Validation of Dental Panoramic Radiography Measures for Identifying Postmenopausal Women with Spinal Osteoporosis

Akira Taguchi1 Yoshikazu Suei1 Mitsuhiro Sanada2 Masahiko Ohtsuka3 Takashi Nakamoto1 Hiroomi Sumida3 Koso Ohama2 Keiji Tanimoto3
OBJECTIVE. Measurements of mandibular inferior cortical shape and width detected on dental panoramic radiographs may be a useful screening tool for spinal osteoporosis in postmenopausal women. The purposes of this study were to clarify whether these measures are validated compared with simple screening tools based on questionnaires, such as the osteoporosis self-assessment tool (OST) and whether these measures can be used in postmenopausal women with histories of hysterectomy, oophorectomy, or estrogen use. SUBJECTS AND METHODS. We calculated the diagnostic performances of panoramic measurements and the OST for identifying women with spinal osteoporosis in both 159 healthy postmenopausal and 157 postmenopausal women with histories of hysterectomy, oophorectomy, or estrogen use. Spinal osteoporosis was dened as a bone mineral density T score of 2.5 or less at the lumbar spine. Cortical shape and width were evaluated on dental panoramic radiographs. Receiver operating characteristic curve analyses were used to determine the optimal cutoff thresholds for cortical width and the OST in healthy postmenopausal women. RESULTS. The sensitivity and specicity, respectively, for identifying women with spinal osteoporosis were 89.5% and 33.9% for cortical width, 86.8% and 57.8% for the OST, and 86.8% and 63.6% for cortical shape in healthy postmenopausal women. Sensitivity and specicity, respectively, were 92.5% and 35.0% for cortical width, 72.5% and 58.1% for the OST, and 80.0% and 64.1% for cortical shape in postmenopausal women with histories of hysterectomy, oophorectomy, or estrogen use. CONCLUSION. Dentists may be able to refer postmenopausal women with suspected spinal osteoporosis for bone densitometry on the basis of dental panoramic radiographs with diagnostic performance similar to that of osteoporosis screening tools based on questionnaires.

Received March 8, 2004; accepted after revision May 11, 2004. Supported by a grant-in-aid for scientic research by the Japan Society for the Promotion of Science (grant no. 14571786).
1

Department of Oral and Maxillofacial Radiology, Hiroshima University Hospital, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8553, Japan. Address correspondence to A. Taguchi (akiro@hiroshima-u.ac.jp).
2 Department of Obstetrics and Gynecology, Division of Clinical Medical Science, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima, Japan. 3 Department of Oral and Maxillofacial Radiology, Division of Medical Intelligence and Informatics, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima, Japan.

AJR 2004;183:17551760 0361803X/04/18361755 American Roentgen Ray Society

steoporotic fractures are associated with substantial morbidity rates, increased medical cost, and high mortality risk in the elderly [1]. New equipment for performing bone assessment has been developed and applied to identify individuals, especially postmenopausal women, with increased risk of osteoporotic fractures. However, the results from the National Osteoporosis Risk Assessment showed that large segments of the population of postmenopausal women 50 years old or older with no previous diagnosis of osteoporosis had low skeletal bone mineral density (BMD) [2]. BMD testing for all postmenopausal women may be preferable to reduce the incidence of osteoporotic fractures and subsequent complications; however, it is difcult in clinical practice to refer all postmenopausal women for BMD testing when

considering the cost-effectiveness, limited number of facilities, and trained personnel. Some simple tools based on questionnaires have been designed to identify postmenopausal women who should undergo BMD testing [38]. The sensitivity of these tools in identifying postmenopausal women with low skeletal BMD was reported to be 90% or more, although the specicity was low. On the other hand, the response rate for questionnaires in the Canadian Multicentre Osteoporosis Study was only 42% [6]. This fact implies that many women may not respond to questionnaires if they have no knowledge of or interest in osteoporosis. Recent studies indicate that dental panoramic radiography measures may be useful tools in identifying postmenopausal women with low skeletal BMD [914], high bone turnover rate [15], or high risk of os-

AJR:183, December 2004

1755

Taguchi et al. teoporotic fractures [16]. Because dental panoramic radiography is frequently performed for the diagnosis of diseases affecting the teeth and jaws in general dental practice, the incidental ndings detected on dental panoramic radiographs maybe used to identify women who have no awareness of their low BMD and would benet from BMD testing. However, little is known about whether these measures are valid in comparison with other simple risk indexes based on responses to questionnaires. The osteoporosis self-assessment tool (OST) is one of the simple tools based on questionnaires and is considered to be a useful index for identifying Asian and Caucasian postmenopausal women with low skeletal BMD [7, 17]. Comparison of diagnostic performances between dental panoramic radiography measures and the OST index would be of interest to explore the validation of the panoramic measures. Furthermore, it is likely that the gynecologic histories such as hysterectomy, oophorectomy, or estrogen use might inuence the diagnostic performance of screening tools by altering the ndings of skeletal BMD tests in postmenopausal women [18, 19]. It is still unknown whether dental panoramic radiography measures can be used in postmenopausal women with such histories and in healthy postmenopausal women. The purpose of this study was therefore to examine the diagnostic performances of dental panoramic radiography measurements and the OST index in postmenopausal women with and without histories of hysterectomy, oophorectomy, or estrogen use.
Subjects and Methods
Of 676 women who visited our clinic for BMD assessment between 1996 and 2003, 316 women 33 84 years old (mean [ SD], 56.8 7.7 years) were recruited for this study. Subjects were divided into two groups: 159 healthy postmenopausal women without histories of hysterectomy, oophorectomy, or estrogen use (group A) and 157 postmenopausal women with one or more of these histories (group B). Of the women in group B, 107 (68%) had undergone hysterectomy, 25 (16%) had undergone unilateral oophorectomy, 67 (43%) had undergone bilateral oophorectomy, and 54 (34%) used estrogen. Excluded from the study were postmenopausal women who had used tobacco or medications that affect bone metabolism except estrogen or who had metabolic bone diseases, diabetes, signicant renal impairment, bone-destructive lesions in the jaw, nonvertebral osteoporotic fractures, and vertebral osteoporotic fracture detected on radiography at skeletal BMD assessment. The vertebral fracture was assessed using a semiquantitative assessment [20]. All subjects gave informed consent before inclusion. Hiroshima University institutional human subjects committee approved obtaining dental panoramic radiographs in subjects who had given informed consent. BMD at the lumbar spine (L2L4) was measured by a dual-energy X-ray absorptiometry (DEXA) scanner (DPX-alpha, Lunar). The in vivo short-term precision error for spine BMD in our clinic is 1.0%. Spine BMD was categorized as normal (T score > 1.0), osteopenic (T score, 1.0 to 2.5), or osteoporotic (T score < 2.5), according to the World Health Organization (WHO) classication [21]. Height and weight were measured at the time of the DEXA scanning. All dental panoramic radiographs were obtained at the time of the DEXA scan with an AZ3000 scanner (Asahi) at 12 mA and 15 sec; the kilovoltage varied between 70 and 80 kV. Screens of speed group 200 (HG-M, Fuji Photo lm) and lm (UR-2, Fuji Photo lm) were used. Two dental panoramic radiography measures, mandibular cortical shape and width, were estimated on dental panoramic radiographs by one oral radiologist with 16 years of clinical experience. Measurement of mandibular cortical width was made bilaterally on the radiographs at the site of the mental foramen according to our previous study [10] (Fig. 1). We drew a line parallel to the long axis of the mandible and tangential to the inferior border of the mandible and constructed a line perpendicular to this tangent intersecting the inferior border of mental foramen, along which the mandibular cortical width was measured with a caliper. The mean cortical width on both sides of the mandible was used in this study. The coefcient of variation due to positioning and operator errors in cortical width measurements was less than 2%. Intraobserver variation in cortical width measurements was 0.1 mm, which was similar to the interobserver variation. The OST index was obtained from the integer of 0.2 times the patients weight minus the integer of 0.2 times her age on the basis of a previous study [7]. Mandibular cortical shape on dental panoramic radiographs was determined by observing the mandible distally from the mental foramen bilater-

Fig. 1.Dental panoramic radiograph of 60-year-old woman with normal cortex. We drew line parallel to long axis of mandible and tangential to inferior border of mandible and constructed dotted line perpendicular to this tangent intersecting inferior border of mental foramen, along which mandibular cortical width was measured. Distance between two parallel solid lines is cortical width. Arrow shows mental foramen.

Fig. 2.57-year-old woman with normal left mandibular inferior cortex detected on dental panoramic radiograph.

1756

AJR:183, December 2004

Dental Panoramic Radiography for Spinal Osteoporosis

ally and categorized into one of three groups according to the method of Klemetti et al. [9] as follows: normal cortex, the endosteal margin of the cortex is even and sharp on both sides (Fig. 2); mildly to moderately eroded cortex, the endosteal margin shows semilunar defects (lacunar resorption) or appears to form endosteal cortical residues (Fig. 3); or severely eroded cortex, the cortical layer forms heavy endosteal cortical residues and is clearly porous (Fig. 4). Overall agreements for intraobserver and interobserver performances were 92% and 82%, respectively. Characteristics between group A and group B were compared using the unpaired Students t test or the chi-square test. For mandibular cortical shape, subjects were divided into two groups on the basis of the assessment of panoramic radiographs: women with and without an eroded mandibular cortex (eroded cortex). The subjects were also divided into two groups on the basis of BMD measurements of the lumbar spine: women with and without spinal osteoporosis (T score < 2.5). The sensitivities, specicities, predictive values, and accuracies for the identication of women with spinal osteoporosis by eroded mandibular cortex were calculated in dichotomous 2 2 tables in group A and group B. The likelihood ratio and 95% condence interval (CI) for a positive risk-index result were also calculated as described by Simel et al. [22]. The likelihood ratio farther away from 1.0 indicates better performance for discriminating between subjects with and without the condition. Receiver operating characteristic (ROC) curve analyses were used to determine the optimal cutoff thresholds of cortical width and OST index for the identication of spinal osteoporosis in group A. The risk-index range corresponding to a sensitivity of approximately 90% was chosen to dene the low-risk group. The sensitivity, specicity, predictive value, accuracy, and likelihood ratio for the identication of women with spinal osteoporosis by optimal cutoff thresholds of cortical width and the OST index determined on ROC analyses in group A were calculated in group A and group B. In the subgroup that excluded subjects with spinal osteoporosis, the likelihood ratios for the identication of women with spinal osteopenia (T score, 1.0 to 2.5) by cortical width, OST index, and cortical shape were also calculated in group A and group B. Data analyses were performed by using SPSS, version 8.0 (Statistical Package for the Social Sciences) for Windows (Microsoft). A p value of less than 0.05 was considered statistically signicant.

Fig. 3.58-year-old woman with mildly to moderately eroded left mandibular inferior cortex detected on dental panoramic radiograph.

Fig. 4.59-year-old woman with severely eroded left mandibular inferior cortex detected on dental panoramic radiograph.

in group A was 0.771 (95% CI, 0.6780.864) for cortical width and 0.829 (95% CI, 0.748 0.910) for the OST index. No signicant difference in Az was seen between them. When the risk-index range corresponding to a sen-

sitivity of approximately 90% was chosen to determine the optimal cutoff threshold, the cutoff thresholds for cortical width and the OST index were 4.5 mm and 1.0, respectively. Tables 2 and 3 show the number of

TABLE 1

Characteristics of Study Subjects Postmenopausal Women Without and With Hysterectomy, Oophorectomy, or Estrogen Use Without (Group A) With (Group B) 157 56.3 8.0 153.4 5.7 52.0 7.2 10.8 7.6 4.0 1.0 47.1 25.5

Characteristic

Results

Characteristics of group A and group B are shown in Table 1. There was no signicant difference in characteristics between two groups. The area under the ROC curve (Az) for identifying women with spinal osteoporosis

No. of patients Age (yr) Height (cm) Weight (kg) Time since menopause (yr) Mandibular cortical width (mm) Mandibular cortical erosion (%) Spinal osteoporosis (%)
Note.For rows 26, data are mean SD.

159 57.4 7.5 152.8 5.3 51.0 7.4 8.3 7.4 4.0 1.0 48.4 23.9

AJR:183, December 2004

1757

Taguchi et al.
Spine Bone Mineral Density in Postmenopausal Women Without Hysterectomy, Oophorectomy, or Estrogen Use (Group A) Spine Bone Mineral Density Variable Normal Cortical width 4.5 mm > 4.5 mm Osteoporosis self-assessment tool index 1.0 > 1.0 Cortical shape Erosion Normal
Note.Data are numbers of patients.

TABLE 2

Osteopenia 47 17 32 32 31 33

Osteoporosis 34 4 33 5 33 5

33 24 19 38 13 44

teoporosis was 1.35 (95% CI, 1.141.60) for cortical width, 2.06 (95% CI, 1.622.63) for the OST index, and 2.39 (95% CI, 1.83 3.12) for cortical shape in group A; and 1.42 (95% CI, 1.211.67) for cortical width, 1.73 (95% CI, 1.302.30) for the OST index, and 2.22 (95% CI, 1.672.97) for cortical shape in group B. Furthermore, the likelihood ratio for a positive risk-index result for identifying women with spinal osteopenia was 1.27 (95% CI, 0.971.65) for cortical width, 1.50 (95% CI, 0.962.33) for the OST index, and 2.12 (95% CI, 1.243.65) for cortical shape in the group A; and 1.24 (95% CI, 0.93 1.59) for cortical width, 1.85 (95% CI, 1.16 2.94) for the OST index, and 2.77 (95% CI, 1.544.97) for cortical shape in group B.

TABLE 3

Spine Bone Mineral Density in Postmenopausal Women with Hysterectomy, Oophorectomy, or Estrogen Use (Group B) Spine Bone Mineral Density Variable Normal Osteopenia 42 17 32 27 31 28 Osteoporosis 37 3 29 11 32 8

Discussion

Cortical width 4.5 mm > 4.5 mm Osteoporosis self-assessment tool index 1.0 > 1.0 Cortical shape Erosion Normal
Note.Data are numbers of patients.

34 24 17 41 11 47

subjects in each group with normal spine BMD, osteopenia, or osteoporosis in postmenopausal women with and without histories of hysterectomy, oophorectomy, or estrogen use as determined by cortical width ( 4.5 mm), the OST index ( 1.0), and cortical shape (any cortical erosion). Using the cutoff thresholds of 4.5 mm for cortical width and 1.0 for the OST index and selecting any cortical erosion, we found that the sensitivity and specicity, respectively, for identifying women with spinal osteoporosis were 89.5% and 33.9% for cortical width, 86.8% and 57.8% for the OST index, and 86.8% and 63.6% for cortical shape in group A (Table 4). The specicity and accuracy for cortical width were signicantly lower than those for the OST index and cortical shape in this group. No signicant difference in the diagnostic performance

was found between the OST index and cortical shape in group A. The sensitivity and specicity, respectively, for identifying women with spinal osteoporosis were 92.5% and 35.0% for cortical width, 72.5% and 58.1% for the OST index, and 80.0% and 64.1% for cortical shape in group B (Table 5). The specicity for cortical width was signicantly lower than that for the OST index and cortical shape in this group. Accuracy for cortical width was signicantly lower than that for cortical shape. No signicant differences were seen in the diagnostic performances between the OST index and cortical shape in group B and for cortical width, the OST index, and cortical shape between group A and group B. The likelihood ratio for positive risk-index results in identifying women with spinal os-

In this study, no signicant difference in the diagnostic performance was found between mandibular cortical shape detected on dental panoramic radiographs and the OST index for identifying women with spinal osteoporosis. This tendency was similar to that observed for identifying women with spinal osteopenia in the subgroup of normal plus osteopenic women. These results suggest that postmenopausal women with spinal osteoporosis who should undergo BMD testing may be identied by mandibular cortical shape and by the OST index. The response rate for questionnaires may be relatively low if postmenopausal women have little information or no interest regarding osteoporosis; however, because dental panoramic radiographs are taken for the diagnosis of conditions affecting the teeth and jaws in clinical dental practice worldwide, the incidental cortical shape nding detected on dental panoramic radiographs may be an additional tool for the identication of postmenopausal women with spinal osteoporosis who cannot be detected with simple screening tools based on questionnaires such as the OST index. The sensitivity and accuracy for mandibular cortical shape and the OST index tended to be lower in postmenopausal women with histories of hysterectomy, oophorectomy, or estrogen use than in those without these histories, although the differences were relatively small for cortical shape compared with the OST index. The OST index was rst developed to identify postmenopausal women with low hip BMD dened by BMD T scores equal to 2.5 or less in the Asian female popu-

1758

AJR:183, December 2004

Dental Panoramic Radiography for Spinal Osteoporosis

Performance of Cortical Width, Osteoporosis Self-Assessment Tool (OST) Index, and Cortical Shape for Identifying Spinal Osteoporosis in Postmenopausal Women Without Hysterectomy, Oophorectomy, or Estrogen Use (Group A) Sensitivity Identifying Factor % Cortical width OST index Cortical shape
Note.CI = confidence interval.

TABLE 4

Specificity % 33.9 57.8 63.6 95% CI 25.542.3 49.166.6 55.172.2

Positive Predictive Value Negative Predictive Value % 29.8 39.2 42.9 95% CI 21.438.1 28.849.6 31.854.0 % 91.1 93.3 93.9 95% CI 82.899.4 87.699.0 88.799.1 %

Accuracy 95% CI 39.354.9 57.472.2 62.076.4

95% CI 79.799.2 76.197.6 76.197.6

89.5 86.8 86.8

47.1 64.8 69.2

lation without oophorectomy and estrogen use [7]. However, Geusens et al. [17] showed that the diagnostic performance of the OST index was similar to that of simple calculated osteoporosis risk estimation and the osteoporosis risk assessment instrument; estrogen use is included in the calculation for both of these tools. The presence of subjects with a history of hysterectomy and oophorectomy may have more inuence on the diagnostic performance of a simple screening tool based on questionnaires such as the OST index than that of cortical shape directly detected on radiographs. The Az for identifying women with spinal osteoporosis by cortical width was 0.771 (95% CI, 0.6780.864), corresponding to moderate accuracy [22]. This result was similar to that of a recent report in which the Az for identifying women with reduced BMD at the spine, hip, or forearm by means of cortical width detected on dental panoramic radiographs was 0.733 (95% CI, 0.6180.830) in 74 Caucasian women [11]. This nding implies that mandibular cortical width detected on dental panoramic radiographs may be used worldwide as well, although the diagnostic performance of cortical width was lower than that of cortical shape and the OST index in our study. Devlin and Horner [11] suggested that a diagnostic threshold for cortical width of 3 mm (or less) may be the most appropriate threshold for referral for bone densitometry in Caucasian postmenopausal women. This

threshold may contribute to the high specicity and high positive predictive value; however, the low sensitivity may lead to a large number of postmenopausal women with undetectable spinal osteoporosis or osteopenia. In fact, if we selected the cortical width of 3 mm (or less) as a threshold in this study, 21 (55%) subjects with spinal osteoporosis in group A and 24 (60%) subjects with spinal osteoporosis in group B could not be identied by cortical width. Because the risk-index range corresponding to a sensitivity of approximately 90% was chosen to determine the optimal cutoff threshold for cortical width in this study, the positive predictive value was rather low. This nding resulted in the high false-positive rate ( 50%) that would lead to unnecessary additional examinations and would decrease the cost-effectiveness. However, in comparison with a general screening, this approach is efcient because it identies a subgroup (postmenopausal women with spinal osteoporosis). In comparison with cortical width, the false-positive rate was 32% for the OST index and 27% for cortical shape in the group A. The false-positive rate was 31% and 28%, respectively, for these measures in group B. These rates indicate that a large number of postmenopausal women with spinal osteoporosis may be identied with minimal additional cost by both the OST index and cortical shape detected on dental panoramic radiographs. This study has some design limitations. Subjects without histories of hysterectomy,

oophorectomy, or estrogen use may not be representative of all Japanese postmenopausal women. Characteristics such as the proportion of subjects with a history of hysterectomy and oophorectomy or the duration of their estrogen use might limit our interpretation. Further investigation in a large number of populations would be necessary to conrm our ndings. Another limitation is that we evaluated the diagnostic performance in identifying postmenopausal women with spinal osteoporosis in this study, although other simple screening tools based on questionnaires were evaluated for the identication of postmenopausal women with low hip BMD or femoral osteoporosis. However, because the prevalence rate of vertebral fractures is higher in Japanese women than in Caucasian women and is higher than the prevalence rate of hip fractures in Japanese women [24, 25], we chose the lumbar spine as a screening site in this study. We previously showed that mandibular cortical BMD measured by quantitative CT (QCT) was signicantly correlated with lumbar spine BMD measured on QCT in postmenopausal women [26]. This nding implies that the measurement of mandibular cortical bone may be useful for predicting spine BMD. Although we did not refer to the identication of postmenopausal women with low hip BMD or femoral osteoporosis in this study, we recently showed that the diagnostic performance in identifying postmenopausal women with low spine BMD was almost similar to that

TABLE 5

Performance of Cortical Width, Osteoporosis Self-Assessment Tool (OST) Index, and Cortical Shape in Identifying Spinal Osteoporosis in Postmenopausal Women with Hysterectomy, Oophorectomy, or Estrogen Use (Group B) Sensitivity Specificity % 35.0 58.1 64.1 95% CI 26.443.7 49.267.1 55.472.8 Positive Predictive Value Negative Predictive Value % 32.7 37.2 43.2 95% CI 24.141.4 26.547.9 32.054.5 % 93.2 86.0 90.4 95% CI 85.7100.6 78.393.7 84.196.7 % 49.7 61.8 68.2 Accuracy 95% CI 41.957.5 54.169.4 60.975.5

Identifying Factor % Cortical width OST index Cortical shape

Note.CI = confidence interval.

95% CI 84.3100.7 58.786.3 67.692.4

92.5 72.5 80.0

AJR:183, December 2004

1759

Taguchi et al. in identifying those with low hip BMD when general dental practitioners assessed mandibular cortical shape on dental panoramic radiographs [13]. This nding implies that there may be no signicant difference in the diagnostic performance between mandibular cortical shape and the OST index even when BMD of the femoral neck is used as a standard. In conclusion, the diagnostic performance of mandibular inferior cortical shape detected on dental panoramic radiographs for identifying postmenopausal women with spinal osteoporosis was similar to that of the OST index developed for the Asian female population. Cortical shape also may be useful for the identication of spinal osteoporosis in postmenopausal women with histories of hysterectomy, oophorectomy, or estrogen use. Dentists may be able to refer asymptomatic postmenopausal women with suspected spinal osteoporosis for bone densitometry using dental panoramic radiographs with diagnostic performance similar to that of other osteoporosis risk indexes based on questionnaires such as the OST index.
References
1. Melton LJ 3rd. Adverse outcomes of osteoporotic fractures in the general population. J Bone Miner Res 2003;18:11391141 2. Siris ES, Miller PD, Barrett-Connor E, et al. Identication and fracture outcomes of undiagnosed low bone mineral density in postmenopausal women: results from the National Osteoporosis Risk Assessment. JAMA 2001;286:28152822 3. Michalsson K, Bergstrm R, Mallmin H, Holmberg L, Wolk A, Ljunghall S. Screening for osteopenia and osteoporosis: selection by body composition. Osteoporos Int 1996;6:120126 4. Lydick E, Cook K, Turpin J, Melton M, Stine R, Byrnes C. Development and validation of a simple 12. questionnaire to facilitate identication of women likely to have low bone density. Am J Manag Care 1998;4:3748 Weinstein L, Ullery B. Identication of at-risk women for osteoporosis screening. Am J Obstet Gynecol 2000;183:547549 Cadarette SM, Jaglal SB, Kreiger N, Mclsaac WJ, Darlington GA, Tu JV. Development and validation of the Osteoporosis Risk Assessment Instrument to facilitate selection of women for bone densitometry. CMAJ 2000;162:12891294 Koh LK, Sedrine WB, Torralba TP, et al. A simple tool to identify Asian women at increased risk of osteoporosis. Osteoporos Int 2001;12:699705 Sedrine WB, Chevallier T, Zegels B, et al. Development and assessment of the Osteoporosis Index of Risk (OSIRIS) to facilitate selection of women for bone densitometry. Gynecol Endocrinol 2002;16:245250 Klemetti E, Kolmakov S, Kroger H. Pantomography in assessment of the osteoporosis risk group. Scand J Dent Res 1994;102:6872 Taguchi A, Suei Y, Ohtsuka M, Otani K, Tanimoto K, Ohtaki M. Usefulness of panoramic radiography in the diagnosis of postmenopausal osteoporosis in women: width and morphology of inferior cortex of the mandible. Dentomaxillofac Radiol 1996;25:263267 Devlin H, Horner K. Mandibular radiomorphometric indices in the diagnosis of reduced skeletal bone mineral density. Osteoporos Int 2002;13:373378 Horner K, Devlin H, Harvey L. Detecting patients with low skeletal bone mass. J Dent 2002;30:171175 Nakamoto T, Taguchi A, Ohtsuka M, et al. Dental panoramic radiograph as a tool to detect postmenopausal women with low bone mineral density: untrained general dental practitioners diagnostic performance. Osteoporos Int 2003;14:659664 Taguchi A, Suei Y, Sanada M, et al. Detection of vascular disease risk in women by panoramic radiography. J Dent Res 2003;82:838843 Taguchi A, Sanada M, Krall E, et al. Relationship between dental panoramic radiographic ndings and biochemical markers of bone turnover. J Bone Miner Res 2003;18:16891694 16. Bollen AM, Taguchi A, Hujoel PP, Hollender LG. Case-control study on self-reported osteoporotic fractures and mandibular cortical bone. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;90:518524 17. Geusens P, Hochberg MC, van der Voort DJ, et al. Performance of risk indices for identifying low bone density in postmenopausal women. Mayo Clin Proc 2002;77:629637 18. Ohta H, Makita K, Komukai S, Nozawa S. Bone resorption versus estrogen loss following oophorectomy and menopause. Maturitas 2002;43:2733 19. Cheng S, Sievanen H, Heinonen A, et al. Does hysterectomy with ovarian conservation affect bone metabolism and density? J Bone Miner Metab 2003;21:1216 20. Genant HK, Wu CY, van Kuijk C, Nevitt MC. Vertebral fracture assessment using a semiquantitative technique. J Bone Miner Res 1993;8:11371148 21. Kanis JA, WHO Study Group. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report. Osteoporos Int 1994;4:368381 22. Simel DL, Samsa GP, Matchar DB. Likelihood ratios with condence: sample size estimation for diagnostic test studies. J Clin Epidemiol 1991;44:763770 23. Greiner M, Pfeiffer D, Smith RD. Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests. Prev Vet Med 2000;45:2341 24. Ross PD, Fujiwara S, Huang C, et al. Vertebral fracture prevalence in women in Hiroshima compared to Caucasians or Japanese in the US. Int J Epidemiol 1995;24:11711177 25. Iki M, Kagamimori S, Kagawa Y, Matsuzaki T, Yoneshima H, Marumo F. Bone mineral density of the spine, hip and distal forearm in representative samples of the Japanese female population: Japanese Population-Based Osteoporosis (JPOS) study. Osteoporos Int 2001;12:529537 26. Taguchi A, Tanimoto K, Suei Y, Ohama K, Wada T. Relationship between the mandibular and lumbar vertebral bone mineral density at different postmenopausal stages. Dentomaxillofac Radiol 1996;25:130135

5.

6.

7.

8.

9.

10.

11.

13.

14.

15.

1760

AJR:183, December 2004