CLIMACTERIC 2014;17:486–491
Association of breast vascular calcifications with low bone mass in postmenopausal women E. J. Nasser∗, E. R. Iglésias†, J. A. S. Ferreira∗, C. E. Fernandes∗ and L. M. Pompei∗ ∗ABC School of Medicine, Gynecology Branch, Santo André; †Diagnosis Clinic ‘Diagnóstico por Imagem de Santo André – DISA’, Santo André, Brazil Key words: OSTEOPOROSIS, BREAST VASCULAR CALCIFICATIONS, SCREENING, LOW BONE MASS, OSTEOPENIA
ABSTRACT Background In developing countries, there is a deficiency of densitometers with which to screen the population for osteoporosis. Thus, strategies with which to select patients for a bone density test are desirable. Objective To determine whether breast vascular calcifications (BVCs) may be employed to identify postmenopausal women with osteoporosis/osteopenia. Methods This was a cross-sectional study of postmenopausal women subjected to bilateral mammography and bone densitometry (DXA) of the spine and hip. A medical interview registered possible confounding factors, such as age, length of menopause, previous use of postmenopausal hormone therapy, family history of osteoporosis, smoking, alcoholism, hypertension, diabetes, cardiovascular diseases, and medication use. Results The study included 211 postmenopausal women aged 62.1 ⫹ ⫺ 9.3 years, 38 of whom (18.0%) exhibited BVC. Osteoporosis was detected in 36 (17.1%), and a T-score ⬍ ⫺1.0 for any site was found in 164 (77.7%). No statistically significant difference was found between the groups without BVC (n ⫽ 173) and with BVC (n ⫽ 38) for the prevalence of ‘osteoporosis’ or ‘moderate/severe osteopenia or osteoporosis’ at the spine or at any other site. There was a difference between the groups in terms of age (59.0 ⫹ ⫺ 7.8 vs. 71.9 ⫹ 8.9 years, respectively; p ⬍ 0.001), sedentary lifestyle (57.8% vs. 84.2%, respectively; p ⫽ 0.002), ⫺ smoking (27.7% vs. 7.9%, respectively; p ⫽ 0.009), and high blood pressure (65.3% vs. 92.1%, respectively; p ⫽ 0.001). Logistic regression analysis confirmed the lack of statistical significance for BVC as a predictor of an osteoporosis diagnosis. Sensitivity values of BVCs to detect osteoporosis or osteopenia ranged from 17.9% to 25.0%. Conclusion BVCs have been shown to be inadequate to identify postmenopausal women with osteoporosis or osteopenia.
INTRODUCTION Osteoporosis, one of the main public health problems in the world1, is a disease characterized by reduction in bone density and quality. Its course is silent and bone becomes progressively porous and weak, more so with age and menopause, thus leading to an increasingly greater risk of fractures. The world population is estimated to be over 7 billion people, 846 million of whom are more than 60 years old. With the increase in life expectancy, it is expected that in 2050 there will be above 1.1 billion people over 70 years world-wide2,
consequently increasing the number of people with chronic diseases like osteoporosis. The chief means of diagnosing and monitoring osteoporosis is bone densitometry, preferably using the dual-energy X-ray absorptiometry (DXA) method. In Latin America, access to DXA is quite limited. Brazil and Chile are the countries with the most accessibility (ten machines per million inhabitants) according to the Latin American Regional Audit 2012 report. In Brazil, there are approximately 1850 densitometers with limited access in the public health system, and the cost of each examination is in the neighborhood of US$ 403. According to
Correspondence: Dr L. M. Pompei, ABC School of Medicine, Gynecology Branch, Rua Dr. Procopio Ribeiro dos Santos 84, Sao Paulo, 04664-130 Brazil; E-mail: [email protected]
ORIGINAL ARTICLE © 2014 International Menopause Society DOI: 10.3109/13697137.2013.869672
Received 27-09-2013 Revised 18-11-2013 Accepted 23-11-2013
Breast vascular calcifications and bone mass the World Health Organization (WHO), with the increase in the population over 65 years in Latin America in the next 50 years, the number of hip fractures will be equivalent to that in the United States or in the European countries today4. It is known, through epidemiological studies, that osteoporosis in postmenopausal women is associated with cardiovascular disease (CVD)5 and that osteoporosis in the lumbar spine and proximal femur is associated with coronary and aortic vascular calcifications6,7. Low bone mineral density (BMD) has been further associated with increased risk of myocardial infarction regardless of other risk factors8. There is also a study which points to the high probability of the connection between breast vascular calcifications (BVC) and CVD9. Although epidemiological studies show the positive association between coronary and aortic vascular calcifications and osteoporosis of the spine and hip, thus far there is only one retrospective study with 228 North American women revealing the connection between BMD and BVC10. The fact that BVC can be easily observed in routine mammography, carried out to screen for breast cancer in postmenopausal women, sparked the interest in conducting a study with the purpose of relating this radiological finding to BMD in a population of Brazilian postmenopausal women. Besides, there is a much greater availability of mammography machines in the public health system in our country than densitometers, and thus the examination would be of great practical value, for it would help to detect perceptible changes signaling the need for performing a densitometry, helping to optimize the use of densitometers.
METHODS This is a cross-sectional study to evaluate the association between BVC and low bone mass. All study procedures were approved by the Internal Review Board of the ABC School of Medicine, and all study participants gave their written informed consent. The women participating in the study were being treated at the gynecology outpatient clinic. To be included in the study, the women had to be in postmenopause, aged 45 years or older, to have undergone mammography and bone densitometry of the lumbar spine and hip in the previous 12 months with no more than a 6-month interval between the two examinations. Diagnosis of postmenopause was established for a woman over 45 years old with intact uterus if she presented amenorrhea for at least 12 months. For a hysterectomized woman, we considered her to be postmenopausal when two serum follicle stimulating hormone and estradiol levels with a minimum interval of 60 days between them were ⬎ 30 UI/l and ⬍ 35 pg/ml, respectively. Those patients using any medication that might interfere with bone metabolism, such as bisphosphonates, estrogen therapy, systemic steroids, anticoagulants, anticonvulsants and aromatase inhibitors, and who were carriers of uncompensated
Climacteric
Nasser et al. thyroid dysfunction, hyperparathyroidism, osteogenesis imperfecta, Paget’s bone disease, rheumatoid arthritis, malignant neoplasia (except for non-melanoma skin cancer), and of mammographic changes categorized as BI-RADS® 4 and 5 were excluded from the study. Following inclusion, all women were subjected to a standard questionnaire aimed at gathering information about possible confounding factors, such as age, ethnicity, profession, schooling, age at last menstruation, length of menopause, previous use of postmenopausal hormone therapy, a sedentary lifestyle, a family history of osteoporosis, smoking habits, alcoholism, hypertension, dyslipidemia, diabetes, cardiovascular diseases, as well as medication use. Weight and height, as well as any previous osteoporotic fractures, were recorded at the time of bone densitometry. Participants were required to have had bone densitometry measured by the DXA technique (Lunar, General Electric Healthcare, Chalfont St. Giles, UK) within the previous 12 months. Results were classified as osteoporosis or osteopenia in accordance with the WHO criteria: osteopenia when the T-score was less than -1.0 and more than ⫺2.5 (mild: ⬍ ⫺1.0 and ⫺ ⬎ ⫺1.5; moderate: ⬍ ⫺1.5 and ⫺ ⬎ ⫺2.0; severe: ⬍ ⫺2.0 11 and ⬎ ⫺2.5) and osteoporosis when the T-score ⬍ ⫺ ⫺2.5 . The mammograms were reviewed by a single radiology technologist (E.R.I.) blinded to the participants’ clinical data and bone mass. Each examination was classified as positive or negative for BVCs based on the easily recognizable characteristic pattern of amorphous calcifications between two parallel lines12. Sample size was initially calculated considering the prevalence of BVC to be approximately 10%9,13,14, and of low bone mass, around 40%, for the type of population included in the study10,15. Based on this assumption, 29 BVC cases and a total of 290 participants were deemed necessary for an 80% power to detect an odds ratio of 3.0 for BVCs to be a predictor of low bone mass, considering a type I error of 5%. As the exact prevalence of BVC and low bone mass in the study population were not known, a definite sample size was calculated after gathering data from the first 150 participants (interim analysis). Since the prevalence of BVC was approximately 18% and the prevalence of a T-score ⬍ ⫺1.5 was 44% at any site, it was estimated that 33 BVC cases and a total of 184 participants would be necessary as long as the other parameters remained the same. An additional 15% of cases were included as a safety measure for possible loss of data or variations in known prevalences. The continuous numerical data were presented as mean ⫹ ⫺ standard deviation. Categorical data were presented as absolute values and percentage rates. The comparisons of continuous numerical data between the groups were made with the Student t-test for independent samples when the distribution was normal and with the Mann–Whitney test when the distribution was not normal. The categorical data were compared using the χ2 test, except when the expected values in contingency tables were less than 5 in more than 20% of the cells. In this case, the Fisher exact test was employed. All data were included in an electronic spreadsheet
487
Breast vascular calcifications and bone mass
Nasser et al.
(Microsoft® Excel® 2007, Microsoft, Redmond, USA) and statistical analysis was performed with the WinSTAT® software, version 2007.1 (R. Fitch Software, Bad Krozingen, Germany). Multivariate logistic regressions were carried out in which the independent variables were age, body mass index (BMI), sedentary lifestyle, present or past smoking habits, diagnosis of high blood pressure, diabetes, dyslipidemia, or any cardiovascular disease, and presence of BVC. For the dependent variable, three models were analyzed, one with spine osteoporosis, another with hip osteoporosis, and the third one with osteoporosis at any site. The logistic regressions were performed with the OpenStat® v. 2013 software (Miller WG). The sensitivities and specificities of BVCs were also calculated for the detection of moderate or severe osteopenia and detection of osteoporosis.
RESULTS Of the 211 patients included in the study, 166 were of white ethnicity (78.7%), 40 of black ethnicity (18.9%), and five of yellow ethnicity (2.4%). BVCs were identified in 38 participants (18.0%). The prevalence of low BMD (osteoporosis or osteopenia) at the lumbar spine was 63.0% (133 cases), and at the hip, 58.8% (124 cases). The prevalence of osteoporosis at the spine was 13.3% (28 cases), and at the hip, 7.6% (16 cases). Other data from the total sample are displayed in Table 1.
Table 1 Characteristics of the study population. Data are given as mean ⫹ ⫺ standard deviation or frequency (%) 62.1 ⫹ ⫺ 9.3 48.9 ⫹ ⫺ 4.4 13.4 ⫹ ⫺ 9.4 68.4 ⫹ ⫺ 13.5 153.2 ⫹ ⫺ 5.5 29.1 ⫹ ⫺ 5.3
Age at densitometry (years) Age at menopause (years) Length of postmenopause (years) Weight (kg) Height (cm) Body mass index (kg/m2) Bone mass density Lumbar spine (g/cm2) T-score for lumbar spine Neck (g/cm2) T-score for neck Total femur (g/cm2) T-score for total femur Osteoporosis or osteopenia Osteoporosis at spine T-score ⬍ ⫺1.0 at spine Osteoporosis at hip T-score ⬍ ⫺1.0 at hip Osteoporosis at any site T-score ⬍ ⫺1.0 at any site
1.058 ⫹ ⫺ 0.164 ⫺1.15 ⫹ ⫺ 1.35 0.864 ⫹ ⫺ 0.125 ⫺1.01 ⫹ ⫺ 1.03 0.910 ⫹ ⫺ 0.131 ⫺0.75 ⫹ ⫺ 1.10 28 133 16 124 36 164
(13.3%) (63.0%) (7.6%) (58.8%) (17.1%) (77.7%)
Table 2 shows the comparison between the group with BVCs and that without BVCs in terms of some clinical parameters. Statistically significant differences between the two groups were found for the following variables: age, length of postmenopause, sedentary lifestyle, smoking habits, hypertension, diabetes, dyslipidemia, or any cardiovascular disease.
Table 2 Comparison between the groups with and without breast vascular calcifications (BVC) in terms of clinical parameters. Data are given as mean ⫹ standard deviation or frequency (%)
⫺
Parameter
Without BVC (n ⫽ 173)
With BVC (n ⫽ 38)
Age (at DXA, years) Age at menopause (years) Length of postmenopause (years) Weight (kg) Height (cm) Body mass index (kg/m2) Sedentary lifestyle Postmenopausal hormone therapy in past Family history of osteoporosis Smoker (present or past) Smoker (only present) Hypertension Diabetes Hyperlipidemia BAH or DM or dyslipidemia or any CVD Previous fracture due to bone weakness
59.9 ⫹ ⫺ 7.8 48.8 ⫹ ⫺ 4.6 11.4 ⫹ ⫺ 8.1 68.4 ⫹ ⫺ 13.4 153.8 ⫹ ⫺ 5.5 28.9 ⫹ ⫺ 5.2 100 (57.8%) 19 (11.0%) 30 (17.3%) 48 (27.7%) 22 (12.7%) 81 (46.8%) 27 (15.6%) 63 (36.4%) 113 (65.3%) 0
71.9 ⫹ ⫺ 8.9 49.6 ⫹ ⫺ 3.0 22.3 ⫹ ⫺ 9.9 68.1 ⫹ ⫺ 13.8 150.4 ⫹ ⫺ 4.7 30.2 ⫹ ⫺ 5.7 32 (84.2%) 2 (5.3%) 2 (5.3%) 3 (7.9%) 1 (2.6%) 28 (73.7%) 10 (26.3%) 17 (44.7%) 35 (92.1%) 0
p Value ⬍ 0.001* 0.405† ⬍ 0.001† 0.939† 0.003* 0.242* 0.002‡ 0.229** 0.060‡ 0.009‡ 0.053** 0.003‡ 0.116‡ 0.338‡ 0.001‡ NA
DXA, dual-energy X-ray absorptiometry; BAH, blood arterial hypertension; DM, diabetes mellitus; CVD, cardiovascular disease; NA, not applicable *, Student t-test for independent samples; †, Mann–Whitney U-test; ‡, χ2 test; **, Fisher exact test
488
Climacteric
Breast vascular calcifications and bone mass Table 3 shows the comparison between the groups with respect to DXA results and diagnosis of osteoporosis and osteopenia. Only the following parameter had statistical significance: osteoporosis or moderate or severe osteopenia at the hip (T-score ⬍ ⫺1.5). Logistic regression analysis revealed that in all models (spine osteoporosis, hip osteoporosis, or osteoporosis at any site as a dependent variable) only age and BMI exhibited a statistically significant effect. In none of the models did BVC have the statistical significance for predicting the osteoporosis diagnoses. Table 4 shows the details of the logistic regression. Table 5 displays the sensitivity and specificity values of BVCs to detect osteoporosis and low bone mass.
DISCUSSION This is the first study which has attempted to evaluate the association between BVC and low BMD in a Brazilian population. Our results did not show any association, except in the case of osteoporosis or moderate/severe osteopenia at the hip (T-score ⬍ ⫺1.5). The regression logistic analyses confirmed the lack of association. The low sensitivity of BVC for detection of osteoporosis or low bone mass became evident.
Nasser et al. Several authors, including those in our group, have shown that BVC was an independent risk marker for cardiovascular disease9,16–18. Thus, it seemed plausible to imagine BVCs as predictors of other age-related chronic diseases like osteoporosis. To date, only one study has tested a possible association between low bone mass and BVC in a North American population of Hispanic predominance. It was found that there is a strong independent correlation between the two factors10. In the study, the attempt was made to relate low bone mass to the vascular disease detected by BVC. However, the authors reported a possible screening bias in their study because, as screening for osteoporosis by means of DXA is recommended for women older than 65, the cases they included of younger women may have had additional risk factors for low bone mass. In our study, we required that all postmenopausal patients, regardless of age, underwent DXA. This probably corrected the screening bias. The study population was another point of divergence. In the study by Reddy and colleagues10, the population was predominantly Hispanic, whereas in ours it was mostly Caucasian, a population at higher risk for osteoporosis19. Furthermore, Reddy and colleagues evaluated the patients’ medical histories without having any contact with them10. In our study, along with the review of their medical histories, all
Table 3 Comparison between the groups with and without breast vascular calcifications (BVC) in terms of results of dual-energy X-ray absorptiometry and diagnosis of osteoporosis and osteopenia. Data are given as mean ⫹ ⫺ standard deviation or frequency (%) Parameter Bone mineral density Spine (g/cm2) T-score for spine Neck (g/cm2) T-score for neck Total femur (g/cm2) T-score for total femur Osteoporosis or osteopenia At spine osteoporosis osteopenia osteoporosis or osteopenia osteoporosis or (moderate/severe) osteopenia At hip osteoporosis osteopenia osteoporosis or osteopenia osteoporosis or (moderate or severe) osteopenia At any site osteoporosis osteoporosis or osteopenia osteoporosis or (moderate/severe) osteopenia
Without BVC (n ⫽ 173)
With BVC (n ⫽ 38)
1.052 ⫹ ⫺ 0.149 ⫺1.19 ⫹ ⫺ 1.22 0.867 ⫹ ⫺ 0.118 ⫺0.98 ⫹ ⫺ 1.00 0.914 ⫹ ⫺ 0,125 ⫺0.73 ⫹ ⫺ 1.04
1.084 ⫹ ⫺ 0.220 ⫺0.934 ⫹ ⫺ 1.813 0.850 ⫹ ⫺ 0.151 ⫺1.14 ⫹ ⫺ 1.20 0.894 ⫹ ⫺ 0.157 ⫺0.83 ⫹ ⫺ 1.35
p Value 0.390* 0.403* 0.535* 0.395* 0.422* 0.675*
(12.7%) (50.0%) (63.0%) (43.9%)
6 18 24 19
(15.8%) (47.4%) (63.2%) (50.0%)
0.613‡ 0.744‡ 0.986‡ 0.496‡
12 (6.9%) 85 (49.4%) 97 (56.1%) 57 (32.9%)
4 23 27 20
(10.5%) (60.5%) (71.1%) (52.6%)
0.862** 0.203‡ 0.089‡ 0.022‡
22 87 109 76
27 (15.6%) 135 (78.0%) 93 (53.8%)
9 (23.7%) 29 (76.3%) 25 (65.8%)
0.231‡ 0.818‡ 0.176‡
*, Student t-test for independent samples; ‡, χ2 test; **, Fisher exact test
Climacteric
489
Breast vascular calcifications and bone mass
Nasser et al.
Table 4 Logistic regression analysis with diagnosis of osteoporosis as the dependent variable Independent variable
Odds ratio (95% confidence interval)
p Value
Model 1: Osteoporosis at lumbar spine as the dependent variable Age 1.08 (1.02–1.14) Body mass index 0.88 (0.79–0.98) Sedentary lifestyle 1.77 (0.65–4.84) Smoking habits 1.25 (0.42–3.77) BAH or DM or dyslipidemia or any CVD 1.17 (0.40–3.39) Breast vascular calcification 0.65 (0.18–2.38)
0.009 0.018 0.265 0.689 0.774 0.515
Model 2: Osteoporosis at hip as the dependent variable Age Body mass index Sedentary lifestyle Smoking habits BAH or DM or dyslipidemia or any CVD Breast vascular calcification
1.14 0.79 1.19 1.97 1.00 0.51
(1.05–1.23) (0.67–0.94) (0.34–4.20) (0.47–8.17) (0.26–3.79) (0.10–2.66)
0.001 0.006 0.782 0.352 0.994 0.423
Model 3: Osteoporosis at any site as the dependent variable Age Body mass index Sedentary lifestyle Smoking habits BAH or DM or dyslipidemia or any CVD Breast vascular calcification
1.08 0.86 1.46 1.47 1.00 1.05
(1.02–1.14) (0.78–0.95) (0.60–3.59) (0.53–4.06) (0.38–2.62) (0.33–3.38)
0.004 0.003 0.406 0.456 0.999 0.938
BAH, blood arterial hypertension; DM, diabetes mellitus; CVD, cardiovascular disease
Wparticipants were personally interviewed by either of two physicians (E.J.N. and L.M.P.). Reddy and colleagues reported not having access to relevant clinical data, such as diagnoses of diabetes and hypertension, in 23 of the 228 participants (around 10%)10, while we were able to obtain data for all participants in our study. On the other hand, in analyzing the features of the groups with and without BVC, some characteristics stood out. The group without BVC exhibited a statistically significant higher rate of present or past smokers than the group with BVC. Furthermore, the former group had a larger number of participants, albeit only near significance, with a family history of osteoporosis than the group with BVC. These two factors
may explain lower BMDs than expected in the group without BVC if the percentage of smokers and of women with significant family history had been similar to those in the group with BVC. It is conceivable that such differences may have contributed to the inadequacy of BVC to detect osteoporosis or osteopenia. It is possible that our results were influenced by the participants’ BMI, whose average value corresponded to the diagnosis of overweight and the protective effect of high BMI on bone mass is well recognized. Although the BMI was higher in the group with BVC, it was not statistically different from the value observed in the group without BVC. Moreover, the goal was to test the ability of BVC to better select candidates
Table 5 Sensitivity and specificity and their 95% confidence interval (95% CI) for breast vascular calcification as a screening tool for osteoporosis or for osteoporosis or moderate or severe osteopenia To detect At lumbar spine or hip Osteoporosis Osteoporosis or moderate or severe osteopenia At lumbar spine Osteoporosis Osteoporosis or moderate or severe osteopenia At hip Osteoporosis Osteoporosis or moderate or severe osteopenia
490
Sensitivity (95% CI)
Specificity (95% CI)
25.0% (10.9–39.1) 20.3% (13.1–27.6)
83.4% (77.9–88.9) 86.0% (79.0–93.1)
17.9% (3.7–32.0) 18.9% (11.1–26.8)
82.5% (77.0–88.0) 83.6% (76.9–90.4)
25.0% (3.8–46.2) 24.7% (15.0–34.3)
83.1% (77.8–88.3) 86.6% (80.8–92.3)
Climacteric
Breast vascular calcifications and bone mass for DXA from a population of postmenopausal women; thus we understand that overweight or obese participants should not be excluded, and it is known that such diagnoses are frequent in postmenopausal women. In our multivariate analysis, only age and BMI were shown to be predictors with statistical significance of a diagnosis of osteoporosis. This result is in agreement with the clinical Risk Assessment Tool (OsteoRisk), which takes into account only age and weight to screen women at higher risk of osteoporosis20. Our group found a sensitivity of 86.5% for identifying women with osteoporosis in a population of 812 postmenopausal women with those two factors only21. Our study had some limitations. First, the fact that it was a cross-sectional study, which allows one to see only the data of the moment, preventing an assessment over time. Second, data collection was based on information contained in medical histories or supplied by the patients themselves, raising the possibility of memory biases. Third and last, although all densitometry tests were performed by the same type of equipment, they were not done by the same machine, or not even by the same technologist. On the other hand, there are some strong points. The same radiologist, blinded to all clinical information, evaluated all
Nasser et al. mammograms. Only two gynecologists examined all patients clinically and administered the questionnaires. Finally, the recalculation of the sample size during the study was yet another strong point, which made it possible to keep the power of our sample size adequate for our purposes. Thus, our data allow us to conclude that BVCs detected on mammograms are inadequate for identifying postmenopausal women with osteoporosis or osteopenia, at least in the Brazilian population that took part in our study. An effort should be made to find appropriate tools for screening low bone mass, especially when access to densitometry is limited. Confict of interest L. M. Pompei is a lecturer for the following pharmaceutical companies: Bayer, MSD, Libbs, and TEVA. C. E. Fernandes is a lecturer for the following pharmaceutical companies: Bayer, Sanofi and TEVA. E. R. Iglésias, E. J. Nasser and J. A. S. Ferreira have no confl ict of interest to declare. Source of funding This study was supported exclusively by the authors themselves.
References 1. Notomi T, Okimoto N, Okazaki Y, Nakamura T, Suzuki M. Tower climbing exercise started 3 months after ovariectomy recovers bone strength of the femur and lumbar vertebrae in aged osteopenic rats. J Bone Miner Res 2003;18:140–9 2. United States Census Bureau. International Programs Data. 2013. Available at: http://www.census.gov/population/international/data 3. Regional Advisory Council for the Latin America region Auditoria regional da America Latina. The Latin America Regional Audit. Epidemiology, costs & burden of osteoporosis in 2012. Nyon, Switzerland: International Osteoporosis Foundation, 2012. Available at: http://www.osteoporosisinlatinamerica.com 4. Pinheiro MM, dos Reis Neto ET, Machado FS, et al. Risk factors for osteoporotic fractures and low bone density in pre and postmenopausal women. Rev Saúde Pública 2010;44:479–85 5. von der Recke P, Hansen MA, Hassager C. The association between low bone mass at the menopause and cardiovascular mortality. Am J Med 1999;106:273–8 6. Wild RA. Coronary artery calcification and osteoporosis: is estrogen an important link? Menopause 2005;12:661–3 7. Schulz E, Arfai K, Liu X, Sayre J, Gilsanz V. Aortic calcification and the risk of osteoporosis and fractures. J Clin Endocrinol Metab 2003;89:4246–53 8. Wiklund P, Nordström A, Jansson JH, Weinehall L, Nordström P. Low bone mineral density is associated with increased risk for myocardial infarction in men and women. Osteoporos Int 2012;23:963–70 9. Ferreira JAS, Pompei LM, Fernandes CE, Azevedo LH,Peixoto S. Breast arterial calcification is a predictive factor of cardiovascular disease in Brazilian postmenopausal women. Climacteric 2009;12:439–44 10. Reddy J, Bilezikian JP, Smith SJ, Mosca L. Reduced bone mineral density is associated with breast arterial calcification. J Clin Endocrin Metab 2008;93:208–11
Climacteric
11. World Health Organization. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Geneva, Switzerland: WHO, 1994,1–129 12. Sickles EA. Breast calcifications: mammographic evaluation. Radiology 1986;160:289–93 13. Baum JK, Comstock CH, Joseph L. Intramammary arterial calcifications with diabetes. Radiology 1980;136:61–2 14. Kemmeren JM, van Nord PA, Beijerinck D, et al. Arterial calcification found on breast cancer screening mammograms and cardiovascular mortality in women: The DOM Project. Am J Epidemiol 1998;147:333–41 15. Siris ES, Miller PD, Barrett-Connor E, et al. Identification and fracture outcomes of undiagnosed low bone mineral density in postmenopausal women: results from the National Osteoporosis Risk Assessment. JAMA 2001;286:2815–22 16. Crystal P, Crystal E, Leor J, et al. Breast arterial calcification on routine mammography as a potential marker for increased cardiovascular disease. Am J Cardiol 2000;86:216–17 17. Iribarren C, Go AS, Tolstykh I, et al. Breast vascular calcification and risk of coronary heart disease, stroke and heart failure. J Women’s Health 2004;13:381–9 18. Kemmeren JM, Beijerinck D, van Noord PA, et al. Breast arterial calcifications: associations with diabetes mellitus and cardiovascular mortality. Radiology 1996;201:75–8 19. Cauley JA, Wu L, Wampler NS, et al. Clinical risk factors for fractures in multi-ethnic women: the Women’s Health Initiative. J Bone Miner Res 2007;22:1816–26 20. Sen SS, Rives VP, Messina OD, et al. A risk assessment tool (OsteoRisk) for identifying Latin American women with osteoporosis. J Gen Intern Med 2005;20:245–50 21. Steiner ML, Fernandes CE, Strufaldi R, et al. Application of OsteoRisk to postmenopausal patients with osteoporosis. Sao Paulo Med J 2010;128:24–9
491
Copyright of Climacteric is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Association of breast vascular calcifications with low bone mass in postmenopausal women E. J. Nasser∗, E. R. Iglésias†, J. A. S. Ferreira∗, C. E. Fernandes∗ and L. M. Pompei∗ ∗ABC School of Medicine, Gynecology Branch, Santo André; †Diagnosis Clinic ‘Diagnóstico por Imagem de Santo André – DISA’, Santo André, Brazil Key words: OSTEOPOROSIS, BREAST VASCULAR CALCIFICATIONS, SCREENING, LOW BONE MASS, OSTEOPENIA
ABSTRACT Background In developing countries, there is a deficiency of densitometers with which to screen the population for osteoporosis. Thus, strategies with which to select patients for a bone density test are desirable. Objective To determine whether breast vascular calcifications (BVCs) may be employed to identify postmenopausal women with osteoporosis/osteopenia. Methods This was a cross-sectional study of postmenopausal women subjected to bilateral mammography and bone densitometry (DXA) of the spine and hip. A medical interview registered possible confounding factors, such as age, length of menopause, previous use of postmenopausal hormone therapy, family history of osteoporosis, smoking, alcoholism, hypertension, diabetes, cardiovascular diseases, and medication use. Results The study included 211 postmenopausal women aged 62.1 ⫹ ⫺ 9.3 years, 38 of whom (18.0%) exhibited BVC. Osteoporosis was detected in 36 (17.1%), and a T-score ⬍ ⫺1.0 for any site was found in 164 (77.7%). No statistically significant difference was found between the groups without BVC (n ⫽ 173) and with BVC (n ⫽ 38) for the prevalence of ‘osteoporosis’ or ‘moderate/severe osteopenia or osteoporosis’ at the spine or at any other site. There was a difference between the groups in terms of age (59.0 ⫹ ⫺ 7.8 vs. 71.9 ⫹ 8.9 years, respectively; p ⬍ 0.001), sedentary lifestyle (57.8% vs. 84.2%, respectively; p ⫽ 0.002), ⫺ smoking (27.7% vs. 7.9%, respectively; p ⫽ 0.009), and high blood pressure (65.3% vs. 92.1%, respectively; p ⫽ 0.001). Logistic regression analysis confirmed the lack of statistical significance for BVC as a predictor of an osteoporosis diagnosis. Sensitivity values of BVCs to detect osteoporosis or osteopenia ranged from 17.9% to 25.0%. Conclusion BVCs have been shown to be inadequate to identify postmenopausal women with osteoporosis or osteopenia.
INTRODUCTION Osteoporosis, one of the main public health problems in the world1, is a disease characterized by reduction in bone density and quality. Its course is silent and bone becomes progressively porous and weak, more so with age and menopause, thus leading to an increasingly greater risk of fractures. The world population is estimated to be over 7 billion people, 846 million of whom are more than 60 years old. With the increase in life expectancy, it is expected that in 2050 there will be above 1.1 billion people over 70 years world-wide2,
consequently increasing the number of people with chronic diseases like osteoporosis. The chief means of diagnosing and monitoring osteoporosis is bone densitometry, preferably using the dual-energy X-ray absorptiometry (DXA) method. In Latin America, access to DXA is quite limited. Brazil and Chile are the countries with the most accessibility (ten machines per million inhabitants) according to the Latin American Regional Audit 2012 report. In Brazil, there are approximately 1850 densitometers with limited access in the public health system, and the cost of each examination is in the neighborhood of US$ 403. According to
Correspondence: Dr L. M. Pompei, ABC School of Medicine, Gynecology Branch, Rua Dr. Procopio Ribeiro dos Santos 84, Sao Paulo, 04664-130 Brazil; E-mail: [email protected]
ORIGINAL ARTICLE © 2014 International Menopause Society DOI: 10.3109/13697137.2013.869672
Received 27-09-2013 Revised 18-11-2013 Accepted 23-11-2013
Breast vascular calcifications and bone mass the World Health Organization (WHO), with the increase in the population over 65 years in Latin America in the next 50 years, the number of hip fractures will be equivalent to that in the United States or in the European countries today4. It is known, through epidemiological studies, that osteoporosis in postmenopausal women is associated with cardiovascular disease (CVD)5 and that osteoporosis in the lumbar spine and proximal femur is associated with coronary and aortic vascular calcifications6,7. Low bone mineral density (BMD) has been further associated with increased risk of myocardial infarction regardless of other risk factors8. There is also a study which points to the high probability of the connection between breast vascular calcifications (BVC) and CVD9. Although epidemiological studies show the positive association between coronary and aortic vascular calcifications and osteoporosis of the spine and hip, thus far there is only one retrospective study with 228 North American women revealing the connection between BMD and BVC10. The fact that BVC can be easily observed in routine mammography, carried out to screen for breast cancer in postmenopausal women, sparked the interest in conducting a study with the purpose of relating this radiological finding to BMD in a population of Brazilian postmenopausal women. Besides, there is a much greater availability of mammography machines in the public health system in our country than densitometers, and thus the examination would be of great practical value, for it would help to detect perceptible changes signaling the need for performing a densitometry, helping to optimize the use of densitometers.
METHODS This is a cross-sectional study to evaluate the association between BVC and low bone mass. All study procedures were approved by the Internal Review Board of the ABC School of Medicine, and all study participants gave their written informed consent. The women participating in the study were being treated at the gynecology outpatient clinic. To be included in the study, the women had to be in postmenopause, aged 45 years or older, to have undergone mammography and bone densitometry of the lumbar spine and hip in the previous 12 months with no more than a 6-month interval between the two examinations. Diagnosis of postmenopause was established for a woman over 45 years old with intact uterus if she presented amenorrhea for at least 12 months. For a hysterectomized woman, we considered her to be postmenopausal when two serum follicle stimulating hormone and estradiol levels with a minimum interval of 60 days between them were ⬎ 30 UI/l and ⬍ 35 pg/ml, respectively. Those patients using any medication that might interfere with bone metabolism, such as bisphosphonates, estrogen therapy, systemic steroids, anticoagulants, anticonvulsants and aromatase inhibitors, and who were carriers of uncompensated
Climacteric
Nasser et al. thyroid dysfunction, hyperparathyroidism, osteogenesis imperfecta, Paget’s bone disease, rheumatoid arthritis, malignant neoplasia (except for non-melanoma skin cancer), and of mammographic changes categorized as BI-RADS® 4 and 5 were excluded from the study. Following inclusion, all women were subjected to a standard questionnaire aimed at gathering information about possible confounding factors, such as age, ethnicity, profession, schooling, age at last menstruation, length of menopause, previous use of postmenopausal hormone therapy, a sedentary lifestyle, a family history of osteoporosis, smoking habits, alcoholism, hypertension, dyslipidemia, diabetes, cardiovascular diseases, as well as medication use. Weight and height, as well as any previous osteoporotic fractures, were recorded at the time of bone densitometry. Participants were required to have had bone densitometry measured by the DXA technique (Lunar, General Electric Healthcare, Chalfont St. Giles, UK) within the previous 12 months. Results were classified as osteoporosis or osteopenia in accordance with the WHO criteria: osteopenia when the T-score was less than -1.0 and more than ⫺2.5 (mild: ⬍ ⫺1.0 and ⫺ ⬎ ⫺1.5; moderate: ⬍ ⫺1.5 and ⫺ ⬎ ⫺2.0; severe: ⬍ ⫺2.0 11 and ⬎ ⫺2.5) and osteoporosis when the T-score ⬍ ⫺ ⫺2.5 . The mammograms were reviewed by a single radiology technologist (E.R.I.) blinded to the participants’ clinical data and bone mass. Each examination was classified as positive or negative for BVCs based on the easily recognizable characteristic pattern of amorphous calcifications between two parallel lines12. Sample size was initially calculated considering the prevalence of BVC to be approximately 10%9,13,14, and of low bone mass, around 40%, for the type of population included in the study10,15. Based on this assumption, 29 BVC cases and a total of 290 participants were deemed necessary for an 80% power to detect an odds ratio of 3.0 for BVCs to be a predictor of low bone mass, considering a type I error of 5%. As the exact prevalence of BVC and low bone mass in the study population were not known, a definite sample size was calculated after gathering data from the first 150 participants (interim analysis). Since the prevalence of BVC was approximately 18% and the prevalence of a T-score ⬍ ⫺1.5 was 44% at any site, it was estimated that 33 BVC cases and a total of 184 participants would be necessary as long as the other parameters remained the same. An additional 15% of cases were included as a safety measure for possible loss of data or variations in known prevalences. The continuous numerical data were presented as mean ⫹ ⫺ standard deviation. Categorical data were presented as absolute values and percentage rates. The comparisons of continuous numerical data between the groups were made with the Student t-test for independent samples when the distribution was normal and with the Mann–Whitney test when the distribution was not normal. The categorical data were compared using the χ2 test, except when the expected values in contingency tables were less than 5 in more than 20% of the cells. In this case, the Fisher exact test was employed. All data were included in an electronic spreadsheet
487
Breast vascular calcifications and bone mass
Nasser et al.
(Microsoft® Excel® 2007, Microsoft, Redmond, USA) and statistical analysis was performed with the WinSTAT® software, version 2007.1 (R. Fitch Software, Bad Krozingen, Germany). Multivariate logistic regressions were carried out in which the independent variables were age, body mass index (BMI), sedentary lifestyle, present or past smoking habits, diagnosis of high blood pressure, diabetes, dyslipidemia, or any cardiovascular disease, and presence of BVC. For the dependent variable, three models were analyzed, one with spine osteoporosis, another with hip osteoporosis, and the third one with osteoporosis at any site. The logistic regressions were performed with the OpenStat® v. 2013 software (Miller WG). The sensitivities and specificities of BVCs were also calculated for the detection of moderate or severe osteopenia and detection of osteoporosis.
RESULTS Of the 211 patients included in the study, 166 were of white ethnicity (78.7%), 40 of black ethnicity (18.9%), and five of yellow ethnicity (2.4%). BVCs were identified in 38 participants (18.0%). The prevalence of low BMD (osteoporosis or osteopenia) at the lumbar spine was 63.0% (133 cases), and at the hip, 58.8% (124 cases). The prevalence of osteoporosis at the spine was 13.3% (28 cases), and at the hip, 7.6% (16 cases). Other data from the total sample are displayed in Table 1.
Table 1 Characteristics of the study population. Data are given as mean ⫹ ⫺ standard deviation or frequency (%) 62.1 ⫹ ⫺ 9.3 48.9 ⫹ ⫺ 4.4 13.4 ⫹ ⫺ 9.4 68.4 ⫹ ⫺ 13.5 153.2 ⫹ ⫺ 5.5 29.1 ⫹ ⫺ 5.3
Age at densitometry (years) Age at menopause (years) Length of postmenopause (years) Weight (kg) Height (cm) Body mass index (kg/m2) Bone mass density Lumbar spine (g/cm2) T-score for lumbar spine Neck (g/cm2) T-score for neck Total femur (g/cm2) T-score for total femur Osteoporosis or osteopenia Osteoporosis at spine T-score ⬍ ⫺1.0 at spine Osteoporosis at hip T-score ⬍ ⫺1.0 at hip Osteoporosis at any site T-score ⬍ ⫺1.0 at any site
1.058 ⫹ ⫺ 0.164 ⫺1.15 ⫹ ⫺ 1.35 0.864 ⫹ ⫺ 0.125 ⫺1.01 ⫹ ⫺ 1.03 0.910 ⫹ ⫺ 0.131 ⫺0.75 ⫹ ⫺ 1.10 28 133 16 124 36 164
(13.3%) (63.0%) (7.6%) (58.8%) (17.1%) (77.7%)
Table 2 shows the comparison between the group with BVCs and that without BVCs in terms of some clinical parameters. Statistically significant differences between the two groups were found for the following variables: age, length of postmenopause, sedentary lifestyle, smoking habits, hypertension, diabetes, dyslipidemia, or any cardiovascular disease.
Table 2 Comparison between the groups with and without breast vascular calcifications (BVC) in terms of clinical parameters. Data are given as mean ⫹ standard deviation or frequency (%)
⫺
Parameter
Without BVC (n ⫽ 173)
With BVC (n ⫽ 38)
Age (at DXA, years) Age at menopause (years) Length of postmenopause (years) Weight (kg) Height (cm) Body mass index (kg/m2) Sedentary lifestyle Postmenopausal hormone therapy in past Family history of osteoporosis Smoker (present or past) Smoker (only present) Hypertension Diabetes Hyperlipidemia BAH or DM or dyslipidemia or any CVD Previous fracture due to bone weakness
59.9 ⫹ ⫺ 7.8 48.8 ⫹ ⫺ 4.6 11.4 ⫹ ⫺ 8.1 68.4 ⫹ ⫺ 13.4 153.8 ⫹ ⫺ 5.5 28.9 ⫹ ⫺ 5.2 100 (57.8%) 19 (11.0%) 30 (17.3%) 48 (27.7%) 22 (12.7%) 81 (46.8%) 27 (15.6%) 63 (36.4%) 113 (65.3%) 0
71.9 ⫹ ⫺ 8.9 49.6 ⫹ ⫺ 3.0 22.3 ⫹ ⫺ 9.9 68.1 ⫹ ⫺ 13.8 150.4 ⫹ ⫺ 4.7 30.2 ⫹ ⫺ 5.7 32 (84.2%) 2 (5.3%) 2 (5.3%) 3 (7.9%) 1 (2.6%) 28 (73.7%) 10 (26.3%) 17 (44.7%) 35 (92.1%) 0
p Value ⬍ 0.001* 0.405† ⬍ 0.001† 0.939† 0.003* 0.242* 0.002‡ 0.229** 0.060‡ 0.009‡ 0.053** 0.003‡ 0.116‡ 0.338‡ 0.001‡ NA
DXA, dual-energy X-ray absorptiometry; BAH, blood arterial hypertension; DM, diabetes mellitus; CVD, cardiovascular disease; NA, not applicable *, Student t-test for independent samples; †, Mann–Whitney U-test; ‡, χ2 test; **, Fisher exact test
488
Climacteric
Breast vascular calcifications and bone mass Table 3 shows the comparison between the groups with respect to DXA results and diagnosis of osteoporosis and osteopenia. Only the following parameter had statistical significance: osteoporosis or moderate or severe osteopenia at the hip (T-score ⬍ ⫺1.5). Logistic regression analysis revealed that in all models (spine osteoporosis, hip osteoporosis, or osteoporosis at any site as a dependent variable) only age and BMI exhibited a statistically significant effect. In none of the models did BVC have the statistical significance for predicting the osteoporosis diagnoses. Table 4 shows the details of the logistic regression. Table 5 displays the sensitivity and specificity values of BVCs to detect osteoporosis and low bone mass.
DISCUSSION This is the first study which has attempted to evaluate the association between BVC and low BMD in a Brazilian population. Our results did not show any association, except in the case of osteoporosis or moderate/severe osteopenia at the hip (T-score ⬍ ⫺1.5). The regression logistic analyses confirmed the lack of association. The low sensitivity of BVC for detection of osteoporosis or low bone mass became evident.
Nasser et al. Several authors, including those in our group, have shown that BVC was an independent risk marker for cardiovascular disease9,16–18. Thus, it seemed plausible to imagine BVCs as predictors of other age-related chronic diseases like osteoporosis. To date, only one study has tested a possible association between low bone mass and BVC in a North American population of Hispanic predominance. It was found that there is a strong independent correlation between the two factors10. In the study, the attempt was made to relate low bone mass to the vascular disease detected by BVC. However, the authors reported a possible screening bias in their study because, as screening for osteoporosis by means of DXA is recommended for women older than 65, the cases they included of younger women may have had additional risk factors for low bone mass. In our study, we required that all postmenopausal patients, regardless of age, underwent DXA. This probably corrected the screening bias. The study population was another point of divergence. In the study by Reddy and colleagues10, the population was predominantly Hispanic, whereas in ours it was mostly Caucasian, a population at higher risk for osteoporosis19. Furthermore, Reddy and colleagues evaluated the patients’ medical histories without having any contact with them10. In our study, along with the review of their medical histories, all
Table 3 Comparison between the groups with and without breast vascular calcifications (BVC) in terms of results of dual-energy X-ray absorptiometry and diagnosis of osteoporosis and osteopenia. Data are given as mean ⫹ ⫺ standard deviation or frequency (%) Parameter Bone mineral density Spine (g/cm2) T-score for spine Neck (g/cm2) T-score for neck Total femur (g/cm2) T-score for total femur Osteoporosis or osteopenia At spine osteoporosis osteopenia osteoporosis or osteopenia osteoporosis or (moderate/severe) osteopenia At hip osteoporosis osteopenia osteoporosis or osteopenia osteoporosis or (moderate or severe) osteopenia At any site osteoporosis osteoporosis or osteopenia osteoporosis or (moderate/severe) osteopenia
Without BVC (n ⫽ 173)
With BVC (n ⫽ 38)
1.052 ⫹ ⫺ 0.149 ⫺1.19 ⫹ ⫺ 1.22 0.867 ⫹ ⫺ 0.118 ⫺0.98 ⫹ ⫺ 1.00 0.914 ⫹ ⫺ 0,125 ⫺0.73 ⫹ ⫺ 1.04
1.084 ⫹ ⫺ 0.220 ⫺0.934 ⫹ ⫺ 1.813 0.850 ⫹ ⫺ 0.151 ⫺1.14 ⫹ ⫺ 1.20 0.894 ⫹ ⫺ 0.157 ⫺0.83 ⫹ ⫺ 1.35
p Value 0.390* 0.403* 0.535* 0.395* 0.422* 0.675*
(12.7%) (50.0%) (63.0%) (43.9%)
6 18 24 19
(15.8%) (47.4%) (63.2%) (50.0%)
0.613‡ 0.744‡ 0.986‡ 0.496‡
12 (6.9%) 85 (49.4%) 97 (56.1%) 57 (32.9%)
4 23 27 20
(10.5%) (60.5%) (71.1%) (52.6%)
0.862** 0.203‡ 0.089‡ 0.022‡
22 87 109 76
27 (15.6%) 135 (78.0%) 93 (53.8%)
9 (23.7%) 29 (76.3%) 25 (65.8%)
0.231‡ 0.818‡ 0.176‡
*, Student t-test for independent samples; ‡, χ2 test; **, Fisher exact test
Climacteric
489
Breast vascular calcifications and bone mass
Nasser et al.
Table 4 Logistic regression analysis with diagnosis of osteoporosis as the dependent variable Independent variable
Odds ratio (95% confidence interval)
p Value
Model 1: Osteoporosis at lumbar spine as the dependent variable Age 1.08 (1.02–1.14) Body mass index 0.88 (0.79–0.98) Sedentary lifestyle 1.77 (0.65–4.84) Smoking habits 1.25 (0.42–3.77) BAH or DM or dyslipidemia or any CVD 1.17 (0.40–3.39) Breast vascular calcification 0.65 (0.18–2.38)
0.009 0.018 0.265 0.689 0.774 0.515
Model 2: Osteoporosis at hip as the dependent variable Age Body mass index Sedentary lifestyle Smoking habits BAH or DM or dyslipidemia or any CVD Breast vascular calcification
1.14 0.79 1.19 1.97 1.00 0.51
(1.05–1.23) (0.67–0.94) (0.34–4.20) (0.47–8.17) (0.26–3.79) (0.10–2.66)
0.001 0.006 0.782 0.352 0.994 0.423
Model 3: Osteoporosis at any site as the dependent variable Age Body mass index Sedentary lifestyle Smoking habits BAH or DM or dyslipidemia or any CVD Breast vascular calcification
1.08 0.86 1.46 1.47 1.00 1.05
(1.02–1.14) (0.78–0.95) (0.60–3.59) (0.53–4.06) (0.38–2.62) (0.33–3.38)
0.004 0.003 0.406 0.456 0.999 0.938
BAH, blood arterial hypertension; DM, diabetes mellitus; CVD, cardiovascular disease
Wparticipants were personally interviewed by either of two physicians (E.J.N. and L.M.P.). Reddy and colleagues reported not having access to relevant clinical data, such as diagnoses of diabetes and hypertension, in 23 of the 228 participants (around 10%)10, while we were able to obtain data for all participants in our study. On the other hand, in analyzing the features of the groups with and without BVC, some characteristics stood out. The group without BVC exhibited a statistically significant higher rate of present or past smokers than the group with BVC. Furthermore, the former group had a larger number of participants, albeit only near significance, with a family history of osteoporosis than the group with BVC. These two factors
may explain lower BMDs than expected in the group without BVC if the percentage of smokers and of women with significant family history had been similar to those in the group with BVC. It is conceivable that such differences may have contributed to the inadequacy of BVC to detect osteoporosis or osteopenia. It is possible that our results were influenced by the participants’ BMI, whose average value corresponded to the diagnosis of overweight and the protective effect of high BMI on bone mass is well recognized. Although the BMI was higher in the group with BVC, it was not statistically different from the value observed in the group without BVC. Moreover, the goal was to test the ability of BVC to better select candidates
Table 5 Sensitivity and specificity and their 95% confidence interval (95% CI) for breast vascular calcification as a screening tool for osteoporosis or for osteoporosis or moderate or severe osteopenia To detect At lumbar spine or hip Osteoporosis Osteoporosis or moderate or severe osteopenia At lumbar spine Osteoporosis Osteoporosis or moderate or severe osteopenia At hip Osteoporosis Osteoporosis or moderate or severe osteopenia
490
Sensitivity (95% CI)
Specificity (95% CI)
25.0% (10.9–39.1) 20.3% (13.1–27.6)
83.4% (77.9–88.9) 86.0% (79.0–93.1)
17.9% (3.7–32.0) 18.9% (11.1–26.8)
82.5% (77.0–88.0) 83.6% (76.9–90.4)
25.0% (3.8–46.2) 24.7% (15.0–34.3)
83.1% (77.8–88.3) 86.6% (80.8–92.3)
Climacteric
Breast vascular calcifications and bone mass for DXA from a population of postmenopausal women; thus we understand that overweight or obese participants should not be excluded, and it is known that such diagnoses are frequent in postmenopausal women. In our multivariate analysis, only age and BMI were shown to be predictors with statistical significance of a diagnosis of osteoporosis. This result is in agreement with the clinical Risk Assessment Tool (OsteoRisk), which takes into account only age and weight to screen women at higher risk of osteoporosis20. Our group found a sensitivity of 86.5% for identifying women with osteoporosis in a population of 812 postmenopausal women with those two factors only21. Our study had some limitations. First, the fact that it was a cross-sectional study, which allows one to see only the data of the moment, preventing an assessment over time. Second, data collection was based on information contained in medical histories or supplied by the patients themselves, raising the possibility of memory biases. Third and last, although all densitometry tests were performed by the same type of equipment, they were not done by the same machine, or not even by the same technologist. On the other hand, there are some strong points. The same radiologist, blinded to all clinical information, evaluated all
Nasser et al. mammograms. Only two gynecologists examined all patients clinically and administered the questionnaires. Finally, the recalculation of the sample size during the study was yet another strong point, which made it possible to keep the power of our sample size adequate for our purposes. Thus, our data allow us to conclude that BVCs detected on mammograms are inadequate for identifying postmenopausal women with osteoporosis or osteopenia, at least in the Brazilian population that took part in our study. An effort should be made to find appropriate tools for screening low bone mass, especially when access to densitometry is limited. Confict of interest L. M. Pompei is a lecturer for the following pharmaceutical companies: Bayer, MSD, Libbs, and TEVA. C. E. Fernandes is a lecturer for the following pharmaceutical companies: Bayer, Sanofi and TEVA. E. R. Iglésias, E. J. Nasser and J. A. S. Ferreira have no confl ict of interest to declare. Source of funding This study was supported exclusively by the authors themselves.
References 1. Notomi T, Okimoto N, Okazaki Y, Nakamura T, Suzuki M. Tower climbing exercise started 3 months after ovariectomy recovers bone strength of the femur and lumbar vertebrae in aged osteopenic rats. J Bone Miner Res 2003;18:140–9 2. United States Census Bureau. International Programs Data. 2013. Available at: http://www.census.gov/population/international/data 3. Regional Advisory Council for the Latin America region Auditoria regional da America Latina. The Latin America Regional Audit. Epidemiology, costs & burden of osteoporosis in 2012. Nyon, Switzerland: International Osteoporosis Foundation, 2012. Available at: http://www.osteoporosisinlatinamerica.com 4. Pinheiro MM, dos Reis Neto ET, Machado FS, et al. Risk factors for osteoporotic fractures and low bone density in pre and postmenopausal women. Rev Saúde Pública 2010;44:479–85 5. von der Recke P, Hansen MA, Hassager C. The association between low bone mass at the menopause and cardiovascular mortality. Am J Med 1999;106:273–8 6. Wild RA. Coronary artery calcification and osteoporosis: is estrogen an important link? Menopause 2005;12:661–3 7. Schulz E, Arfai K, Liu X, Sayre J, Gilsanz V. Aortic calcification and the risk of osteoporosis and fractures. J Clin Endocrinol Metab 2003;89:4246–53 8. Wiklund P, Nordström A, Jansson JH, Weinehall L, Nordström P. Low bone mineral density is associated with increased risk for myocardial infarction in men and women. Osteoporos Int 2012;23:963–70 9. Ferreira JAS, Pompei LM, Fernandes CE, Azevedo LH,Peixoto S. Breast arterial calcification is a predictive factor of cardiovascular disease in Brazilian postmenopausal women. Climacteric 2009;12:439–44 10. Reddy J, Bilezikian JP, Smith SJ, Mosca L. Reduced bone mineral density is associated with breast arterial calcification. J Clin Endocrin Metab 2008;93:208–11
Climacteric
11. World Health Organization. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Geneva, Switzerland: WHO, 1994,1–129 12. Sickles EA. Breast calcifications: mammographic evaluation. Radiology 1986;160:289–93 13. Baum JK, Comstock CH, Joseph L. Intramammary arterial calcifications with diabetes. Radiology 1980;136:61–2 14. Kemmeren JM, van Nord PA, Beijerinck D, et al. Arterial calcification found on breast cancer screening mammograms and cardiovascular mortality in women: The DOM Project. Am J Epidemiol 1998;147:333–41 15. Siris ES, Miller PD, Barrett-Connor E, et al. Identification and fracture outcomes of undiagnosed low bone mineral density in postmenopausal women: results from the National Osteoporosis Risk Assessment. JAMA 2001;286:2815–22 16. Crystal P, Crystal E, Leor J, et al. Breast arterial calcification on routine mammography as a potential marker for increased cardiovascular disease. Am J Cardiol 2000;86:216–17 17. Iribarren C, Go AS, Tolstykh I, et al. Breast vascular calcification and risk of coronary heart disease, stroke and heart failure. J Women’s Health 2004;13:381–9 18. Kemmeren JM, Beijerinck D, van Noord PA, et al. Breast arterial calcifications: associations with diabetes mellitus and cardiovascular mortality. Radiology 1996;201:75–8 19. Cauley JA, Wu L, Wampler NS, et al. Clinical risk factors for fractures in multi-ethnic women: the Women’s Health Initiative. J Bone Miner Res 2007;22:1816–26 20. Sen SS, Rives VP, Messina OD, et al. A risk assessment tool (OsteoRisk) for identifying Latin American women with osteoporosis. J Gen Intern Med 2005;20:245–50 21. Steiner ML, Fernandes CE, Strufaldi R, et al. Application of OsteoRisk to postmenopausal patients with osteoporosis. Sao Paulo Med J 2010;128:24–9
491
Copyright of Climacteric is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
