0001 405 -0.228 0.0001
2 105 -0.164 0.0183 51 0.001 0.6174 436 -0.012 0.2854 405 -0.118 0.0757
3 94 -0.184 0.0063 43 0.054 0.5600 400 -0.017 0.2608 388 -0.106 0.0781
QCT Spine L-vBMD 1 97 -0.357 0.0001 49 -0.197 0.2679 392 -0.196 0.0002 372 -0.296 0.0001
2 97 -0.159 0.0493 49 -0.027 0.9988 392 0.031 0.9932 372 -0.148 0.0106
3 86 -0.182 0.0339 41 0.080 0.8442 358 0.005 0.5285 356 -0.131 0.0250
DXA Spine L-aBMD 1 87 -0.084 0.6203 45 0.104 0.2015 376 -0.018 0.5934 348 0.053 0.6229
2 87 -0.049 0.4097 45 0.148 0.1558 376 0.106 0.1422 348 0.050 0.3719
3 77 -0.036 0.6214 40 0.206 0.3377 345 0.061 0.4438 333 0.045 0.4326
DXA Hip aBMD 1 86 -0.187 0.1499 43 -0.064 0.9256 378 -0.209 0.0001 347 -0.057 0.4298
2 86 -0.074 0.9276 43 0.040 0.7532 378 -0.037 0.3240 347 0.004 0.6569
3 76 -0.079 0.6255 38 0.275 0.0001 346 -0.084 0.0613 332 -0.012 0.9009
Abdominal Aorta QCT Spine T-vBMD 1 100 -0.348 0.0069 49 -0.253 0.0350 392 -0.357 0.0001 378 -0.259 0.0001
2 100 -0.119 0.6689 49 -0.074 0.5597 392 -0.085 0.0628 378 -0.146 0.0130
3 88 -0.136 0.3982 41 -0.102 0.6149 359 -0.045 0.2344 362 -0.115 0.0391
QCT Spine L-vBMD 1 100 -0.383 0.0009 49 -0.242 0.1054 393 -0.339 0.0001 377 -0.316 0.0001
2 100 -0.148 0.4544 49 -0.031 0.8963 393 -0.045 0.2159 377 -0.157 0.0160
3 88 -0.171 0.2511 41 0.006 0.5632 359 -0.030 0.4135 361 -0.129 0.0625
DXA Spine L-aBMD 1 90 -0.122 0.4096 45 -0.020 0.6349 377 -0.099 0.1514 353 0.082 0.2653
2 90 -0.066 0.8400 45 0.032 0.8892 377 0.071 0.3045 353 0.079 0.1769
3 79 -0.018 0.7030 40 0.137 0.6380 346 0.049 0.5604 338 0.066 0.2482
DXA Hip aBMD 1 89 -0.146 0.0663 43 -0.222 0.2389 380 -0.269 0.0001 352 -0.031 0.1215
2 89 0.008 0.7277 43 -0.106 0.9861 380 -0.034 0.4475 352 0.036 0.9282
3 78 0.042 0.6893 38 0.219 0.0032 348 0.016 0.8548 337 0.006 0.7442
Open in a new tab

Models: 1) unadjusted; 2) adjusted for age; 3) adjusted for age, smoking, BMI, and prevalent cardiovascular disease.

In EA men an inverse association was present between the vascular sites (coronary, carotid and abdominal aorta) and the two sites of vBMD (thoracic and lumbar) before and after age adjustment (model 1 with 6 of 6 associations demonstrating inverse correlations with p<0.05; model 2 with 5 of 6 associations demonstrating inverse correlations with p <0.05 and 1 borderline p=0.076) and in the expanded model adjusted for age, sex, ethnicity, BMI, smoking and prevalent CVD (model 3 with 3 of 6 associations demonstrating inverse correlations with p<0.05 and highest borderline p=0.078). In AA women an inverse association is present between the vascular sites (coronary, carotid and abdominal aorta) and the two sites of vBMD (thoracic and lumbar) before adjustment (model 1 with 6 of 6 associations demonstrating inverse correlations with p<0.05). After age adjustment, the relation between aortic CP and vBMD remains negative but is no longer significant. However, AA women continue to have a significant inverse relation between coronary and carotid CP and vBMD after age adjustment and in the expanded model adjusted for age, sex, ethnicity, BMI, smoking and prevalent CVD (model 2 with 4 of 4 associations demonstrating inverse correlations with p <0.05; model 3 with 3 of 4 associations demonstrating inverse correlations with p <0.05 and the fourth p=0.10). In EA women, significant associations were observed only in the unadjusted models, once accounting for age (or other covariates) no associations between CP and BMD are observed although the relation between thoracic vBMD and aortic CP approached significance (p=0.062). In AA men, this study had limited power to detect an association secondary to the smaller sample size. Prior to age adjustment significant inverse associations between CP and vBMD were observed (model 1, 3 of 6 p < 0.05 and 2 of 6 p < 0.11). All of these associations become non-significant after adjusting for age.

DXA-derived aBMD measures were less well correlated with CP. Significant relationships between aBMD and CP were observed in AA women and EA men. In AA women, the inverse associations between spine aBMD and coronary artery CP were present in the unadjusted and age adjusted models, but not in the fully adjusted model. In EA men, unadjusted and age adjusted associations between spine aBMD and coronary CP were positive, but again not present in the fully adjusted model. In EA women, hip aBMD was negatively associated with coronary, carotid, and aortic CP only in the unadjusted model. In AA men, DXA aBMD of the hip is inversely associated with coronary CP prior to age adjustment and with abdominal aortic CP only in model 3, observations based on a smaller number of individuals.

Stratified analyses were performed in post-menopausal women based upon estrogen use, since this factor is known to influence BMD and atherogenesis (Table 3). AA women reporting use of estrogen had significant inverse correlations between coronary, carotid and abdominal aortic CP with thoracic vBMD after adjustment for age, and these remained significant after further adjustment for BMI, smoking and prevalent CVD. Coronary artery CP and abdominal aorta CP were also inversely associated with L-vBMD in AA women taking estrogen. Among EA estrogen users, inverse relationships between carotid CP with T-vBMD and L-vBMD and hip aBMD were observed before, but not after adjustment for age, although a significant relationship was observed with L-vBMD in the fully adjusted model. Among estrogen non-users, inverse relationships between CP and BMD at multiple sites were accounted for by adjustment for age, except that coronary CP and spine aBMD remained significantly associated in AA women.

Table 3.

Spearman Correlation Coefficients for Arterial Calcified Plaque versus Bone Mineral Density in Post-Menopausal Women by Ethnicity and Estrogen Use

African America Women European American Women
Estrogen (Yes) Estrogen (No) Estrogen (Yes) Estrogen (No)
Artery Measure Bone Measure Model N R* GEE p-value N R* GEE p-value N R* GEE p-value N R* GEE p-value
Coronary QCT Spine T-vBMD 1 17 -0.399 0.0012 77 -0.168 0.0685 120 -0.241 0.2573 280 -0.186 0.0041
2 17 -0.402 0.0001 77 -0.092 0.3056 120 -0.03 0.3981 280 -0.066 0.2938
3 16 -0.639 0.0020 70 -0.064 0.4552 105 0.055 0.1949 262 -0.049 0.4557
QCT Spine L-vBMD 1 16 -0.203 0.0003 70 -0.216 0.0677 106 -0.154 0.4358 254 -0.192 0.0304
2 16 -0.159 0.0001 70 -0.143 0.2739 106 0.09 0.4343 254 -0.057 0.6644
3 15 -0.628 0.0005 63 -0.113 0.5124 93 0.162 0.1418 235 -0.061 0.7519
DXA Spine L-aBMD 1 15 -0.131 0.1875 65 -0.242 0.0317 102 0.08 0.3721 246 0.001 0.8384
2 15 -0.179 0.0706 65 -0.225 0.0329 102 0.169 0.2449 246 0.08 0.1912
3 14 -0.753 0.0005 58 -0.054 0.7826 90 0.226 0.1191 228 0.043 0.2931
DXA Total Hip aBMD 1 15 0.155 0.7922 64 -0.104 0.4364 102 -0.104 0.818 248 -0.143 0.143
2 15 0.200 0.0840 64 0.001 0.8717 102 0.081 0.1683 248 -0.034 0.9645
3 14 -0.468 0.1673 57 0.059 0.4221 90 0.236 0.0338 229 -0.031 0.7219
Carotid Bifurcation QCT Spine T-vBMD 1 17 -0.764 0.0358 75 -0.236 0.0002 119 -0.309 0.0002 280 -0.115 0.0081
2 17 -0.833 0.0949 75 -0.035 0.1047 119 -0.146 0.2307 280 0.075 0.8453
3 16 -0.851 0.0110 68 -0.053 0.3322 104 -0.147 0.1231 262 0.06 0.9646
QCT Spine L-vBMD 1 16 -0.522 0.0576 68 -0.259 0.0004 106 -0.257 0.0161 254 -0.096 0.1491
2 16 -0.499 0.2022 68 -0.081 0.0609 106 -0.077 0.716 254 0.117 0.345
3 15 -0.594 0.0285 61 -0.053 0.5059 93 -0.102 0.0148 235 0.085 0.9954
DXA Spine L-aBMD 1 15 -0.193 0.5769 63 -0.029 0.7168 102 -0.039 0.6423 246 0.08 0.3423
2 15 -0.361 0.1270 63 0.003 0.1993 102 0.022 0.8388 246 0.193 0.0276
3 14 -0.423 0.0066 56 0.122 0.5879 90 -0.034 0.3964 228 0.146 0.1234
DXA Total Hip aBMD 1 15 -0.211 0.2845 62 -0.161 0.169 102 -0.238 0.0033 248 -0.155 0.0552
2 15 -0.140 0.2835 62 -0.031 0.6653 102 -0.103 0.5779 248 -0.005 0.7392
3 14 -0.329 0.1458 55 0.018 0.6164 90 -0.177 0.1968 229 -0.037 0.3203
Abdominal Aorta QCT Spine T-vBMD 1 16 -0.407 0.0002 70 -0.243 0.0895 106 -0.439 0.0234 254 -0.279 0.0001
2 16 -0.404 0.0001 70 -0.005 0.8701 106 -0.222 0.8593 254 -0.037 0.502
3 15 -0.624 0.0001 63 0.044 0.6338 93 -0.212 0.2267 236 0.009 0.9526
QCT Spine L-vBMD 1 16 -0.247 0.0003 70 -0.289 0.1241 106 -0.335 0.0602 255 -0.29 0.0016
2 16 -0.186 0.0001 70 -0.056 0.758 106 -0.064 0.8742 255 -0.031 0.9699
3 15 -0.667 0.0001 63 0.02 0.3131 93 -0.094 0.6593 236 -0.011 0.8261
DXA Spine L-aBMD 1 15 -0.043 0.0001 65 -0.084 0.5477 102 -0.04 0.5956 247 -0.069 0.8624
2 15 -0.112 0.0001 65 -0.04 0.7094 102 0.059 0.3876 247 0.073 0.0992
3 14 -0.564 0.0036 58 0.23 0.1545 90 0.025 0.7098 229 0.058 0.1185
DXA Total Hip aBMD 1 15 0.094 0.0888 64 -0.178 0.3849 102 -0.296 0.1932 249 -0.219 0.1338
2 15 0.160 0.0001 64 -0.002 0.5246 102 -0.107 0.3226 249 -0.019 0.6042
3 14 -0.445 0.0145 57 0.176 0.1634 90 -0.042 0.5871 230 0.021 0.2188
Open in a new tab

Post-menopausal is defined as age > 55 years. Models: 1) unadjusted; 2) adjusted for age; 3) adjusted for age, smoking, BMI, and prevalent cardiovascular disease.

Discussion

Type 2 diabetes mellitus, osteoporosis and CVD are major causes of morbidity and mortality worldwide. Calcified plaque and bone mineralization can be precisely quantified, and their relationships analyzed non-invasively, to provide insight for disease prevention and therapy. This analysis is the first to define the relationship between BMD and CP using extensive phenotyping of both traits in a large high risk population that included men and subjects with DM2. A particular strength of this report was our ability to measure BMD using two complimentary approaches. DXA measurement of aBMD is the de facto standard for determining BMD in both clinical practice and the research setting. QCT provides a measurement of vBMD specific for the metabolically active trabecular bone in the vertebrae. Our findings indicate that the relationship between calcified atherosclerotic plaque and bone mineralization in diabetic individuals is complex and influenced by a variety of factors. In addition, vertebral trabecular bone measured by QCT (vBMD) but not the more commonly measured and reported combined cortical and trabecular bone of the spine or hip (aBMD measured by DXA) demonstrated significant associations with CP at multiple vascular sites. The strongest and most consistent inverse associations between vBMD and CP were observed in diabetic EA men. EA men had the highest calcified atherosclerotic plaque burden at all three vascular sites and the lowest vertebral trabecular vBMD of all sex-ethnic groups (all contrasts highly significant - see Table 1b). AA women also demonstrated significant inverse associations between CP and BMD, and AA women had the highest average vertebral trabecular bone (vBMD by QCT) and the lowest calcified plaque burden. Stratified analyses in AA women taking estrogen demonstrated significant inverse association between coronary, carotid and aortic CP and thoracic vBMD and between coronary and aortic CP and lumbar vBMD (all p<0.05). Significant age-independent associations between BMD and CP in the larger sample of EA women, even when stratified by estrogen use, were largely absent except for the carotid bifurcation and lumbar spine. The complex association between BMD and CP in women with potential effect modification by ethnicity and estrogen use is interesting and requires further study. Estrogen use is associated with reduced atheromatous plaque formation in animal models and with reduced incidence of clinical CHD and plaque progression in younger women (19), while being a potent bone protectant known to increase bone mass in women (20). Recent results from the Women’s Health Initiative demonstrated a significant reduction in coronary artery calcified plaque burden among women 5o-59 years of age when randomized to estrogen compared to placebo(20a) Thus, estrogen use may strengthen the inverse relation between bone density and CP. It is not clear why such an effect would be observed in AA women and not in EA women in this study. We have insufficient prescribing and compliance data to carefully examine the influence of dose, duration of treatment, or type of estrogen or progestogen on the relation between bone and CP. Skeletal status determined using aBMD (DXA) of the hip and spine captures the combined cortical and trabecular BMD, whereas vBMD by QCT of the thoracic and lumber vertebra provides measures of trabecular vBMD while avoiding interference from cortical bone, osteoarthritic changes or adjacent vascular (or other) calcifications. Our results demonstrated stronger associations between CP and trabecular vBMD measured using QCT, than with aBMD using DXA. This may be related to the fact that the trabecular bone compartment is more metabolically active than the cortical compartment due to its larger surface area and exposure. In this population of relatively overweight or obese subjects, biomechanical influences on bone remodeling may have an exaggerated influence on site-specific vBMD as well as aBMD, which may alter relationships between BMD and CP.

Atherosclerosis was assessed by measuring CP at three important vascular territories directly associated with the clinical expression of CVD: the carotid bifurcation, abdominal aorta and coronary arteries. Coronary artery CP measured by cardiac CT has been demonstrated to add significantly to traditional risk factors in the prediction of near-term coronary artery disease events, CVD events and all-cause mortality in adults. We detected significant negative associations between CP in the coronary, carotid and aortic territories with vBMD in EA men and AA women. Associations between CP in these vascular beds and CVD risk factors have been demonstrated(21). Carotid bifurcation plaques are the proximate cause of embolic stroke and aorto-iliac atherosclerosis is associated with the clinical expression of symptomatic peripheral arterial disease and risk of future coronary heart disease events(22).

Our results show the relations between sub-clinical atherosclerosis and BMD vary by an individual’s ethnicity and sex. Significant ethnicity and sex differences in coronary CP have previously been reported in the Multi-Ethnic Study of Atherosclerosis (MESA) population. In the MESA study population, recruited at 6 sites across North America and selected based upon the absence of clinical CVD, the relative risk of an AA having coronary calcification was 0.78 compared to EA participants(23). Among MESA participants with coronary CP, EAs had the greatest CP burden, followed by Chinese, Hispanic and AA participants. Similar lower prevalence of coronary CP was observed in the subset of diabetic MESA participants(24), a report from Los Angeles(25), and previously reported by us in our Diabetes Heart Study families(26), where the most marked ethnic differences in CP were observed among men.

Evidence for a systematic and possibly genetic process controlling the process of calcification in arteries and bone is supported by the marked ethnic variation in the prevalence of osteoporosis(27) and calcified vascular plaque (23-26). African Americans, who typically ingest less dietary calcium than European Americans, nonetheless manifest increased BMD, skeletal resistance to the effects of parathyroid hormone and lower levels of calcified coronary artery plaque. African Americans appear to preferentially deposit calcium in the skeletal system. The pathologic condition of atherosclerotic calcified plaque is less prevalent and extensive in both AA men and women. Given equal access to medical care, African Americans have a reduced risk of bone fractures(28) and myocardial infarction(29-31), relative to European Americans.

The Study of Osteoporotic Fractures demonstrated that BMD using DXA in the heel, femoral neck and forearm was higher in older diabetic women (mean age 72 years), compared to non-diabetic women(8). However, diabetic women in this report had higher fracture rates related to factors that increased the risk of fractures. In the Health ABC study, older diabetic subjects (age 70-79 years) had higher baseline hip BMD using DXA(11). Longitudinal data from Health ABC indicated that the 4 year change in femoral neck BMD (a site of trabecular bone measured using DXA) declined significantly only in EA diabetic women. No decline in trabecular BMD was observed in AA women, AA men or EA men. Previously identified variation in BMD and atherosclerosis between sexes and ethnic groups support differential relationships between BMD and CP based upon sex and ethnicity.

We have previously demonstrated the independent heritability of BMD and calcified atherosclerotic plaque(32) in Diabetes Heart Study families. It is likely that genetic determinants influence the association between calcified atherosclerotic plaque and BMD. Several biologic pathways, particularly those involving mineralization inhibitors, may participate in the propensity for development of osteoporosis with atherosclerosis. For example, fetuin-A, matrix GLA protein and pyrophosphates all inhibit calcium deposition and are present or active in blood vessel walls(33). It is possible that polymorphisms in these mineralization inhibitor genes may account for a portion of the risk of atherosclerosis and osteoporosis. In the Coronary Artery Disease in Young Adults (CARDIA) study, single nucleotide polymorphisms (SNPs) in the osteopontin and matrix GLA protein genes were not associated with BMD and calcified atherosclerotic plaque(34), although CARDIA was perhaps hampered by a relatively young study population with limited expression of calcified atherosclerotic plaque or osteoporosis. In rodents, statins activate the expression of bone morphogenic protein 2 (BMP-2)(35) and variable results on BMD have been observed in humans following statin therapy (36).

There are several limitations in the interpretation of our results. First, the DHS is a genetic study, by design based on a family sampling strategy. To prevent overstating statistical significance, we present only a conservative definition of significance to account for the inherent correlation that is present within family members. Second, the observations are limited to individuals with DM2. It is possible and perhaps even likely that the observed relations in diabetics would be different in non-diabetic subjects. The complexity of the relations between bone mineralization and calcified atherosclerotic plaque, as well as the ethnic-specific and sex-specific findings, support the contention that genetic causes likely impact the relations between bone and atherosclerotic calcified plaque. Replication of our results in a large non-diabetic population would be informative as to whether these relations are specific to diabetes. Furthermore, evaluating the relations between CP and BMD in non-diabetic populations will likely provide greater insight to the biologic pathways involved and whether they are unique to diabetes. Third, although we report the largest study evaluating the relation of BMD to CP in AA and EA men and women to date, our inferences related to AA men and the stratified analyses on the effects of estrogen in women are limited by relatively small sample sizes but document the importance of including minorities and both sexes in clinical studies in this area of science.

In summary, our results indicate that a complex but significant relation is present between atherosclerotic calcification and bone mineralization in type 2 diabetics. The observation of this relation in humans supports the hypothesis that metabolic, cellular and/or genetic processes related to both atherosclerosis and bone metabolism play a role in osteoporosis and atherosclerosis.

Furthermore, we have shown the relation between vascular calcification and bone density is strongly influenced by age, sex and ethnicity, with further effect modification in women receiving supplemental estrogen. EA men and AA women with diabetes have a significant inverse associations between vertebral trabecular bone and CP at multiple vascular sites, effects that remained significant after adjustment for age and other covariates. Thus, inverse relationships between bone and vascular calcification exist in a population of overweight and obese diabetic subjects at high risk for atherosclerotic vascular diseases. Subjects with obesity and diabetes are often advised to lose weight, and weight loss is usually paralleled by loss of bone. The existence of an independent inverse relationship between the skeletal and vascular calcium fluxes could lead to increased vascular calcification. It would be interesting, but technically challenging, to assess whether the calcium released from bone is preferentially relocated to the arterial wall over other calcium pools. Such an effect might occur if released skeletal calcium circulates in a mineral form or complexed with phospholipids or peptides which might be more readily incorporated into the vascular mineralization process. Whether these events occur or lead to increases or decreases in clinical events is a largely unexplored area worthy of investigation. These results support the performance of additional studies designed to better understand interactions between bone mineralization and atherosclerotic calcified plaque, as well as the influences of genes and hormone replacement therapy on this interaction.

Acknowledgements

Software and technical support was provided by Image Analysis; Columbia, KY and General Electric Healthcare; Waukesha, WI. The investigators acknowledge the cooperation of our participants; study recruiters Ms. Carrie Smith and Ms. Sue Ann Backus; CT analysts Lining Du, Caresse Hightower and Susan Pillsbury; technical support Chris O’Rourke, Bob Ellison, Josh Tan and Cynthia Hogan; and CT technologists. This study was supported by NIAMS and NHLBI grants R01 AR048797 (JJC) and R01 HL67348 (DWB) and the General Clinical Research Center of the Wake Forest University School of Medicine grant M01 RR07122. The authors report no conflicts of interest.

Abbreviations

aBMD

Areal BMD

CP

Atherosclerotic calcified plaque

BMD

Bone mineral density

CVD

Cardiovascular disease

CT

Computed tomography

DXA

Dual X-ray absorptiometry

GEE1

Generalized estimating equations

L-vBMD

Lumbar vertebrae vBMD

QCT

Quantitative CT

T-vBMD

Thoracic vertebrae vBMD

vBMD

Trabecular volumetric BMD

DM2

Type 2 diabetes mellitus

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Anderson JB, Barnett E, Nordin BE. The relationship between osteoporosis and aortic calcification. Br J Radiol. 1964;37:910–912. doi: 10.1259/0007-1285-37-444-910. [DOI] [PubMed] [Google Scholar]; 1b Marcovitz PA, Tran HH, Franklin BA, O’Neill WW, Yerkey M, Boura J, Kleerekoper M, Dickinson CZ. Usefulness of bone mineral density to predict significant coronary artery disease. Am J Cardiol. 2005;96:1059–63. doi: 10.1016/j.amjcard.2005.06.034. [DOI] [PubMed] [Google Scholar]; 1c Ness J, Aronow WS. Comparison of prevalence of atherosclerotic vascular disease in postmenopausal women with osteoporosis or osteopenia versus without osteoporosis or osteopenia. Am J Cardiol. 2006 May 15;97(10):1427–8. doi: 10.1016/j.amjcard.2005.12.033. [DOI] [PubMed] [Google Scholar]
  • 2.Silverman SL, Delmas PD, Kulkarni PM, Stock JL, Wong M, Plouffe L. Comparison of fracture, cardiovascular event, and breast cancer rates at 3 years in postmenopausal women with osteoporosis. J Am Geriatr Soc. 2004;52:1543–1548. doi: 10.1111/j.1532-5415.2004.52420.x. [DOI] [PubMed] [Google Scholar]
  • 3.Tanko LB, Christiansen C, Cox DA, Geiger MJ, McNabb MA, Cummings SR. Relationship between osteoporosis and cardiovascular disease in postmenopausal women. J Bone Miner Res. 2005;20:1912–1920. doi: 10.1359/JBMR.050711. [DOI] [PubMed] [Google Scholar]
  • 4.Browner WS, Pressman AR, Nevitt MC, Cauley JA, Cummings RG. Association between low bone density and stroke in elderly women. The study of osteoporotic fractures. Stroke. 1993;24:940–946. doi: 10.1161/01.str.24.7.940. [DOI] [PubMed] [Google Scholar]
  • 5.Kiel DP, Kauppila LI, Cupples LA, Hannan MT, O’Donnell CJ, Wilson PW. Bone loss and the progression of abdominal aortic calcification over a 25 year period; the Framingham Heart Study. Calcif Tissue Int. 2001;68:271–276. doi: 10.1007/BF02390833. [DOI] [PubMed] [Google Scholar]
  • 6.Schulz E, Arfai K, Liu X, Sayre J, Gilsanz V. Aortic calcification and the risk of osteoporosis and fractures. J Clin Endocrinol Metab. 2004;89:4243–4245. doi: 10.1210/jc.2003-030964. [DOI] [PubMed] [Google Scholar]
  • 7.Bagger YZ, Tanko LB, Alexandersen P, Qin G, Christiansen C, Prospective Epidemiological Risk Factors Study Group Radiographic measure of aorta calcification is a site-specific predictor of bone loss and fracture risk at the hip. J Intern Med. 2006;259:598–605. doi: 10.1111/j.1365-2796.2006.01640.x. [DOI] [PubMed] [Google Scholar]
  • 8.Schwartz AV, Sellmeyer DE, Ensrud KE, Cauley JA, Tabor HK, Schreiner PJ, Jamal SA, Black DM, Cummings SR, Study of Osteoporotic Features Research Group Older women with diabetes have an increased risk of fracture: a prospective study. J Clin Endocrinol Metab. 2001;86:32–38. doi: 10.1210/jcem.86.1.7139. [DOI] [PubMed] [Google Scholar]
  • 9.Strotmeyer ES, Cauley JA, Schwartz AV, Nevitt MC, Resnick HE, Bauer DC, Tylavsky FA, deRekeneire N, Harris TB, Newman AB. Nontraumatic fracture risk with diabetes mellitus and impaired fasting glucose in older white and black adults: the health, aging, and body composition study. Arch Intern Med. 2005;165:1612–1617. doi: 10.1001/archinte.165.14.1612. [DOI] [PubMed] [Google Scholar]
  • 10.Vestergaard P, Rejnmark L, Mosekilde L. Relative fracture risk in patients with diabetes mellitus, and the impact of insulin and oral antidiabetic medication on relative fracture risk. Diabetologia. 2005;48:1292–1299. doi: 10.1007/s00125-005-1786-3. [DOI] [PubMed] [Google Scholar]
  • 11.Schwartz AV, Sellmeyer DE, Strotmeyer ES, Tylavsky FA, Feingold KR, Resnick HE, Shorr RI, Nevitt MC, Black DM, Cauley JA, Cummings SR, Harris TB, Health ABC Study Diabetes and bone loss at the hip in black and white adults. J Bone Miner Res. 2005;20:596–603. doi: 10.1359/JBMR.041219. [DOI] [PubMed] [Google Scholar]
  • 12.Lenchik L, Hsu FC, Register TC, Lohman KK, Freedman BI, Langefeld CD, Bowden DW, Carr JJ. Heritability of spinal trabecular volumetric bone mineral density measured by QCT in the Diabetes Heart Study. Calcif Tissue Int. 2004;75:305–312. doi: 10.1007/s00223-004-0249-z. [DOI] [PubMed] [Google Scholar]
  • 13.Carr JJ, Nelson JC, Wong ND, McNitt-Gray M, Arad Y, Jacobs DR, Jr, Sidney S, Bild DE, Williams OD, Detrano RC. Calcified coronary artery plaque measurement with cardiac CT in population-based studies: standardized protocol of Multi-Ethnic Study of Atherosclerosis (MESA) and Coronary Artery Risk Development in Young Adults (CARDIA) study. Radiology. 2005;234:35–43. doi: 10.1148/radiol.2341040439. [DOI] [PubMed] [Google Scholar]
  • 14.Carr JJ, Crouse JR, Goff DC, D’Agostino RA, Peterson NP, Burke GL. Evaluation of subsecond gated helical CT for quantification of coronary artery calcium and comparison with electron beam CT. AJA Am J Roentgenol. 2000;174:915–921. doi: 10.2214/ajr.174.4.1740915. [DOI] [PubMed] [Google Scholar]
  • 15.Detrano RC, Anderson M, Nelson J, Wong ND, Carr JJ, McNitt-Gray M, Bild DE. Coronary calcium measurements: effect of CT scanner type and calcium measure on rescan reproducibility--MESA study. Radiology. 2005;236:477–484. doi: 10.1148/radiol.2362040513. [DOI] [PubMed] [Google Scholar]
  • 16.Lenchik L, Register TC, Hsu FC, Lohman K, Nicklas BJ, Freedman BI, Langefeld CD, Carr JJ, Bowden DW. Adiponectin as a novel determinant of bone mineral density and visceral fat. Bone. 2003;33:646–651. doi: 10.1016/s8756-3282(03)00237-0. [DOI] [PubMed] [Google Scholar]
  • 17.Lenchik L, Shi R, Register TC, Beck SR, Langefeld CD, Carr JJ. Measurement of trabecular bone mineral density in the thoracic spine using cardiac gated quantitative computed tomography. J Comput Assist Tomogr. 2004;28:134–139. doi: 10.1097/00004728-200401000-00023. [DOI] [PubMed] [Google Scholar]
  • 18.Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a generalized estimating equation approach. Biometrics. 1988;44:1049–1060. [PubMed] [Google Scholar]
  • 19.Ouyang P, Michos ED, Karas RH. Hormone replacement therapy and the cardiovascular system lessons learned and unanswered questions. J Am Coll Cardiol. 2006;47:1741–53. doi: 10.1016/j.jacc.2005.10.076. [DOI] [PubMed] [Google Scholar]
  • 20.Lindsay R. Hormones and bone health in postmenopausal women. Endocrine. 2004;24:223–30. doi: 10.1385/ENDO:24:3:223. [DOI] [PubMed] [Google Scholar]; 20a Manson JE, Allison MA, Rossouw JE, Carr JJ, Langer RD, Hsia J, Kuller LH, Cochrane BB, Hunt JR, Ludlam SE, Pettinger MB, Gass M, Margolis KL, Nathan L, Ockene JK, Prentice RL, Robbins J, Stefanick ML, WHI and WHI-CACS Investigators Estrogen therapy and coronary artery calcification. N Engl J Med. 2007;25:2591–2602. doi: 10.1056/NEJMoa071513. [DOI] [PubMed] [Google Scholar]
  • 21.Allison MA, Criqui MH, Wright CM. Patterns and risk factors for systemic calcified atherosclerosis. Arterioscler Thromb Vasc Biol. 2004;24:331–336. doi: 10.1161/01.ATV.0000110786.02097.0c. [DOI] [PubMed] [Google Scholar]
  • 22.Maruyama T, Miyazawa I, Oguchi T, Miyashita T, Katagiri Y, Sasaki Y, Kiyosawa K. Association between the extent of sclerotic changes in iliac arteries and long-term prognosis in patients with ischemic heart disease. J Cardiol. 1996;28:33–39. [PubMed] [Google Scholar]
  • 23.Bild DE, Detrano R, Peterson D, Guerci A, Liu K, Shahar E, Outyang P, Jackson S, Saad MF. Ethnic differences in coronary calcification: The multi-ethnic study of atherosclerosis (MESA) Circulation. 2005;111:1313–1320. doi: 10.1161/01.CIR.0000157730.94423.4B. [DOI] [PubMed] [Google Scholar]
  • 24.Carnethon MR, Bertoni AG, Shea S, Greenland P, Ni H, Jacobs DR, Saad M, Liu K. Racial/Ethnic differences in subclinical atherosclerosis among adults with diabetes: the multiethnic study of atherosclerosis. Diabetes Care. 2005;28:2768–2770. doi: 10.2337/diacare.28.11.2768. [DOI] [PubMed] [Google Scholar]
  • 25.Budoff MJ, Nasir K, Mao S, Tseng PH, Chau A, Liu ST, Flores F, Blumenthal RS. Ethnic differences of the presence and severity of coronary atherosclerosis. Atherosclerosis. 2006;187:343–50. doi: 10.1016/j.atherosclerosis.2005.09.013. Atherosclerosis. [DOI] [PubMed] [Google Scholar]
  • 26.Freedman BI, Hsu FC, Langefeld CD, Rich SS, Herrington DM, Carr JJ, Xu J, Bowden DW, Wagenknecht LE. The impact of ethnicity and sex on subclinical cardiovascular disease: the Diabetes Heart Study. Diabetologia. 2005;48:2511–2518. doi: 10.1007/s00125-005-0017-2. [DOI] [PubMed] [Google Scholar]
  • 27.Luckey MM, Wallenstein S, Lapinski R, Meier DE. A prospective study of bone loss in African-American and white women--a clinical research center study. J Clin Endocrinol Metab. 1996;81:2948–2956. doi: 10.1210/jcem.81.8.8768857. [DOI] [PubMed] [Google Scholar]
  • 28.Acheson LS. Bone density and the risk of fractures: should treatment thresholds vary by ethnicity? JAMA. 2005;293:2151–2154. doi: 10.1001/jama.293.17.2151. [DOI] [PubMed] [Google Scholar]
  • 29.Young BA, Rudser K, Kestenbaum B, Seliger SL, Andress D, Boyko EJ. Racial and ethnic differences in incident myocardial infarction in end-stage renal disease patients: The USRDS. Kidney Int. 2006;69:1691–1698. doi: 10.1038/sj.ki.5000346. [DOI] [PubMed] [Google Scholar]
  • 30.Young BA, Maynard C, Boyko EJ. Racial differences in diabetic nephropathy, cardiovascular disease, and mortality in a national population of veterans. Diabetes Care. 2003;26:2392–2399. doi: 10.2337/diacare.26.8.2392. [DOI] [PubMed] [Google Scholar]
  • 31.Karter AJ, Ferrara A, Liu JY, Moffet HH, Ackerson LM, Selby JV. Ethnic disparities in diabetic complications in an insured population. JAMA. 2002;287:2519–2527. doi: 10.1001/jama.287.19.2519. [DOI] [PubMed] [Google Scholar]
  • 32.Wagenknecht LE, Bowden DW, Carr JJ, Langefeld CD, Freedman BI, Rich SS. Familial aggregation of coronary artery calcium in families with type 2 diabetes. Diabetes. 2001;50:861–866. doi: 10.2337/diabetes.50.4.861. [DOI] [PubMed] [Google Scholar]
  • 33.Hruska KA, Mathew S, Saab G. Bone morphogenetic proteins in vascular calcification. Circ Res. 2005;97:105–114. doi: 10.1161/01.RES.00000175571.53833.6c. [DOI] [PubMed] [Google Scholar]
  • 34.Taylor BC, Schreiner PJ, Doherty TM, Fornage M, Carr JJ, Sidney S. Matrix Gla protein and osteopontin genetic associations with coronary artery calcification and bone density: the CARDIA study. Hum Genet. 2005;116:525–528. doi: 10.1007/s00439-005-1258-3. [DOI] [PubMed] [Google Scholar]
  • 35.Mundy G, Garrett R, Harris S, Chan J, Chen D, Rossini G, Boyce B, Zhao M, Gutierrez G. Stimulation of bone formation in vitro and in rodents by statins. Science. 1999;286:1825–1826. doi: 10.1126/science.286.5446.1946. [DOI] [PubMed] [Google Scholar]
  • 36.Hatzigeorgiou C, Jackson JL. Hydroxymethylglutaryl-coenzyme A reductase inhibitors and osteoporosis: a meta-analysis. Osteoporos Int. 2005;16:990–8. doi: 10.1007/s00198-004-1793-0. [DOI] [PubMed] [Google Scholar]

RESOURCES