Importance of measurement of spongious vertebral bone mineral density in the assessment of osteoporosis

Bone, 8, 201-206 (1987) Printed in the USA. All rights reserved

Copyright

8756-3282187 $3.00 + .OO 0 1987 Pergamon Journals Ltd.

Importance of Measurement of Spongious Vertebral Bone Mineral Density in the Assessment of Osteoporosis C.D.

JONES,‘,*

Departments

A.-M.

of Rad/o/ogy,

Address correspondence Medicine, San Francwo,

LAVAL-JEANTET,’ lUnwers~ty of Pax,

M.H.

LAVAL-JEANTET,’

and H.K. GENANT2

and *Univers/ty of Cahfornla School of Medione.

and reprints:: Harry K. Genant, M D CA 94143.0628, USA

Department

Abstract

of RadIology,

San Francrsco,

C-309, Uwersity

Cahfornia, USA

of Callfornla

School of

especially in the verteDral bodies, since this is a region that is composed in part of spongious bone and provides a significant proportion of the body’s support (Weaver and Chalmers. 1966); it is spongious bone that responds to stress more rapidly and more extensively than compact bone (Mazess, 1979). In women mineral loss in vertebral spongious bone is the earliest modification of bone due to oophorectomy (Genant et al., 1982) or physiologic menopause (Johnston. 198313: Parfltt, 1984). Effective preventive strategies and treatment planning require sensitive detection of early. postmenopausal, bone mineral loss. Currently. two methods are being applied for the noninvaslve evaluation of mineral content in vertebrae: dual-photon absorptiometty (DPA) (Wilson, 1975; Krolner and Nielsen. 1982; Dunn et al., 1980) and quantitative computed tomography (QCT) (Genarlt and Boyd, 1977: Rohloff et al 1982: Laval-Jeantet et al.. 1979). QCT can provide a bone mineral measurement of the total vertebra and of the spongious portions of the anterior vertebral bone and transverse processes whereas DPA can provide a bone mineral measurement of only the totai vertebral body and adjacent mlneral deposits (Genant et al.. 1984) QCT and DPA have approximately the same longitudinal precision and both are becoming widely available It has been suggested (Mazess, 1979) that DPA is sufficient for early diagnosis of vertebral bone mineral loss. However because the spongious portions of ilertebral bone show the earllest loss in bone mineral. QCT might be a better diagnostic tool In evaluating the initial stages of osteoporosis. In addition QCT is faster to perform and does not require dedicated equipment. whereas DPA gives a lower radiation dose. The present work uses the unique capability of QCT to spatially separate trabecular from more compact bone in the spine. Using QCT measurements of bone mineral denslty. we compared the amounts of mineral density and content In spongious vertebral bone with that In more compact vertebral bone with the goal of determining what differences might be demonstrated between normal and osteoporotic women and men

Quantitative computed tomography (QCT) measurements of bone mineral density, volume, and content were made of the spongious and compact portions of a vertebral body and of the total vertebral body in 19 osteoporotic patients (13 women and 6 men) and 20 normal (control) age- and sex-matched patients. Specifically, CT scans of 1Omm thick transverse sections of L2 or L3 of each osteoporotic and control patient were subjected to QCT bone measurement with the use of K,HPO, calibration for converting CT values into units (mg/ml) equivalent to bone mineral density. In every instance, a significant difference was found in the spongious vertebral bone mineral density (and content) between the osteoporotic and the paired normal groups. Even when the total vertebral bone mineral content was normal, the spongious vertebral bone mineral density was significantly less in the osteoporotic patients. Only in the osteoporotic men was a difference in compact bone density (and content) found; the osteoporotic men had a significantly lower compact bone mineral than did the normal male patients. This study shows that QCT can be applied to measurement of bone mineral density (and content) of both the anterior spongious and the more dense compact vertebral bone from a single section of a lumbar vertebra and that it is the anterior, spongious, vertebral bone rather than the compact vertebral bone that exhibits the more severe decrease in bone mineral density (and content) due to osteoporosis. Key Words: Osteoporosis-Bone, Trabecular (Spongious), Compact-Quantitative Computed TomographyDual-Photon Absorptiometry

Introduction Bone decreases throughout life as a normal part of aging. Although both trabecular (spongious) and cortical (compact) bone eventually are lost. the earliest ioss occurs primarily In sponglous bone (Johnston 1983a) Osteoporosis, the end result of normal and abnormal bone loss. can be manifest by various atraumatic skeletal fractures (Gordan and Genant. 1985). These fractures appear

Materials

and Methods

Abdominal CT scans of 39 patients at the Saint-Louls Hospltal In Paris were selected for study The scans were ob201

202 tained with a CGR CE 10000 CT scanner operated at 130 kvp. Nineteen of the 39 patients were osteoporotic patients who had radiographically determined atraumatic vertebral fractures. These 19 patients included six women 50 to 59 years old (mean age 52.5 years), seven women 60 to 69 years old (mean age 64.0 years), and six men 34 to 59 years old (mean age 44.8 years). The other 20 of the 39 patients had undergone CT scanning (with calibration phantom) for reasons unrelated to osteoporosis and metabolic bone disease and were found to have no abnormality on abdominal CT scans and no evidence of spinal fracture on lateral scout view; these 20 patients were otherwise similar to the first group and included six women 50 to 59 years old (mean age 54.7 years), six women 60 to 69 years old (mean age 63.8 years), and eight men 34 to 56 years (mean age of 44.9 years). A IOmm-thick, transverse section of the vertebral level of L2 or L3 at the region of the basilar venous plexus (the midplane of the vertebral body) was selected for the CT scan of each patient; in the osteoporotic patients, If L2 or 13 was found to be compressed, an adjacent vertebra would be used for analysis. From each patient’s scan the srze (number of pixels) and the CT density (in Hounsfield units) of a vertebra were measured. Specifically analyzed, by means of regions-of-interest (ROl’s) were (1) a rectangular region that fully enclosed the entire vertebra (Fig. 1); (2) the entire area occupied by the vertebra (the area surrounded by the outer boundary of the vertebral bone) (Figs. 2 and 3); (3) the anterior, spongious portion of the vertebra (Fig. 4); (4) the vertebral canal (Fig. 3); (5) the trabecular portion of each vertebral process; and (6) muscle. With use of QCT incorporating K,HP04 calibration (Cann and Genant, 1980; Laval-Jeantet et al., 1979) for our particular scanner, the CT density measurements of bone could be expressed in units (mg/ml) equivalent to mineral density or in units (g) equivalent to bone mineral content. From the above analyses, for each patient group, the mean mineral density and volume for the spongious portions of the vertebra, the compact bone of the vertebra, and the total vertebra area within the CT section were determined using standard CT software (Brassow et al., 1982; Genant et al., 1984; Cann and Genant, 1981; Mazess and Vetter, 1985; Block et al., 1986; Laval-Jeantet, 1985). Compact bone was defined as the difference between total vertebral bone and the low-density bone mineral of both the anterior, spongious portion of the vertebra and the spongious portions of the vertebral transverse processes. The mean CT attenuation (a reflection of bone mineral content) of each region for each patient group was also determined, by multiplyrng the mean volume (voxels) occupied by bone by the mean CT density (HU). The data for the osteoporotic patients were then compared with those of the age- and sex-matched normal (control) patrents with use of the Student’s t test for paired data.

Results All vertebrae studied were either L2 or L3; none of those measured was found to be compressed on lateral scout views. The mean bone mrneral density of the anterior spongious portion of the vertebra for the osteoporotic women in their 50s was found to be significantly lower than that for the normal (control) women in this age group (t = 3.2, DF (degrees-of-freedom) = 10, P < 0.005) (Table I and Fig. 5). The average mineral density of anterior spon-

C D Jones et al

Sponglous

vertebral

bone In osteoporosis

grous vertebral bone for the osteoporotrc women In their 60s was also significantly less than that for the normal (control) age-matched women (t = 3.81, DF = 11, P < 0.001). The osteoporotic men too had a mean bone mmeral densrty for anterior, spongious vertebral bone that was significantly lower than normal (Fig. 6). The differences rn mineral density, or content, of compact bone between the osteoporotic and age-matched control groups were less evident. There was no signrficant difference In compact bone mineral density among the groups of women (Fig. 5); while a significant difference was found between the male groups (Fig. 6). There was no significant difference in compact bone mineral content between the osteoporotic and control women in their 50s whereas there were significant differences in compact bone mineral content between these groups of women In their 60s and between the groups of men (Fig 6). The ratio of mineral density of anterior spongious vertebral bone to that of all compact vertebral bone was found to be significantly lower in the osteoporotlc patrents in all three pairs of matched groups (Fig. 7) (P c 0.0004 for women in therr 50s. P < 0.002 for women in their 60s and P < 0 0003 for the men). In the control women, this ratro appeared to decrease only minimally with age (54% vs. 49%. N.S.); the ratio for the control men (48%) was similar to that of the control women (Table II). Total vertebral bone mrneral density reflects the weighted average of bone mineral density of the anterior spongious vertebral bone, compact vertebral bone, and spongious bone of the transverse processes. The osteoporotic women in their 50s had a mean total vertebral bone mineral density (and content) that was close to that of the age-matched normal (control) women: 161.2 mgiml and 186.0 mg/ml. respectively (t = 1.08, P = 0 15) (Table I). However, in the osteoporotic women in their 60s and In the osteoporotic men, significantly less total vertebral bone mineral denslty (and content) was found than In the ageand sex-matched normal (control) groups (t = 4.28, DF = 11, and P < 0.0006 for these osteoporotlc vs. normal women and t = 3.28, DF = 12, and P = 0.004 for the osteoporotic vs. normal men). The CT attenuation due to the anterior spongious vertebral segment was a small part of the total CT attenuation (or bone mineral content) of the entire vertebral cross section for all groups. The portion of the total attenuation due to the anterior spongrous vertebral bone was less in the osteoporotic women and men than in the respective normal (control) women and men (Fig. 8 and Table II). The osteoporotic women in their 50s had approxrmately 17% of their total vertebral bone attenuation due to the anterior spongrous portion while the normal control women of this age had about 22% of the total vertebral bone attenuation due to the anterior spongious portion A slmllar drfference was found In the women in their 60s: the osteoporotrc women had 14% of their total vertebral bone attenuation due to the anterior spongious portion of the vertebra, while the matched control women were found to have 21% of the total vertebral bone attenuation due to the anterior spongious portion In the osteoporotlc men, the portion of the total vertebral bone attenuation due to the anterior spongious vertebra was approximately 16%; In the normal men, it was 29%. There were no srgnrfrcant differences In the number of voxels of the anterior spongious vertebral bone, compact bone, and total vertebral bone between any of the three pairs of age- and sex-matched groups. No variations in the

CD

Jones et al.: Spongious

vertebral

Fig. 1. CT scan of L3 showing a rectangular

region-of-interest

of bone

were thus established

Fig. 3. CT scan of L3 showing the vertebral canal (2) and the entire vertebral

(ROI) that encloses the entrre vertebra

volume

203

bone In osteoporosls

by our study (Table

III).

Discussion The present study demonstrated that QCT can be used to evaluate the density (or bone mineral content) of both vertebral spongrous and compact bone from a IOmm-thick section from the middle of the vertebra. In all the parred groups a difference was found between the osteoporotic and normal (control) patients in the anterior vertebral spongious bone mineral density: the osteoporotic mea-

body (1) as assessed with the ROI functton

surements were always significantly lower than the normal values. A significant decrease in spongious bone mineral density, or content, can exist in women with early postmenopausal osteoporosis even when the total vertebral bone mineral density or content is normal; this was demonstrated particularly by the women in their 50s. In such cases, osteoporosis could go undetected if total bone mineral content measured by DPA were used for examination because DPA cannot measure bone mineral density or content in the spongious bone alone-a requisite in determination of the small changes of early osteoporosis Usually the decrease in compact bone mineral content IS relatively late, and CT (x-ray) absorption by compact bone in the spine does not correlate well with age (Brassow et

Fig. 2. CT scan of L3 showing use of the ROI function.

In this case, the entire area of the vertebra within the outline of the outer boundary of the vertebral bone IS being quantified.

Fig. 4. CT scan of L3 showing use of the ROI function the anterior spongious portion of the vertebral body

enclosing

204

C.0

-G

350

& E ‘;;

300

: y” E v) E: P, m

Jones et al

Spongious

vertebral

bone In osteoporosis

NORMAL (SD)

TT

250

OSTEOPOROTIC

(SD)

50-59 yrs

60-69 yrs

200 150

g w

100

F p

50

0” 50-59 yrs

50-59 yrs

60-69 yrs

Anterior Spongious Bone

60-69 yrs

Total Vertebral Bone

Compact Bone

Fig. 5. Measurements of bone mineral density (in equivalent of K,HPO,) in normal and osteoporotic women. A statistically significant difference between the osteoporotic and normal patients exists in the bone mlneral density of the anterior sponglous bone even when little or no difference between these groups IS found In bone mrneral density in compact bone or in total ver!ebral bone

al., 1982). Although “compact vertebral bone” as defined in our study is lamellar and probably partially spongious, it showed nc differences in bone mineral density or content between the normal and osteoporotic women in their 50s and small differences (N.S. for density but significant for content) between the normal and osteoporotic women in their 60s. The ability to detect small but real differences In compact bone may be limited, however, because of the larger errors In determining thrs component (as the difference between total vertebral mineral and trabecular mineral). In the case of men, on the other hand, we observed a large and highly significant difference between the normal and osteoporotic groups in compact bone mineral density (and content). Osteoporosis in the men appears to be a different disease or different aspect of osteoporosis than that observed in the postmenopausal osteoporotic women: In the men rt was more severe and was related to corticosteroid treatment In two and to idiopathic causes in four. The lower-than-normal ratio of anterior spongious to compact bone mineral density in the osteoporotic women

Table I. XT-determined bone mineral density (equivalents K,HPO, in mg/ml)* of vertebral bone in osteoporotic normal (control) women and men

Group

Anterior Spongious Vertebral Bone Compact

Women (50-59 years old) 1198 2 273 6 Normal 74 4 2 21 4 6 Osteoporotlc P < 0.005 Women (60-69 years old) 6 Normal 94 8 2 26 5 34 8 -t 29.7 7 Osteoporotlc P < 0 002 Men (34-56 years old) 8 Normal 137 1 k 54 4 449 2 278 6 Osteoporotlc P < 0 002

T

NORMAL (SD) OSTEOPOROTIC

(SD) tl

of and

Total Bone Vertebral Bone

224.6 -+ 50.1 223.0 ” 56.4 P = 0.5

186.0 k 39.6 161.2 2 39.3 P <015

192 2 t 25.0 170 1 + 26 2 P
155.2 -t 23.1 102.5 k 21.2 P < 0.0006

279 8 2 70 7 1674 -c 599 P i 0 005

213.3 5 66.5 1101 t 446 P < 0 004

* Mean value f 1 standard deviation

and men is an lndicatlon of an increasing difference between the two types of bone tissue and can also be used to detect bone loss. The ratio of bone mineral density of the anterior spongious vertebral body to that of compact bone was lower in the osteoporotic women in their 60s than in the osteoporotlc women tn their 50s indicating greater difference in the bone mineral density between the two bone types in the older osteoporotic group. The osteoporotic men also had a lower ratio of anterior sponglous to comoact bone mineral density than did the osteoporotic

50

Anterior Sponglous Bone

Compact Bone

Total Vertebral Bone

Fig. 6. Measurements of bone mlneral density (In equivaleni of K,HPO,) in normal and osteoporotic men In anterior spongious bone, in compact bone, and in total vertebral bone, the osteoporotlc men average statistically significant lower bone mineral denslty In all three compartments than do the normal men

C D Jones et al

Sponglous

vertebral

bone in osteoporosis

205

-r

NORMAL (SD) @j

NORMAL (SD) &I

80

OSTEOPOROTIC (SD) fi

OSTEOPOROTIC (SD) h

70 60 %

T

6ot %

50 30

50

20

40

10 0

30

FEMALES 1 Age 50-59 yrs

20

FEMALES 2 Age 60-69 yrs

MALES Age 34-59 yrs

Fig. 8. The mean CT attenuation

(bone mIneral content) ratlo of anterior sponglous vertebral bone to total vertebral bone That portion of total CT attenuation due to anterior sponglous bone was less In the osteoporotic patients than tn the respective normal (control) patients

10 0I FEMALES 1 Age 50-59 yrs

FEMALES 2 Age 60-69 yrs

MALES Age 34-59 yrs

women In their 50s. There was always a significant difference in this ratio between the osteoporotlc patients and their matched normal counterparts. The number of voxels (or volume) was approximately the same between the normal and osteoporotic patients in the three palred groups and therefore results for “density” and “content“ are generally comparable since the latter reflects density times volume. No discrepancy in number of pixels altered the measurements of CT attenuation (or bone mineral content) The CT attenuation from the ante-

nor spongious vertebral body was less than 30% of that of the total vertebral bone in normal patients; in the osteoporotlc patients. it was less than 17%. These results support the concept that trabecular bone mineral constitutes a relatively small part of the total vertebral bone mlneral mass As with that measured by DPA the CT attenuation of total vertebral bone In our study was a direct index of the total vertebral bone mineral content. Significant dtfferences were found in the measure of total vertebral bone CT attenuation between the normal and osteoporotic women In their 60s and between the normal and osteoporotic men; however. the normal and osteoporotrc women in their 50s showed no significant difference-in this latter case. DPA would be less sensitive. Results of our study show the early. dominant decrease of bone mineral density of the antenor spongious portion of the vertebral body in osteoporosis This calculated decrease. determined by a single-energy QCT measure-

Table II. Ratios of bone mlneral density and CT attenuation

Table Ill. Vertebral bone volume (number of vowels) in

Fig. 7. The bone mineral density ratto of anterior spongious bone to compact bone In all three osteoporotlc groups studied. the ratlo of bone mineral density of anterior spongious bone to that of compact bone was slgnkantly lower that in the matched control groups

(bone mineral content) in osteoporotic (control) women and men *

Group

Bone Mineral Density Ratlo of Anterior SpongiousiCompact Vertebral Bone

Women (50 59 years old) Normal 0539 Osteoporotlc 0 336 P < Women (60-69 years old) Normal 0494 Osteoporotic 0198 P < Men (34-56 years old) Normal 0 478 Osteoporotic 0.255 P i * Mean value 2 1 standard

deviation

osteoporotlc

and normal

CT Attenuation (Bone Mineral Content) Ratio of Anterior Sponglous:Total Vertebral Bone

2 0082 t 0 063 0 0004

0 22 2 0 04 0 167 -c 0035 P
i 0 129 t 0156 0 002

0 208 -c 0 031 0144 i- 0107 PC010

-t 0 073 2 0 112 0 0003

0285 i- 0061 0 163 r 0075 P ( 0 003

Group Women (50-59 Normal Osteoporotic

and normal (control) women and men

Anterior Sponglous Vertebral Bone

Compact

Bone

*

Total Vertebral Bone

years old) 6833 f 680 12732 5 178620308 -+ 710 3 t 83 2 31420 2 2233 19846 2 PC015 f-028 P-038 Women (60-69 years old) Normal 7315~143513043r226721053-_3261 Osteoporotlc 861 9 2 193 6 948 6 t 279 3 1954.7 z P <’ 002 P-07 PC011 Men (34-56 years old) Normal 10838 z 1435 12960 i- 3391 24528 t Osteoporotic 1078 0 i- 98 0 1361 7 I 179 3 2646 5 t P-- 047 P -- 0 34 P-017

2406 2480

404 3

4468 144 9

* The difference between the total vertebral bone and the sum of the anterior sponglous vertebral body and the compact bone IS equal to the voxel count of the trabecular areas of the transverse processes Value given IS the mean z 1 standard deviatton

206

ment, corresponds to the decrease in the actual bone mineral density and the increase in the fat content of the anterior spongious vertebra-both demineralization and increase of marrow fat converge in aging and osteoporosis (Minarre et al., 1984). With srngle-energy QCT the age-related rate of bone mineral loss is over estrmated by 10 to 15% (Cann et al., 1985; Genant et al., 1983). Nevertheless, QCT values exhibit a significant positive correlation with direct measurements of the bone mineral density of vertebral specrmens, with accuracy errors in perimenopausal and early postmenopausal women of about 5 to 10% (Reiser et al., 1985; Rohloff et al., 1982; McBroom et al., 1985). The average CT absorption values of spongious bone from a single slice of various vertebral bodies (T7 to L5 vertebra) are relatively constant in the same indrvrdual (Brassow et al., 1982; Lampmann, 1983). The use of a single CT section from L2 or L3 measurement of mean vertebral bone mineral content should be sufficient for diagnosis, although the averaging of several vertebral sections should give a more accurate result and better longitudinal reproducibility. The density of vertebrae is related to Its structural strength, and both cortical (compact) and trabecular (spongious) bone contribute to overall vertebral strength (Rockoff et al , 1969), although the latter appears to be of greater Importance (McBroom et al., 1985). In older patients the cortex contributes 45-75% of the peak strength of the vertebral body under compressive load-but this is after 70% of the strength has been lost princrpally in the trabecular component of bone (Rockoff et al., 1969). However, recent investigation (McBroom et al., 1985) has shown that removal of the cortex is associated with about a 10% reduction in vertebral strength. Therefore, even in older patients, the vertebral CT measurement of the anterior spongious vertebral bone mineral density may be more important than total bone mineral content In determining risk for spinal fracture (Cann et al., 1985). In conclusron, our study shows that wrth use of QCT one can obtarn measurements of bone mineral density or content in the anterior spongious vertebral body, the more dense compact bone, or a combination of the two (as can be obtained with DPA). The CT attenuation of the anterror spongrous bone is less than 30% of the total vertebra In normal subjects and less than 17% in osteoporotic patients, indicating a disproportionate loss of sponglous compared to compact bone in osteoporosis. Therefore, for the early detection of vertebral bone mineral loss with its resulting spinal osteoporosis, QCT measurement of bone mineral density in the anterior spongious bone appears to be of considerable importance. References Block J E Genanf H K and Black D Greater vertebral bone mineral mass rn exercrsng young men Wesr J Med 145 39942 1986 Brassow F Crone-Munzebrock W Weh L Kranz R and Eggers-Stroeder G Correlations between breaking load and CT absorpt’on values of vertebral bodres Eur J &d/o/ 2 999101 1982 Cann C E and Genant H K Comparrson of cancellous and integral spinal mineral loss In oophorectomrzed women usrng quantrtative computed tomography Caiof T/ssue Inl 33 307. 1981 Cann C E and Genant H K Precise measurement of vertebral mrnerai content using computed tomography J Compui Assisl iomogr 4 493-500 1980 Cann C E Genant. H K Kolb F 0 and Etttnger B Quantitative computed tomography for predictIon of vertebral fracture risk Bone 6 I-7 1985 Dunn W L Wahner H W and Riggs B L Work in progress Measurement of bone mineral content in human vertebrae and hrp by dual photon absorptrometry Rad/oiogy 136 485-487, 1980

C.D Jones et al

Sponglous

vertebral

bone In osteoporosis

Genant H K and Boyd D Quantttatrve bone mIneral analysis using dual energy computed tomography invest &d/o/ 12 545 -551 1977 Genant H K Cann C E Boyd D P Kolb F 0, Ettrnger B and Gordan G S Quantitative computed tomography for vertebral mineral delerm~natrorr in Ciinrcai Dsorders of Bone and MInera/ Mekbolwn B Frame antr J T Potts. Jr eds PrInceton Excerpta MedIca, inter Congress Senes No 617, 1983, pp 40 -47 Genant H K Cann C E Etttnger B and Gordan G S Quantltatlve corn puted tomography of vertebral sponglosa a sensrtrve method for de tecting early bone loss after oophorectomy Ann lnfern Med 97 699 705, 1982 Genant H K Powell M R Cann C E et al Comparison of melhods for 111 vrvo sprnal mrneral measurement In Osteoporosis C Chrrstranserr, C D Amaud, B E C Nordrn. et al eds Denmark, Aalborg Slrftsbog trykkerl 1984 pp 97 102 Gordan G S and Genant H K The agrng skeleton icine 1 95- 118 1985

Clrnics in Genalric Med

Johnston C Jr Osteoporosis An overvIew In Cirn/ca/ Disorders of Bone and MIneral Metaboiwn B Frame and J T Potts, Jr, eds Prtnceton Excerpta MedIca. Inter Congress Series No 617, 1983a pp 317 322 Johnston C C 1983b

Jr

Osteoporosrs

Henry Ford Hasp

Med

J 31 223

226

Krolner B and Nielsen S P Bone mrneral content of the lumbar spine 111 normal and osteoporotlc women cross-sectlonal and long~tud~na studies C/in SCI 62 32%36 1982 Lampmann L E H Aria/ Skeietai Dens/tomeiry Utrecht. Holland schappelrfke Urtgeverrf Bunge Proefschrift 1983. pp 82-84

Weten

Laval-Jeantel A M Jones C D Bergot C Laval-Jeantet M H and Genant H K Comparrson of bone loss from sponglosa and from compact ver tebral bone rn the aging process and rn osteoporotlcs Proceedings of the 5th lnternat~onal Workshop on Bone and Soft Trssue Usrng Come puted Tomography, Bretton Woods, October 14 18 1985 LavaI-Jeantet M Laval-Jeantet A M Lamarque J L and Demou,in B Evaluation de la mineral~sat~on osseuse vertebrale par !omographe computerlsee J Radio/ 2 87 -93, 1979 Mazess R B Non~tnvaslve measurement of bone In Osteoporosl~ Barmel, ed Orlando FL Grune & Stratton Inc 1979

/i U S

Mazess R B ana Vetter J Comparrson of dual-photon absorptrome!ry and dual-energy computed tomography for vertebral mineral J Compu: Assist iomogr 9 624-625. 1985 M&room H J Hayes W C Edwards W T Goldberg R P and Whrle A A Predrctron of vertebral body compressrve iracture using quantltalrve computed tomography J Bone Joint Surg 67A 1206 1214 1985 Mtnalre P Edouard C ArIot M and Meunler P J pleglc patients ‘&ICI/ ‘ifssue inl 36 338340

Marrow changes ‘I? pard 1984

Parfltt A M Age-related struc:ural changes In trabecular ana corilcal bone cellular mechanrsms and b:omechanicai consequences Calcif T,SSIIC int 36 S123%S128 1984 Retser U J Ru:t B K Genant H K ana Reinbo,o W D Analyss of taclors nfluenclng the accuracy of QCT and DPA Proceedings of !he 5th Inter natlonal Workshop on Bone and Soft Tissue Using Computea Tomog raphy, Bretton Woods Gctober 14-18, 1985 Rockoff S D Sweet E and BluesteIn J The relative contnbu!ion of ‘rdbec ular and cortical bone to the strength of human lumbar ver’ehrae Caicif Tissue Res 3 163 175 1969 Rohloff R Hrlzler H Aincfl W and Frey K W Vergletichende Messunger: des Kalksalrgehaltes spongroser Knochen m:ttels Computerlomogra phle und J-125 Photonen Absorptionsmethode In CT ‘8? In/erna~ 0onaies Computeriomographie Symposium J Llssner J L Doppmar eds Konstanz Schetztor~Verlag 1982 pp 124- 130 Weaver J K and Chalmers J Cancellous bone Its s!rength and changes with aging and an evaluatron of some methods for measur#ng Its rnrin era1 content J Bone Joux Surg 48A 2899299 1966 Wilson C R Predlctron of femoral neck and spmne bone m neral con!erlt from the BMC of the raolub or ulna and the relatIonshIp between bone strength and BMC In lntemat~onal Conference on Bone MIneral Mea surernenl R B Maze% ed WashIngton DC US Dept of health EdL cation and Welfare Puh:lcation No (NIH) 75683 1975. oo 51 59 ~ ~___ _ ~ ~~ __.___ Received May 30 19R6 Rewed Janilarv 2 I 1987 Acceuled Feb& 9 1987