Calcif Tissue Int (1992) 51:352-355
Calcified Tissue International 9 1992 Springer-Verlag New York Inc.
Comparative Assessment of Bone Mineral Density of the Forearm Using Single Photon and Dual X-Ray Absorptiometry Jeri W. Nieves, Felicia Cosman, Chris Mars, and Robert Lindsay Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York 10993, USA Received January 31, 1992, and in revised form May 11, 1992
Summary: Forearm bone mineral density (BMD) was measured at proximal and distal sites by 1251 single photon absorptiometry (SPA) and by dual energy X-ray absorptiometry (DXA) in 67 consecutive subjects, aged 18-75 years. Correlations and regression equations b e t w e e n these two techniques were determined. All forearm measurements were significantly correlated with each other (r = 0.5990.926; P ~< 0.0001). Although SPA and DXA correct for fat in different ways, we found similar correlation and regression equations in women with body mass index measurements above and below the mean. In addition, forearm measurements by both techniques were moderately correlated with vertebral spine and hip BMD. We conclude that overall, SPA forearm measurements in a population can be calibrated to DXA measurements if necessary, and that DXA forearm measurements are as predictive of the remainder of the skeleton as SPA measurements.
ever, where any radial site could be chosen for analysis, other systems such as our Nuclear Data SPA measure fixed locations, which are anatomically different from those sites measured by DXA. Therefore, it is more difficult to make direct comparisons between these two machines. Although BMD assessment can be used to predict fracture risk [7-11], which site is the most predictive remains unknown. It is therefore important to determine how DXA forearm BMD measurements correlate with lumbar spine and hip BMD, and to assess which of the available technologies for forearm measurement gives results that correlate best with these other sites. To this end, we have compared BMD measurements of the forearm using both SPA and DXA, and, in addition, evaluated the relationships between these peripheral measurements and measurements of the axial skeleton (lumbar spine and femoral neck) performed predominantly by DXA.
Key Words: Bone densitometry - Bone mineral density Forearm - Single photon absorptiometry - Dual X-ray absorptiometry.
Methods Patients
Measurement of the forearm (radius or a combination of radius and ulna) is often used as a quick, precise, and relatively inexpensive method to assess the bone mineral density (BMD) of an individual. For the past two decades, most forearm density measurements have been determined using single photon absorptiometry (SPA). Recent software developments, however, have made it possible to measure BMD at forearm sites using dual X-ray absorptiometry (DXA). DXA has potential advantages as the technology is even quicker, exposes the patient to less radiation, has good image resolution and reproducibility, and is therefore likely to be more precise [1-5]. In addition, DXA covers a larger portion of the forearm, including sections of bone with more widely varying compositions of cancellous and cortical bone than is normally accomplished by SPA [6]. Large data sets of individuals with SPA measurements exist, and in order to equate these values to DXA measurements it is necessary to know the extent to which the results of these two instruments correlate with each other. Weinstein et al. [ 1] recently published a comparison between the Norland 2780 SPA and the DXA proximal (diaphyseal) radius measurements, and found high correlations between DXA and SPA measurements (r = 0.975, P ~ 0.0001). Unlike the Norland, how-
Offprint requests to: R. Lindsay
Sixty-seven consecutive patients had measurements of the nondominant forearm (radius and ulna) performed on both the SPA (Nuclear Data 1100A, Schaumburg, IL) and on DXA (Hologic QDR-1000, Waltham, MA), as well as vertebral spine and hip BMD assessments (on the same day) during July and August 1991. The subjects comprised both patients with osteoporosis as well as those coming for bone mass determinations to be used in assessing their risk of developing osteoporosis. No patients were excluded and all gave informed consent. The population of females was separated into two groups on the basis of their body mass index (above or below the mean), and all comparisons between the SPA and DXA were repeated in these subgroups.
SPA Technique The starting position for the SPA scan, using the standard Nuclear Data protocol, is a selected position on the forearm with an 8-ram interosseous space between the ulna and radius. A series of computer-controlled rectilinear scans are then used to measure both proximally and distally from that point. The proximal measurement involves six scans of the radius and ulna each separated by 4 mm, starting with the 8-ram interosseous space (inclusive). The distal mode collects four passes of the radius and ulna scans each separated by 2 mm distal of the starting interosseous space (exclusive). The SPA is standardized to an aluminum standard, and corrections are made for fat using a specialized algorithm [12]. Forearm measures on the SPA take 13 minutes and the radiation exposure is less than 15 mrem [10]. Reproducibility of the SPA forearm measurements was determined on the same six subjects over a period of 4
J. W. Nieves et al.: DXA vs SPA of the Forearm
353 Table 1. Descriptive characteristics of 67 subjects and bone mineral density at various sites Mean -+ sem Age (years) Height (cm) Weight (kg) Body mass index (kg/m2) Bone mineral density (g/em2) Lumbar spine Hip Femoral neck Trochanteric Wards triangle Forearm (DXA) Ultradistal Mid-distal Proximal third Forearm (SPA) Distal Proximal
Fig. 1. (A) Forearm region scanned by the single photon absorptiometry (Nuclear Data 1100A). (B) Forearm region scanned by dual energy X-ray absorptiometry (Hologic QDR-1000).
weeks, and the resulting coefficients of variation were 1.4% for distal and 1.3% for proximal measurements.
D X A Technique DXA wrist measurements were performed using the Hologic Inc. Model #1000 (software version 5.11 with a 2.3 mm source collimator). With the patient in the sitting position, the nondominant forearm is placed on the scanning table adjacent to a foam rubber spacer used to facilitate positioning. The proximal one-third region is arbitrarily defined as a region 20 mm long centered at a distance equal to one-third of the forearm length measured from the distal tip of the ulna and contains almost entirely cortical bone [6]. The ultradistal site is defined as a region nominally 15 mm in length positioned just proximal to the end plate of the radius (excluding the end plate itself) and contains a greater proportion of cancellous bone [6]. The middistal site is a region between the one-third and ultradistal regions and contains both trabecular and cortical bone. Forearm measurements on DXA take approximately 10 minutes and the radiation exposure is about 1 mrem [2]. Reproducibility was determined using multiple forearm measures on the same 12 subjects over a period of 4 weeks. The resulting CV% was 1.6 for the ultradistal region, 0.6 for the mid-distal region, and 0.8 for the ultradistal region, 0.6 for the mid-distal region, and 0.8 for the proximal third. These values for forearm CV% are in agreement with other reported in vivo CV% values ranging from 0.65% to 1.2% [1, 13]. The higher CV% for the ultradistal region, as compared with the other forearm regions, could be because the ultradistal region requires manual positioning of the marker to exclude the end plate of the radius, whereas the other two forearm regions are mechanically positioned. The DXA scan encompasses an area of the forearm with a total length equal to 88 mm whereas the SPA only scans 32 mm (Fig. 1). The SPA distal site is equivalent to about one-half of the DXA ultradistal site. The SPA proximal site contains the rest of the ultradistal site and the adjacent portion of the mid-distal site. The DXA measures a larger portion of the forearm than the SPA, and in fact the DXA proximal third region has no anatomic counterpart with the SPA.
Nonwrist B M D Measurements BMD of the lumbar spine and hip were made using the Hologic QDR
51.5 163.2 65.2 24.4
-+ 1.75 -+ 0.82 -+ 1.56 --- 0.54
Range 18-75 149-183 43-109 18-40
0.934 + 0.018
0.498-1.254
0.731 -+ 0.017 0.610 -~ 0.015 0.602 -+ 0.019
0.303-1.100 0.220-0.894 0.216-0.999
0.395 -+ 0.009 0.533 -+ 0.009 0.618 -+ 0.011
0.175-0.617 0.257-0.718 0.330-0.812
0.309 -+ 0.008 0.445 -+ 0.010
0.160-0.490 0.210-0.670
(n = 44), Norland XR-26 (Norland, Fort Atkinson, WI, n = 7), or Lunar 042 DPA (Lunar, Madison, WI, n = 16). DPA measurements were converted to DXA measurements using regression equations developed from our database of 31 patients measured on both the DPA and DXA [14]. Correlations resulting from this regression conversion are good---0.96 for lumbar spine and 0.90 for femoral neck. In addition, measurements of the seven patients on the XR-26 were converted to DXA-QDR values using a regression equation based on 32 subjects measured on both instruments (unpublished data). Resuiting correlations are high: 0.99 for lumbar spine and 0.95 for the femoral neck.
Analysis Data were analyzed using software from SAS computer systems to determine Pearson correlation coefficients between the two forearm BMD measurements and between forearm BMD and vertebral and hip BMD measurements. Linear regression analysis by the method of least squares estimates was performed to allow derivation of an equation converting SPA to DXA measurements.
Results T h e d e s c r i p t i v e c h a r a c t e r i s t i c s of t h e 67 s u b j e c t s are listed in T a b l e 1. T h e r e w e r e 61 f e m a l e s (24 p r e m e n o p a u s a l a n d 37 p o s t m e n o p a u s a l ) , a n d 6 m a l e s . M e a n age w a s 51 y e a r s (range 18-75). B o n e m i n e r a l m e a s u r e m e n t s for e a c h o f t h e skeletal sites are also d i s p l a y e d in T a b l e 1. O s t e o p o r o t i c w o m e n w e r e identified as h a v i n g a v e r t e b r a l b o n e m i n e r a l d e n s i t y 2 SD b e l o w y o u n g n o r m a l ( < 0 . 8 6 g/cm2). O n t h e b a s i s of this definition, 22 w o m e n w e r e o s t e o p o r o t i c . T h e m e a n B M D o f the w r i s t sites m e a s u r e d b y D X A w e r e u l t r a d i s t a l 0.4 g/cm 2, mid-distal 0.5 g / c m 2, a n d p r o x i m a l t h i r d 0.6 g/cm 2, a n d in the s a m e p o p u l a t i o n m e a n B M D for the S P A distal site w a s 0.3 g/cm 2 a n d t h e S P A p r o x i m a l site w a s 0.4 g/cm 2. C o r r e l a t i o n coefficients b e t w e e n t h e t w o p r o c e d u r e s for m e a s u r i n g B M D o f the wrist are s h o w n in T a b l e 2. All measures w e r e significantly c o r r e l a t e d , w i t h c o r r e l a t i o n coefficients v a r y i n g f r o m 0.599 to 0.926 ( P ~ 0.0001). P e r h a p s n o t surprisingly, the l o w e s t c o r r e l a t i o n o f 0.599 w a s b e t w e e n t h e S P A distal m e a s u r e a n d t h e m o s t p r o x i m a l m e a s u r e o f t h e D X A for w h i c h t h e r e is n o direct S P A c o m p a r i s o n . T h e S P A p r o x i m a l m e a s u r e m e n t c o r r e l a t e s b e s t w i t h t h e D X A mid-
354
J. W. Nieves et al.: DXA vs SPA of the Forearm
Table 2. Correlations among forearm bone mineral density measurements using SPA and DXA
Table 3. Correlations between forearm bone mineral density and lumbar spine and hip bone mineral density
SPAr DXA
Proximal
Distal
Ultradistal Mid-distal Proximal third
0.777 0.926a 0.813
0.661b 0.698 0.599
a Regression equation for conversion between these sites is: Middistal (DXA) = 0.8772 SPA proximal + 0.1425 u Regression equation for conversion between these sites is: Ultradistal (DXA) = 0.7489 SPA distal + 0.1637 c p ~
Calcified Tissue International 9 1992 Springer-Verlag New York Inc.
Comparative Assessment of Bone Mineral Density of the Forearm Using Single Photon and Dual X-Ray Absorptiometry Jeri W. Nieves, Felicia Cosman, Chris Mars, and Robert Lindsay Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York 10993, USA Received January 31, 1992, and in revised form May 11, 1992
Summary: Forearm bone mineral density (BMD) was measured at proximal and distal sites by 1251 single photon absorptiometry (SPA) and by dual energy X-ray absorptiometry (DXA) in 67 consecutive subjects, aged 18-75 years. Correlations and regression equations b e t w e e n these two techniques were determined. All forearm measurements were significantly correlated with each other (r = 0.5990.926; P ~< 0.0001). Although SPA and DXA correct for fat in different ways, we found similar correlation and regression equations in women with body mass index measurements above and below the mean. In addition, forearm measurements by both techniques were moderately correlated with vertebral spine and hip BMD. We conclude that overall, SPA forearm measurements in a population can be calibrated to DXA measurements if necessary, and that DXA forearm measurements are as predictive of the remainder of the skeleton as SPA measurements.
ever, where any radial site could be chosen for analysis, other systems such as our Nuclear Data SPA measure fixed locations, which are anatomically different from those sites measured by DXA. Therefore, it is more difficult to make direct comparisons between these two machines. Although BMD assessment can be used to predict fracture risk [7-11], which site is the most predictive remains unknown. It is therefore important to determine how DXA forearm BMD measurements correlate with lumbar spine and hip BMD, and to assess which of the available technologies for forearm measurement gives results that correlate best with these other sites. To this end, we have compared BMD measurements of the forearm using both SPA and DXA, and, in addition, evaluated the relationships between these peripheral measurements and measurements of the axial skeleton (lumbar spine and femoral neck) performed predominantly by DXA.
Key Words: Bone densitometry - Bone mineral density Forearm - Single photon absorptiometry - Dual X-ray absorptiometry.
Methods Patients
Measurement of the forearm (radius or a combination of radius and ulna) is often used as a quick, precise, and relatively inexpensive method to assess the bone mineral density (BMD) of an individual. For the past two decades, most forearm density measurements have been determined using single photon absorptiometry (SPA). Recent software developments, however, have made it possible to measure BMD at forearm sites using dual X-ray absorptiometry (DXA). DXA has potential advantages as the technology is even quicker, exposes the patient to less radiation, has good image resolution and reproducibility, and is therefore likely to be more precise [1-5]. In addition, DXA covers a larger portion of the forearm, including sections of bone with more widely varying compositions of cancellous and cortical bone than is normally accomplished by SPA [6]. Large data sets of individuals with SPA measurements exist, and in order to equate these values to DXA measurements it is necessary to know the extent to which the results of these two instruments correlate with each other. Weinstein et al. [ 1] recently published a comparison between the Norland 2780 SPA and the DXA proximal (diaphyseal) radius measurements, and found high correlations between DXA and SPA measurements (r = 0.975, P ~ 0.0001). Unlike the Norland, how-
Offprint requests to: R. Lindsay
Sixty-seven consecutive patients had measurements of the nondominant forearm (radius and ulna) performed on both the SPA (Nuclear Data 1100A, Schaumburg, IL) and on DXA (Hologic QDR-1000, Waltham, MA), as well as vertebral spine and hip BMD assessments (on the same day) during July and August 1991. The subjects comprised both patients with osteoporosis as well as those coming for bone mass determinations to be used in assessing their risk of developing osteoporosis. No patients were excluded and all gave informed consent. The population of females was separated into two groups on the basis of their body mass index (above or below the mean), and all comparisons between the SPA and DXA were repeated in these subgroups.
SPA Technique The starting position for the SPA scan, using the standard Nuclear Data protocol, is a selected position on the forearm with an 8-ram interosseous space between the ulna and radius. A series of computer-controlled rectilinear scans are then used to measure both proximally and distally from that point. The proximal measurement involves six scans of the radius and ulna each separated by 4 mm, starting with the 8-ram interosseous space (inclusive). The distal mode collects four passes of the radius and ulna scans each separated by 2 mm distal of the starting interosseous space (exclusive). The SPA is standardized to an aluminum standard, and corrections are made for fat using a specialized algorithm [12]. Forearm measures on the SPA take 13 minutes and the radiation exposure is less than 15 mrem [10]. Reproducibility of the SPA forearm measurements was determined on the same six subjects over a period of 4
J. W. Nieves et al.: DXA vs SPA of the Forearm
353 Table 1. Descriptive characteristics of 67 subjects and bone mineral density at various sites Mean -+ sem Age (years) Height (cm) Weight (kg) Body mass index (kg/m2) Bone mineral density (g/em2) Lumbar spine Hip Femoral neck Trochanteric Wards triangle Forearm (DXA) Ultradistal Mid-distal Proximal third Forearm (SPA) Distal Proximal
Fig. 1. (A) Forearm region scanned by the single photon absorptiometry (Nuclear Data 1100A). (B) Forearm region scanned by dual energy X-ray absorptiometry (Hologic QDR-1000).
weeks, and the resulting coefficients of variation were 1.4% for distal and 1.3% for proximal measurements.
D X A Technique DXA wrist measurements were performed using the Hologic Inc. Model #1000 (software version 5.11 with a 2.3 mm source collimator). With the patient in the sitting position, the nondominant forearm is placed on the scanning table adjacent to a foam rubber spacer used to facilitate positioning. The proximal one-third region is arbitrarily defined as a region 20 mm long centered at a distance equal to one-third of the forearm length measured from the distal tip of the ulna and contains almost entirely cortical bone [6]. The ultradistal site is defined as a region nominally 15 mm in length positioned just proximal to the end plate of the radius (excluding the end plate itself) and contains a greater proportion of cancellous bone [6]. The middistal site is a region between the one-third and ultradistal regions and contains both trabecular and cortical bone. Forearm measurements on DXA take approximately 10 minutes and the radiation exposure is about 1 mrem [2]. Reproducibility was determined using multiple forearm measures on the same 12 subjects over a period of 4 weeks. The resulting CV% was 1.6 for the ultradistal region, 0.6 for the mid-distal region, and 0.8 for the ultradistal region, 0.6 for the mid-distal region, and 0.8 for the proximal third. These values for forearm CV% are in agreement with other reported in vivo CV% values ranging from 0.65% to 1.2% [1, 13]. The higher CV% for the ultradistal region, as compared with the other forearm regions, could be because the ultradistal region requires manual positioning of the marker to exclude the end plate of the radius, whereas the other two forearm regions are mechanically positioned. The DXA scan encompasses an area of the forearm with a total length equal to 88 mm whereas the SPA only scans 32 mm (Fig. 1). The SPA distal site is equivalent to about one-half of the DXA ultradistal site. The SPA proximal site contains the rest of the ultradistal site and the adjacent portion of the mid-distal site. The DXA measures a larger portion of the forearm than the SPA, and in fact the DXA proximal third region has no anatomic counterpart with the SPA.
Nonwrist B M D Measurements BMD of the lumbar spine and hip were made using the Hologic QDR
51.5 163.2 65.2 24.4
-+ 1.75 -+ 0.82 -+ 1.56 --- 0.54
Range 18-75 149-183 43-109 18-40
0.934 + 0.018
0.498-1.254
0.731 -+ 0.017 0.610 -~ 0.015 0.602 -+ 0.019
0.303-1.100 0.220-0.894 0.216-0.999
0.395 -+ 0.009 0.533 -+ 0.009 0.618 -+ 0.011
0.175-0.617 0.257-0.718 0.330-0.812
0.309 -+ 0.008 0.445 -+ 0.010
0.160-0.490 0.210-0.670
(n = 44), Norland XR-26 (Norland, Fort Atkinson, WI, n = 7), or Lunar 042 DPA (Lunar, Madison, WI, n = 16). DPA measurements were converted to DXA measurements using regression equations developed from our database of 31 patients measured on both the DPA and DXA [14]. Correlations resulting from this regression conversion are good---0.96 for lumbar spine and 0.90 for femoral neck. In addition, measurements of the seven patients on the XR-26 were converted to DXA-QDR values using a regression equation based on 32 subjects measured on both instruments (unpublished data). Resuiting correlations are high: 0.99 for lumbar spine and 0.95 for the femoral neck.
Analysis Data were analyzed using software from SAS computer systems to determine Pearson correlation coefficients between the two forearm BMD measurements and between forearm BMD and vertebral and hip BMD measurements. Linear regression analysis by the method of least squares estimates was performed to allow derivation of an equation converting SPA to DXA measurements.
Results T h e d e s c r i p t i v e c h a r a c t e r i s t i c s of t h e 67 s u b j e c t s are listed in T a b l e 1. T h e r e w e r e 61 f e m a l e s (24 p r e m e n o p a u s a l a n d 37 p o s t m e n o p a u s a l ) , a n d 6 m a l e s . M e a n age w a s 51 y e a r s (range 18-75). B o n e m i n e r a l m e a s u r e m e n t s for e a c h o f t h e skeletal sites are also d i s p l a y e d in T a b l e 1. O s t e o p o r o t i c w o m e n w e r e identified as h a v i n g a v e r t e b r a l b o n e m i n e r a l d e n s i t y 2 SD b e l o w y o u n g n o r m a l ( < 0 . 8 6 g/cm2). O n t h e b a s i s of this definition, 22 w o m e n w e r e o s t e o p o r o t i c . T h e m e a n B M D o f the w r i s t sites m e a s u r e d b y D X A w e r e u l t r a d i s t a l 0.4 g/cm 2, mid-distal 0.5 g / c m 2, a n d p r o x i m a l t h i r d 0.6 g/cm 2, a n d in the s a m e p o p u l a t i o n m e a n B M D for the S P A distal site w a s 0.3 g/cm 2 a n d t h e S P A p r o x i m a l site w a s 0.4 g/cm 2. C o r r e l a t i o n coefficients b e t w e e n t h e t w o p r o c e d u r e s for m e a s u r i n g B M D o f the wrist are s h o w n in T a b l e 2. All measures w e r e significantly c o r r e l a t e d , w i t h c o r r e l a t i o n coefficients v a r y i n g f r o m 0.599 to 0.926 ( P ~ 0.0001). P e r h a p s n o t surprisingly, the l o w e s t c o r r e l a t i o n o f 0.599 w a s b e t w e e n t h e S P A distal m e a s u r e a n d t h e m o s t p r o x i m a l m e a s u r e o f t h e D X A for w h i c h t h e r e is n o direct S P A c o m p a r i s o n . T h e S P A p r o x i m a l m e a s u r e m e n t c o r r e l a t e s b e s t w i t h t h e D X A mid-
354
J. W. Nieves et al.: DXA vs SPA of the Forearm
Table 2. Correlations among forearm bone mineral density measurements using SPA and DXA
Table 3. Correlations between forearm bone mineral density and lumbar spine and hip bone mineral density
SPAr DXA
Proximal
Distal
Ultradistal Mid-distal Proximal third
0.777 0.926a 0.813
0.661b 0.698 0.599
a Regression equation for conversion between these sites is: Middistal (DXA) = 0.8772 SPA proximal + 0.1425 u Regression equation for conversion between these sites is: Ultradistal (DXA) = 0.7489 SPA distal + 0.1637 c p ~
