Journal of Bone and Mineral Metabolism

, Volume 27, Issue 1, pp 89–94

Association between lumber bone mineral density and vascular stiffness as assessed by pulse wave velocity in postmenopausal women

Authors

  • Miho Mikumo
    • Department of Obstetrics and GynecologyTokyo Women’s Medical University
  • Hiroya Okano
    • Department of Obstetrics and GynecologyTokyo Women’s Medical University
  • Remi Yoshikata
    • Department of Obstetrics and GynecologyTokyo Women’s Medical University
  • Ken Ishitani
    • Department of Obstetrics and GynecologyTokyo Women’s Medical University
    • Department of Obstetrics and GynecologyTokyo Women’s Medical University
Original Article

DOI: 10.1007/s00774-008-0014-x

Cite this article as:
Mikumo, M., Okano, H., Yoshikata, R. et al. J Bone Miner Metab (2009) 27: 89. doi:10.1007/s00774-008-0014-x

Abstract

Recent studies have showed a significant correlation between vascular calcification and bone mineral density (BMD). Therefore, an investigation was carried out on the association between arterial stiffness, lumbar BMD and bone metabolic markers in Japanese postmenopausal women. Brachial-ankle PWV (baPWV) and BMD of the lumbar spine and serum bone-specific alkaline phosphatase (BAP) levels in 143 postmenopausal women were measured, where there was a significant negative correlation between baPWV and BMD (r = −0.21; P = 0.0135). An additional analysis included the remaining 75 subjects, but excluded subjects with hypertension and obesity. Here, a more negative correlation between baPWV and BMD (r = −0.315; P = 0.006), and a positive correlation between baPWV and BAP (r = 0.248; P = 0.032) were also significant. A group analysis, where the women were age matched and stratified into three groups of different bone density, i.e., normal BMD, osteopenic and osteoporotic, were further made. This showed lower PWV values in the normal BMD group than in the other two groups. A study also showed that the tertile with the highest BAP was associated with significantly higher PWV values than the other tertiles. However, when the multiple linear regression analysis was carried out, there was no correlation between PWV and BAP values. Low BMD and arterial stiffness show some correlation, suggesting that BAP may reflect the degree of arterial stiffness present.

Keywords

Bone mineral densityBrachial-ankle PWVBone-specific alkaline phosphataseArteriosclerosisOsteoporosis

Introduction

Arteriosclerosis progresses with age, and the risk of arteriosclerosis in women increases significantly after menopause [1, 2]. On the other hand, bone mass decreases with age regardless of sex. However, women are at higher risk especially after menopause, when bone mass decreases rapidly due to a decrease in estrogen [3, 4]. It was recently shown that the degree of vascular calcification is significantly correlated with changes in bone density, suggesting that vascular sclerosis and decreased bone mass are closely linked pathological conditions [58].

Pulse wave velocity (PWV) can be used to measure the elasticity of arteries, thus providing an easy measure of progression of arteriosclerosis. Indeed, despite the fact that PWV values do not directly describe calcification of blood vessels, PWV is used as an effective measure of arteriosclerosis. Carotid-femoral PWV (cfPWV) has been a traditional method used to measure PWV. However, this method requires some technical skills. On the other hand, brachial-ankle PWV (baPWV), now available as a more convenient method, only requires placing blood pressure cuffs on the extremities. It has been reported that baPWV and cfPWV values are extremely well correlated within the same patient, suggesting that baPWV is as equally reliable as cfPWV as an index for the severity of arteriosclerosis as well as a prognostic indicator in the care of patients with hypertension [9, 10].

The most important objective of osteoporosis treatment lies in the prevention of bone fractures, which occur as bone strength, a composite endpoint combining both bone quality and density, diminishes considerably. In clinical studies, an accurate evaluation of bone density and quality is carried out by using the dual-energy X-ray absorptiometry (DXA) method and by the use of appropriate bone metabolism makers.

A negative correlation between vascular sclerosis and BMD has already been reported [58]. However, DXA has seldom been used for evaluation of bone density, and studies exploring the association between vascular sclerosis and bone metabolism markers are very few.

This study aimed to elucidate the relationship between arteriosclerosis and osteoporosis in postmenopausal women by examining their baPWV, bone mineral density by using DXA, and bone metabolic markers for association among these variables.

Subjects and methods

Subjects

This study enrolled 143 postmenopausal women who visited the Menopause Clinic in our department from January 2004 to April 2005. Their menopausal status was confirmed by interview, where those who had not menstruated for 12 months were considered to be menopausal. Those who could not precisely recall the date of their last menstrual period or who did not answer the questionnaire were excluded from the study. None of the subjects had undergone hormonal replacement therapy or had taken any steroid hormones. Women who had hypertension, diabetes, dyslipidemia or osteoporosis took appropriate medication for the condition (Table 1).
Table 1

Clinical characteristics of 143 women in this study

Age (years)

57.9 ± 8.3

Height (cm)

155.7 ± 5.2

Weight (kg)

55.3 ± 9.4

BMI (kg/m2)

23.0 ± 3.6

sBP (mmHg)

128.3 ± 21.6

dBP (mmHg)

76.1 ± 12.2

iPTH (pg/ml)

48.3 ± 19.9

Ca (mg/dl)

9.2 ± 0.8

Urinary Ca/Cr

0.18 ± 0.39

P (mg/dl)

3.6 ± 0.4

NTx (nmolBCE/mmol Cr)

43.1 ± 20.0

BMD (g/m2)

0.884 ± 0.154

PWV (cm/s)

1,450 ± 261

BAP (IU/l)

25.0 ± 11.2

No. of subjects with

 

 Hypertension

3

 Diabetes mellitus

5

 Dyslipidemia

44

 Osteoporosis

25

Age at menopause (years)

48.0 ± 5.0

All results are presented as the mean ± SD

BMI body mass index, sBP systolic blood pressure, dBP diastolic blood pressure, iPTH intact parathyroid hormone, NTx urine cross-linked N-telopeptides of type 1 collagen, BAP bone alkaline phosphatase

Informed consent was obtained from every subject who enrolled in the study. The study protocol was then approved by the Ethics Committee of Tokyo Women’s Medical University.

Measurement of pulse wave velocity

All subjects underwent brachial-ankle pulse wave velocity (baPWV) measurement as an index for arteriosclerosis, using a form pulse wave velocity/ankle brachial pressure index (form PWV/ABI) (Nippon Coli). Details of the methodology have been described previously [8].

The subjects were examined while resting in a supine position with the measurement device set to simultaneously record PWV, blood pressure, electrocardiogram and heart sounds. Electrocardiogram electrodes were placed on both wrists, and a heart sound microphone was placed over the left sternal border. The cuffs were wrapped around both arms and ankles and connected to the plethysmographic sensors to evaluate brachial and post-tibial arterial pressure waveforms and volume pulse forms in the subjects. The subjects were also connected to oscillometric sensors to measure blood pressure. baPWV was measured in all subjects after they had rested for at least 5 min. The mean left and the right baPWV values for each subject were used for analysis.

Measurement of bone mineral density

Lumbar spine (L2–L4) BMD was measured by using DXA (QDR4500, Hologic Inc., USA) as an index for osteoporosis. BMD values were reported as grams per square centimeter.

Laboratory measurements

The following variables were evaluated: blood Ca, P and intact PTH levels; urine Ca/Cr ratios; bone alkaline phosphatase (BAP) levels; and urine cross-linked telopeptides of type I collagen (NTx).

Statistical analysis

Data were expressed as mean ± SD. Simple regression analysis was used to examine correlation between baPWV, BMD and other clinical variables. Multiple regression analysis was further performed for baPWV, BMD and other clinical variables. A value of P < 0.05 was considered statistically significant.

Results

Table 1 summarizes the characteristics of the subjects. There was a significant negative correlation between PWV and BMD in the 143 patients studied (r = −0.21; P = 0.0135) (Table 2). When each of the parameters evaluated was examined for possible correlation with PWV and BMD, there was a positive correlation between PWV and blood pressure, with a stronger correlation found between PWV and systolic blood pressure (r = 0.734; P < 0.0001). A positive correlation was also present between PWV and bone ALP (BAP) (r = 0.166; P = 0.047) (Table 2). BMD showed a positive correlation with both body weight and BMI, where a stronger correlation was seen between BMD and body weight (r = 0.506; < 0.0001) (Table 2). Given these results, patients with hypertension (defined as sBP 140 mmHg or higher or dBP 90 mmHg or higher), a factor affecting PWV values, and those with BMI less than 18.5 as well as those with BMI more than 25, a factor affecting BMD, were all excluded from the study. The remaining 75 subjects were subjected to further review. Analysis of these 75 subjects showed a more significant negative correlation between PWV and BMD (r = −0.315; P = 0.006) than in the earlier analysis from which no subjects were excluded (Fig. 1).
Table 2

Univariate analysis of clinical factors correlated with brachial-ankle pulse wave velocity (baPWV) and lumbar bone mineral density (BMD)

 

With baPWV

With BMD

r

P

r

P

Age

0.587

<0.0001

0.283

0.0006

Height

0.311

0.0002

0.191

0.169

Weight

0.006

NS

0.506

<0.0001

BMI

0.128

NS

0.453

<0.0001

sBP

0.734

<0.0001

0.026

NS

dBP

0.564

<0.0001

0.074

NS

BMD/baPWV

0.206

0.0135

0.206

0.0135

Ca

0.072

NS

0.090

NS

P

0.148

0.0793

0.026

NS

Urinary Ca/Cr

0.044

NS

0.028

NS

iPTH

0.140

0.0947

0.017

NS

NTx

0.030

NS

0.051

NS

BAP

0.166

0.0470

0.018

NS

BMI body mass index, sBP systolic blood pressure, dBP diastolic blood pressure, iPTH intact parathyroid hormone, NTx urine cross-linked N-telopeptides of type 1 collagen, BAP bone alkaline phosphatase

https://static-content.springer.com/image/art%3A10.1007%2Fs00774-008-0014-x/MediaObjects/774_2008_14_Fig1_HTML.gif
Fig. 1

An even stronger correlation was observed between L2–L4 BMD and PWV values after those with hypertension, a determining factor of PWV, and those with obesity as assessed by BMI, a determining factor of L2–L4 BMD, were excluded (r = −0.315; P = 0.006)

In order to eliminate the possibility of age affecting the results, the subjects were age-matched and then stratified into three groups by bone density, i.e., normal BMD (n = 17, L2–L4 BMD, 0.962 ± 0.085 g/cm2), osteopenic (n = 12, 0.755 ± 0.029) and osteoporotic (n = 9, 0.673 ± 0.028). No significant difference was observed among these three groups concerning their age and blood pressure (Table 3). The subjects with normal BMD showed significantly lower PWV values than the other two groups (Fig. 2). Hence, a significant correlation between PWV and BMD was confirmed to be present even after adjustment for age among the subjects.
Table 3

Background factors in the three groups aged-matched and stratified by BMD

 

Osteoporotic

Osteopenic

Normal BMD

P

Number

9

12

17

 

BMD (g/m2)

0.067 ± 0.028

0.755 ± 0.029

0.962 ± 0.085

<0.05

Age (years)

57.8 ± 8.1

57.6 ± 8.1

57.3 ± 5.2

NS

Height (cm)

155.0 ± 4.3

153.3 ± 4.4

157.2 ± 6.3

NS

Weight (kg)

49.6 ± 4.0

49.0 ± 5.1

54.9 ± 6.8

<0.05

BMI (kg/m2)

20.6 ± 0.9

21.1 ± 1.8

22.1 ± 1.7

<0.05

sBP (mmHg)

120 ± 11

114 ± 13

11 ± 10

NS

dBP (mmHg)

71 ± 9

71 ± 8

66 ± 10

NS

iPTH (pg/ml)

47.2 ± 13.5

40.6 ± 11.9

46.8 ± 16.7

NS

Ca (mg/dl)

9.3 ± 0.3

9.3 ± 0.3

9.4 ± 0.6

NS

Urinary Ca/Cr

0.13 ± 0.08

0.16 ± 0.12

0.14 ± 0.07

NS

P (mg/dl)

3.5 ± 0.3

3.7 ± 0.2

3.6 ± 0.3

NS

NTx (nmolBCE/mmol Cr)

42.0 ± 22.5

48.2 ± 26.8

43.6 ± 21.1

NS

BAP (IU/l)

25.9 ± 8.1

24.6 ± 7.5

20.4 ± 6.5

NS

BMI body mass index, sBP systolic blood pressure, dBP diastolic blood pressure, iPTH intact parathyroid hormone, NTx urine cross-linked N-telopeptides of type 1 collagen, BAP bone alkaline phosphatase

https://static-content.springer.com/image/art%3A10.1007%2Fs00774-008-0014-x/MediaObjects/774_2008_14_Fig2_HTML.gif
Fig. 2

The subjects in the normal BMD group showed significantly lower PWV values than the other two groups (1,201.1 ± 18.3 vs. 1,312.6 ± 49.0 cm/s, 1,201.1 ± 18.3 vs. 1,399.5 ± 54.1 cm/s; P < 0.05). All results are presented as mean ± SD. In these subjects, a significant negative correlation was shown between PWV and BMD even after they were adjusted for age

In the remaining 75 patients, PWV values showed a stronger positive correlation with the bone metabolism marker BAP (r = 0.248; P = 0.032) than when all subjects were included for analysis (r = 0.166; P = 0.047) (Fig. 3). Comparison of PWV values among the BAP tertiles showed that the tertile with the highest BAP showed significantly higher PWV values than the other tertiles (P < 0.05) (Fig. 4). A further examination by multiple regression analysis showed no correlation between PWV and BMD or between PWV and BAP (Table 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-008-0014-x/MediaObjects/774_2008_14_Fig3_HTML.gif
Fig. 3

PWV values showed a stronger positive correlation with BAP, a bone formation marker, in 75 subjects with normal blood pressure and BMI (r = 0.248; P = 0.032)

https://static-content.springer.com/image/art%3A10.1007%2Fs00774-008-0014-x/MediaObjects/774_2008_14_Fig4_HTML.gif
Fig. 4

The tertile with the highest BAP showed significantly higher PWV values than the other tertiles (1,308.8 ± 35.0 vs. 1,285.5 ± 27.6 cm/s, 1,308.8 ± 35.0 vs. 1,247.9 ± 35.6 cm/s; P < 0.05)

Table 4

Correlation of PWV and other factors as assessed by multiple regression analysis with a significance level of P < 0.05

Variable

Regression coefficient

P value

Standardized regression coefficient

Age

11.42

<0.001

262.58

Height

−3.72

0.115

−52.11

sBP

6.98

<0.001

408.06

BMD

−99.78

0.213

−35.27

BAP

1.89

0.171

42.06

sBP systolic blood pressure, BAP bone alkaline phosphatase

Discussion

Our study results demonstrate that PWV and BMD are negatively correlated in postmenopausal women. In addition, the greater the PWV values, and the more sclerotic the blood vessels are, the lower the lumbar L2–L4 BMD values. Of note, this negative correlation was shown to be particularly pronounced among women with normal physique and blood pressure. Furthermore, this correlation was confirmed even when the data were adjusted for age, suggesting that decreased bone mass is a risk factor for atherosclerosis, independently of other risk factors, such as hypertension, diabetes or smoking. Our results are in agreement with the report of Hirose et al. [7] that demonstrated correlation between increased PWV and reduced calcaneal quantitative osteo-sono index as assessed by quantitative ultrasound (QUS). In recent years, similar findings have been reported not only in cross-sectional, but also in longitudinal studies [6, 11] that were conducted across races. These studies began to clarify the cellular mechanisms of pathogenesis implicated in both atherosclerosis and decreased bone mass [12, 13]. An osteoblast- or chondrocyte-like phenotypic transformation of vascular smooth muscle cells and myofibroblasts is assumed to be implicated in the process of vascular calcification, suggesting a role for osteochondral metabolism-associated factors in this process [1416]. While aging and menopause are clinical risk factors for both atherosclerosis and osteoporosis, other factors, such as various inflammatory processes, oxidative stress and homocystein, are also reported as risk factors for both conditions [17].

The interrelationship between atherosclerosis and bone metabolism has been corroborated by the fact that anti-atherosclerotic and anti-resorptive agents exert effects on bone metabolism and on atherosclerosis [18, 19]. In this regard, statins as therapeutic agents for hypercholesterolemia are known to exert their anti-atherosclerotic effects through inhibition of HMG-CoA reductase, a key enzyme in the rate-limiting step of the mevalonate pathway; they have also been shown to mediate BMP-2 promoter activation. In both mouse calvarial cultures and in clinical trials [20, 21], statins were shown to increase the number of osteoblasts as well as new bone mass, suggesting a potential role for statins as a new class of pro-osteogenic agents. On the other hand, bisphosphonates as anti-osteoporotic agents have been shown to suppress osteoclast activation as part of their mechanism of action that inhibits the mevalonate pathway [22]. Thus, together, these results suggest that statins and bisphosphonates may act on, and exert similar effects on, the same cells.

In our present study as well, BAP and PWV were found to be correlated, consistently with previous reports showing that when osteoporotic patients were stratified by presence or absence of aortic calcification, those with aortic calcification were associated with significantly higher BAP values [23]. It is also reported that BAP was significantly expressed in calcified vascular smooth muscle cells. Furthermore, in the presence of pro-inflammatory cytokines, there is an increase in the BAP level, thus further promoting vascular calcification [24]. These findings appear to point towards the possibility that BAP values reflect the degree of arteriosclerosis present and that osteoblast-like cells are implicated in arteriosclerosis.

In our analyses using multiple linear regression, we were unable to establish a clear relationship between PWV and BMD or between PWV and BAP. However, the results obtained from the stratified groups do not necessarily exclude the possibility of such relationship, as shown in a comparison of PWV values among the BAP tertiles. The main limitation of this study lies in the fact that the study subjects were not adequately uniform; the study subjects varied greatly in age and included those with medical conditions such as hypertension and diabetes. The limited availability of appropriate cases further enhanced the study limitation. Therefore, increasing the number of subjects, which also helps to ensure inclusion of uniform subjects, may contribute towards a better clarification of the relationship between bone and vasculature.

The management of bone metabolic disorders needs to focus not only on the disease per se, but also on the resulting vascular calcification that will likely lead to ectopic calcification, thus affecting the overall prognosis of affected patients. Mounting evidence suggests a strong correlation between vascular calcification and bone mineral content. While the implication of this finding remains to be further explored, current evidence appears to suggest a role for BMD measurement as an important index that assists in the management of vascular calcification.

It is suggested that patients with low bone mass should undergo careful monitoring for atherosclerosis to better manage the condition, where therapeutic intervention may also be considered. In a fast-aging society, such an approach is not only needed for the health of people and for the social care workers caring for them, but also for health economic reasons.

Copyright information

© The Japanese Society for Bone and Mineral Research and Springer 2008