- Email: [email protected]
Altered bone mineral metabolism in patients with osteoarthritis
Joint Bone Spine 2000 ; 67 : 521-7 © 2000 Éditions scientifiques et médicales Elsevier SAS. All rights reserved S1297319X00002189/FLA
ORIGINAL ARTICLE
Altered bone mineral metabolism in patients with osteoarthritis Yasser M. El Miedany1*, Annie N. Mehanna2, Mohammed A. El Baddini3 1 Rheumatology and Rehabilitation Department, Ain Shams University Faculty of Medicine, Cairo, Egypt ; 2Radiology Department, Ain Shams University Faculty of Medicine, Cairo, Egypt ; 3Clinical Pathology Department, Cairo University Cancer Institute, Cairo, Egypt
(Submitted for publication July 12, 1999; accepted in revised form June 28, 2000)
Summary – Objective. Investigation of the relationship between osteoarthritis (OA) and mineral density, and determination of any alteration in bone mineral, metabolism as assessed by biochemical markers of bone resorption and formation. Methods. Forty females and 20 males were included in the study. Spinal OA as well as knee OA were defined from radiographs and graded according to Lane et al.’s and Spector et al.’s scoring systems. Bone mineral density (BMD) of the lumbar spine was measured by osteo CT. Bone turnover rates were estimated by measuring biochemical markers of bone resorption (urinary deoxypyridinoline) and bone formation (bone-specific alkaline phosphatase). Forty females and 20 males of the same age were studied as a control group. Results. BMD was greater in women with spinal OA as compared to controls (P < 0.05). Also, males with OA had a non-significantly higher BMD than controls. The bone resorption markers were higher than normal values. However, they were lower than the control group. Similarly, the bone formation markers were lower as compared to the control group. Conclusion. Spinal OA is associated with higher BMD. This protective effect of spinal OA against osteoporosis may be mediated through decreased rate of bone turnover. Joint Bone Spine 2000 ; 67 : 521-7. © 2000 Éditions scientifiques et médicales Elsevier SAS bone mineral density / bone turnover / osteoarthirits
INTRODUCTION Although osteoarthritis (OA) and osteoporosis are both common conditions with high prevalence in the older age group, they are not purely due to simple aging and do not anthropometrically affect distinct populations. However, the coexistence of the two disorders in individual patients has been considered to be rare [1]. * Correspondence and reprints: Dr. Yasser El Miedany, 2 Italian Hospital Street Abbassia, Cairo, 11381, Egypt. E-mail address: [email protected] (Y.M. El Miedany).
Hence, an inverse relationship between OA and osteoporosis has been proposed. Evidence for such a relationship may give rise to etiologic clues; however, the studies that are presently available do not allow definite conclusions [2]. Studies that favour an inverse relationship [3-6] as well as studies in which there is no sufficient support for such a relationship [7, 8] have been published. An increase in measured bone mineral density (BMD) in a region affected by OA may result from artifactual error caused by the presence of osteophytes, joint space narrowing and sclerosis within the region of interest.
522
Y.M. El Miedany et al.
Alternatively, it may reflect a generalized effect of OA on the skeleton, or a combination of both. If OA has a generalized effect on the skeleton, the study of bone metabolism in patients with OA may be important in understanding the pathophysiology of osteoporosis. There has been little study of the underlying mechanism of the interaction between bone mass and OA. The development of noninvasive biochemical assays for markers of bone resorption and formation enable estimation of bone turnover in large cohorts of subjects. There have been few reports of biochemical markers in subjects with spinal OA [9]. Likewise, few studies have been done on men suffering from OA [6, 10]. In this work we performed a controlled study to determine whether the presence of spinal OA was associated with a generalized increase in BMD and if so, if there exists a sex difference. This work was done to study as well the underlying mechanism at the tissue level through assessment of biochemical markers of bone metabolism. SUBJECTS AND METHODS Subjects Forty postmenopausal Egyptian females and 20 Egyptian males diagnosed with osteoarthritic changes in the form of lumbar spondylosis and/or knee osteoarthritis were randomly collected from those attending the physical medicine outpatient clinic at Ain Shams University. Full medical history and thorough clinical examination were done for each patient with a stress on symptoms and signs of lumbar spondylosis, knee and hip OA. At initial evaluation the height and weight were measured and body mass index (BMI wt/height2) was calculated as an estimate of obesity. None of our patients had history of fractures. Patients who gave a history of renal affection were excluded from the study. Also, patients with a history of disease (e.g., diabetes mellitus, hyperthyroidism, endogenous or iatrogenic hypercorticism), malignancy or medications (steroids, thyroid hormone, diuretics, antihypertensive drugs containing thiazides, antimitotics, heparin and anticonvulsants) that may alter bone mineral metabolism were also excluded from the study. None of the postmenopausal women were on hormone replacement therapy or any other form of treatment for osteoporosis. None of our patients was a smoker or an alcoholic. None were bedridden.
Radiological assessment Each subject had anteroposterior and lateral radiographs of the cervical, thoracic and lumbar spine in addition to both knee and hip joints. Spinal OA was defined from the radiographs and scored according to the criteria of Lane et al. [11]. They were graded according to their score into: mild (≤ 2), moderate (3–5) and severe (≥ 6). Knee OA was defined and scored from the radiographs according to Spector et al. [12]. They were also graded into mild (≤ 3), moderate (4–7) and severe (≥ 8). The maximum of the left and right sides was used for the analysis. The radiographs were all reviewed by a single observer irrespective of bone mineral density result. Bone densitometry Bone mineral density (BMD) was measured using quantitative CT (Siemens) of L1–L3 lumbar vertebrae. The procedure was done according to the Cann and Gennant method [13]. Briefly, a phantom (QCT Bone Mineral Analysis System, San Francisco, CA) containing potassium phosphate standard was placed under the patient. Cursors were placed on the vertebra image to define a 100-mm thick transverse section through the center of L1, L2, L3. Cross-sectional images of each vertebra were obtained and used to position elliptical cursors in the trabecular area of each vertebral body. CT counts were then obtained for the selected vertebrae. Spinal measurements were referenced to a calibration curve obtained from the standard (Caucasian, European) and were expressed in milligrams per cubic centimeter. The results were expressed in a Z-score (number of standard deviation above or below the normal mean after comparison with age- and sexmatched normal control values supplied by the manufacturer). The T-score for females was calculated according to the equation given by the manufacturer: BMD (trabecular) – 159.1 (BMC of 35-year-old female) ± 27.6 (SD). The T-score for males was calculated according to the equation given by the manufacturer: BMD (trabecular) – 174.5 (BMC of 20-year-old male) ± 26.5 (SD). Biochemical markers of bone turnover The following assays were performed: serum Ca, phosphorous and alkaline phosphatase, in addition to bone resorption marker (urinary deoxypyridinoline) and
523
Altered bone mineral metabolism and OA Table I. Comparison of the osteoarthritic female and male patients with the control group in terms of the variables tested in this study. Variable Age (years) Years since menopause Parity Body mass index (weight/height2) Calcium (mmol/L) Phosphorus (mmol/L) Alkaline phosphatase (mmol/L) Corticol BMC (mg/cm3) Trabecular BMC (mg/cm3) Z-score T-score L-spine X-ray score Knee X-ray score Deoxypyridinoline (nmol/mmol) Bone alkaline phosphatase (U/L)
Female group
Male group
OA patients
Control group
P-value
OA patients
Control group
P-value
53.7 ± 5.1 6.2 ± 4.2 5.4 ± 3.4 33.0 ± 5.4 2.23 ± 0.14 3.74 ± 0.62 161.9 ± 59.9 241.0 ± 37.1 99.2 ± 29.6 –0.66 ± 0.89 –2.17 ± 1.15 4.8 ± 1.6 5.8 ± 1.6 6.74 ± 1.45 17.1 ± 5.9
52.9 ± 5.9 7.8 ± 5.6 5.3 ± 3.8 29.3 ± 3.4 2.19 ± 0.12 3.71 ± 0.48 157.0 ± 58.8 234.6 ± 42.1 81.0 ± 38.7 –1.29 ± 0.87 –2.83 ± 1.03 0.25 ± 0.44 0.30 ± 0.47 7.95 ± 1.97 22.1 ± 5.7
NS NS NS P < .001 NS NS NS NS P < .01 P < .01 P < .01 P < .01 P < .01 P < .01 P < .01
52.3 ± 6.4
51.5 ± 6.4
NS
bone formation marker (bone-specific alkaline phosphatase). Aliquots of early morning urine were stored frozen at –70° C until each assay was done in batches. Venous blood samples were obtained, centrifuged and stored in a similar manner. Deoxypyridinoline was measured in urine by estimation of Pyrilinks-D using competitive enzyme immunoassay method. Metrabiosystemst kits were used. Bone-specific alkaline phosphatase was assessed using the enzyme immunoassay method from Metrabiosystemst, which uses monoclonal anti-bone alkaline phosphatase antibody coated on micrometer strips to capture and quantitative bone alkaline phosphatase activity. Control group Forty Egyptian postmenopausal female and 20 Egyptian males of matched age were randomly collected and included as a control group. They had no history of any disease or medication known to affect bone metabolism. They were subjected to the same laboratory and radiological assessment. The study and aim of the work were explained to all subjects who participated in the study. All patients and controls provided their informed consent. Statistical analysis Data was analyzed using the Statistical Package for Social Sciences (SPSS) Software on an IBM PC. Different variables were described by the mean ± SD. Student’s t test was used to compare between cases and
31.6 ± 3.5 29.7 ± 2.7 2.16 ± 0.11 2.19 ± 0.09 3.69 ± 0.519 3.61 ± 0.386 129.4 ± 27.9 152.0 ± 42.2 318.3 ± 45.05 299.95 ± 58.10 153.5 ± 27.9 146.3 ± 31.3 0.87 ± 0.68 0.82 ± 0.66 –0.93 ± 1.113 –1.10 ± 1.074 4.7 ± 1.6 0.3 ± 0.47 5.6 ± 1.6 0.25 ± 0.44 4.41 ± 1.74 5.09 ± 1.89 16.2 ± 4.0 17.4 ± 3.6
NS NS NS NS NS NS NS NS P < .01 P < .01 NS NS
controls. The Mann-Whitney test was used to compare the variables that were not uniformly distributed. Spearman’s correlation coefficient was used to assess the correlation between trabecular and cortical bone mineral density, and knee and lumbar spine grading of osteoarthritis. Effect of obesity was assessed by Spearman’s correlation and multiple linear regression using Stata 6 for Windows 98. RESULTS Table I compares the female and male osteoarthritis patients with the control groups in terms of the variables tested in this study. It shows that osteoarthritic patients had higher bone mineral content compared to the control group. This difference was significant (P < 0.01) in the female group while it was not significant in the male group. Studying the bone markers revealed that bone resorption and formation markers were lower among osteoarthritic patients than the control group. Again, this difference was significant (P < 0.01) in the female patients while it was nonsignificant in the male group. Osteoarthritic patients had significantly higher BMI compared to the control group (P < 0.01 in the female group). Tables IIa and b are an analysis of variance showing the mean value and the correlation of trabecular and cortical bone mineral content among different grades of knee and lumbar spondylosis. It confirms the significant correlation between BMC and degree of osteoarthritis. Table III shows a comparison between obese osteoarthritic (28/40) (BMI > 30) versus obese non-
524
Y.M. El Miedany et al.
Table IIa. Analysis of variance showing the mean values and the correlation of the bone mineral content between the different grades of lumbar spondylosis (female group). Variables
Mild (Grade I)
Moderate (Grade II)
Severe (Grade III)
Significance
Cortical bone mineral content Trabecular bone mineral content
236.5 ± 40.4 92.7 ± 26.7
286.7 ± 55.1 121.0 ± 37.9
279.2 ± 53.8 139.6 ± 36.43
P < 0.001 P < 0.001
osteoarthritic females (18/40) regarding the Z-score as well as a comparison between non-obese osteoarthritic (BMI < 30) (12/40) versus non-obese nonosteoarthritic females (22/40). In both cases there was a significant difference (P < 0.03 and P < 0.02, respectively). Table IV shows that body weight as expressed by body mass index did not correlate significantly with the osteoarthritis X-ray score nor bone mineral density (r = 0.116 and 0.096, respectively). On the other hand, bone mineral density correlated significantly with the osteoarthritis X-ray score (r = 0.729, P < 0.001). Controlling for body mass index (table V), bone mineral density was still significantly correlated with osteoarthritis. Multiple regression analysis has been performed twice: in the first model the T-score was posed as the dependent variable while in the second bone mineral density was used as the dependent variable. In both models body mass index has been excluded. The regression coefficient of the osteoarthritis score was 0.681 and 0.636, respectively (table VI). DISCUSSION We found that elderly Egyptian women with moderate to severe radiographic features of OA had significantly (P < 0.02) higher BMD in the axial skeleton than did women with no or minimal radiographic findings of OA. BMD was increased in women who had not only spinal but also knee OA. The higher BMD of women
with OA was not explained by potential confounding variables such as age, weight, physical activity or use of medications that increase bone mineral density. Similarly, on the male side, although there was no significant difference, men with OA had also more bone mass than those without OA. Our findings are consistent with those of previous studies [6, 14-17] that showed that an increase in BMD was found not only in the spine but also at sites where there is no influence of osteophytes or subchondral sclerosis. In addition, our results showed significant correlation between BMD and X-ray scores of knee and lumbar spine osteoarthritic changes. This agrees with what has been reported by Nevitt et al. [18], who found that higher bone mineral density at both appendicular and axial sites was associated with the presence and size of hip osteophytes, while isolated narrowing without osteophytes was not associated with higher bone mineral density. Measurement of the biochemical markers of bone turnover enabled us to assess the mechanism of increased bone mineral density in patients with OA. In the female group bone resorption and formation markers were significantly (P < 0.01) lower in patients with OA compared to the control group. On the male side markers of bone resorption and formation were lower in the OA group compared to the control group although the differences were not significant. This might imply that patients with OA had a lower rate of bone turnover.
Table IIb. Analysis of variance showing the correlation of the bone mineral content between the different grades of knee OA (female group). Variables Trabecular bone mineral content Cortical bone mineral content
Mild (Grade I)
Moderate (Grade II)
Severe (Grade III)
Significance
91.2 ± 27.2 236.5 ± 41.6
112.9 ± 32.5 273.7 ± 51.9
144.5 ± 37.3 284.4 ± 57.0
P < 0.001 P < 0.001
Table III. Comparison of Z-score between obese, non-obese and control female groups.
Z-score
Obese OA (28/40)
Obese control (18/40)
P-value
Non-obese OA (12/40)
Non-obese control (22/40)
P-value
–0.33 ± 0.96
–1.07 ± 1.23
P < 0.03
–0.41 ± 1.23
–1.12 ± 2.25
P < 0.02
Altered bone mineral metabolism and OA Table IV. Spearman’s correlation showing the correlation between the variables included in the study. Variable
BMI
OA score
T-score
BMD
BMI
1.00
r = 0.116 NS 1.000
r = 0.151 NS r = 0.729 P < .001 1.00
r = 0.096 NS r = 0.720 P < .001 r = 0.951 P < .001 1.000
OA score T-score BMD
r = 0.116 NS r = 0.151 NS R = 0.96 NS
r = 0.729 P < .001 r = 0.720 P < .001
r = 0.951 P < .001
BMD: bone mineral density; OA: osteoathritis X-ray score.
Our results agree with what has been reported by Peel et al. and Gevers et al. [9, 14], who found in their studies a decrease in both resorption and formation markers in women with spinal OA. Generalized OA could be a good negative indicator for selecting patients at risk for osteoporosis. The generalized increase in bone mineral density in OA indicates that this disease might initially be a subchondral bone disease rather than a cartilage disorder [19]. There are several possible reasons for the inverse relationship between OA and osteoporosis. In our study as well as in others [3, 4, 6], the association exceeds the single joint and even correlates to the degree of OA. This suggests that generalized factors might play a role. Table V. Correlation between the variables tested in this study after controlling for body weight. Variable
BMD
OA score
T-score
BMI
1.000
r = 0.6315 P < .001 1.000
r = 0.9228 P < .001 r = 0.6756 P < .001 1.000
OA score T-score
r = 0.6315 P < .001 r = 0.9228 P < .001
r = 0.6756 P < .001
BMD: Bone Mineral Density; OA: Osteoarthritis X-ray score.
Table VI. Multiple regression analysis showing the correlation between osteoarthritis and bone mineral density in our group of patients. Dependent variable
Constant
Regression coefficient for OA score
P-value
T-score BMD
–4.344 37.681
0.681 0.636
P < .001 P < .001
BMD: Bone mineral density.
525
Elevated levels of insulin-like growth factors (IGF) I and II and transforming growth factor beta (TGFβ), the most abundant growth factors in human bone, have been shown to be increased in the matrix of the iliac crest in patients with OA. This finding is consistent with the hypothesis of a generally increased osteoblastic biosynthetic activity in these patients [20]. The anabolic properties of these growth factors might contribute to an increased stiffness of the subchondral bone and subsequent cartilage damage. Both IGFs and TGFβ are known to enhance osteoblastic replication and to promote the synthesis of different protein components, including procollagen I [21]. In addition, IGF I and II induce matrix opposition and decrease collagen degradation as well as the expression of interstitial collagenase, suggesting a role in the preservation of bone matrix. Besides obesity as a shared determinant [22], physical activity early in life may produce the association, since high levels of physical activity during youth have been shown to be of key importance in reaching peak bone mass [23] and may also lead to an increased risk of OA later in life [24, 25]. Gender difference in terms of the significant inverse relationship between OA and bone mineral density can be explained by the physiological changes that occur in woman at menopause leading to increased bone resorption, hence osteoporosis. The increased bone turnover might magnify the systemic effect of OA in females compared to males, who do have higher backgrounds of bone mass and do not suffer hormonal changes at that age. Moreover, in the work done by Felson et al. [26], the authors reported that women have a 1.5- to twofold higher incidence of knee OA compared to men. Quam et al. [27] reported that the age-adjusted frequency of total knee and hip arthroplasty was respectively 40% and 30% higher in women than in men. The association of obesity with the presence of OA has been examined through analysis of cross-sectional data and longitudinal cohort designs [26, 28-30]. Most studies used BMI as the measure of obesity. In the study done by Hochberg et al. [10], it was reported that body weight was associated with both definite and bilateral knee OA in both sexes, and this work supports contribution of mechanical as opposed to systemic factors to explain the association. Results of our study revealed that BMI was not significantly correlated with trabecular bone mineral content or T-score, nor with the X-ray score. However, the osteoarthritic group of patients studied had a significantly higher body mass index (P < 0.01) compared to the control group, whereas
526
Y.M. El Miedany et al.
after ruling out the role of body weight, osteoarthritis was significantly correlated with bone mineral density. Our results further extend the observation that body weight is associated with OA. Although the association of obesity with osteoporosis in patients with spinal OA is not frequently reported as compared to knee OA, our results support the notion that the systemic effect of obesity might have a protective effect against osteoporosis. But it is not the only factor involved. As conversion of androsterone to estrone in fat tissue represents the main source of estrogen in postmenopausal women [31] and estrogen-induced secretion of both IGFs and TGFβ by osteoblast-like cells is well documented, estrogen-mediated growth factor deposition in bone might be involved in the increased density and rigidity of the trabecular bone in OA, contributing to the pathogenesis of OA but also protecting from osteoporotic fracture [32]. However, in a recent study done on a cohort of 5,552 elderly women [33], subjects with OA did not have a significantly reduced risk of osteoporotic fracture. More interesting are the findings reported by Foss and Byers [34] and Roh et al. [35], who found that the negative association of OA with osteoporosis can only partially be accounted for by an absolute or relative estrogen excess. An increase in circulating growth hormone levels, both baseline growth hormone and growth hormone response to indirect stimuli, has been documented in OA compared to osteoporosis and may be of pathogenetic significance as growth hormone has a positive effect on bone and cartilage. Vertebral bone mineral density is usually measured in the anteroposterior plane though this method may give falsely high values. An increase in measured BMD in a region affected by OA may result from artifactual error caused by the presence of osteophytes, joint space narrowing and sclerosis within the region of interest. Aortic calcification is another potential confounder. Moreover, the anteroposterior BMD measurement using DEXA machines includes the posterior spinal element rich in cortical bone, while the main concern is the trabecular bone present in the vertebral body. In addition, the lower reproducibility and technical difficulties associated with lateral DEXA serve to limit the application of this method [36]. These were the reasons why we preferred to assess BMD using osteo CT rather than anteroposterior DEXA. Osteo CT has the advantage of selectively measuring the trabecular and cortical compartments of the vertebrae and has therefore been recognized as a sensitive method for assessment of bone
mineral density in patients with osteoporosis. Osteo CT also has the advantages of excluding the posterior compact bone elements of the vertebrae, any hypertrophic and degenerative change and/or vascular calcification, which commonly occur in aged people like our group of patients. Lastly, in the study done by Guglielmi et al. [37], the authors reported that although anteroposterior DEXA is a highly precise technique, with a 1% variation in short-term reproducibility studies, quantitative CT has been shown to help discriminate between healthy women and those with osteoporosis better than anteroposterior DEXA. In conclusion, the present study demonstrates that radiological OA is associated with a generalized increase in BMD and a decrease rate of bone turnover. These results are consistent with the hypothesis that OA has a protective effect against bone loss, mediated by a lower rate of bone turnover. This is more pronounced in women than in men. Alternatively, our results may suggest the existence of a subgroup of women with increased bone turnover, who are susceptible to the development of osteoporosis and who do not develop physiological degenerative changes of the skeleton with aging. REFERENCES 1 Foss MVL, Byers PD. Bone density, osteoarthrosis of the hip and fracture of the upper end of the femur. Ann Rheum Dis 1972 ; 31 : 259-64. 2 Hodon LD, Wright V, Smith NI. A. Bone mass in osteoarthritis. Ann Rheum Dis 1992 ; 51 : 823-8. 3 Gevers G, Dequeker J, Geusens P. Physical and histomorphological characteristics of iliac crest bone differ according to the grade of osteoarthritis at the hand. Bone 1989 ; 10 : 173-7. 4 Reid IR, Evans MC, Ames R, Wattie DJ. The influence of osteophytes and aortic calcification on spinal mineral density in postmenopausal women. J Clin Endocrinol Metab 1991 ; 72 : 1372-6. 5 Äström J, Beertema J. Reduced risk of hip fractures in the mothers of patients with osteoarthritis of the hip. J Bone Joint Surg Br 1992 ; 74 : 270-5. 6 Hannan MT, Anderson JJ, Zhang Y. Bone mineral density and knee osteoarthritis in elderly men and women. The Framingham Study. Arthritis Rheum 1993 ; 36 : 1671-6. 7 Price T, Hesp R, Mitchell R. Bone density in generalized osteoarthritis. J Rheumatol 1987 ; 14 : 560-5. 8 Orwall ES, Oviatt SK, Mann T. The impact of osteophytic and vascular calcifications on vertebral mineral density measurements in men. J Clin Endocrinol Metab 1989 ; 70 : 1202-6. 9 Peel NF, Barrington NA, Blumsohn A. Bone mineral density and bone turnover in spinal osteoarthritis. Ann Rheum Dis 1995 ; 54 : 867-71. 10 Hochberg MC, Lethbridge-Cejku M, Scott WW. Upper extremity bone mass and osteoarthritis of the knees. Data from the Baltimore Longitudinal Study of Aging. J Bone Miner Res 1995 ; 10 : 432-6. 11 Lane NE, Nevitt MC, Genant HK, Hochberg MC. Reliability
Altered bone mineral metabolism and OA
12 13 14
15
16
17
18 19 20 21 22 23 24
of new indices of radiographic osteoarthritis of the hand and hip and lumbar disc degeneration. J Rheumatol 1993 ; 20 : 1911-6. Spector TD, Cooper C, Cushanghan DJ, Dieppe P. A radiographic atlas of knee OA. London: Springer-Verlag; 1992. Cann CG, Gennant HK. Precise measurement of vertebral mineral content using computed tomography. J Comput Assist Tomogr 1980 ; 4 : 493-7. Gevers G, Dequeker J, Geusens P, Verstraeten A, Vanderschueren D. Serum bone Gla protein and other markers of bone mineral metabolism in postmenopausal osteoporosis and osteoarthritis. J Orthop Rheum 1988 ; 1 : 22-7. Hart DJ, Mootoosamy I, Doyle DV, Spector TD. The relationship between osteoarthritis and osteoporosis in the general population. The Chingford Study. Ann Rheum Dis 1994 ; 53 : 158. Marcelli C, Favier F, Kotzki PO. The relationship between osteoarthritis of the hands, bone mineral density, and osteoporotic fractures in elderly women. Osteoporos Int 1995 ; 5 : 382-6. Burger H, Van Daele PLA, Odding E. Association of radiographically evident osteoarthritis with higher bone mineral density: an increased bone loss with age. The Rotterdam Study. Arthritis Rheum 1996 ; 39 : 81-6. Nevitt MC, Lane NE, Scott JC. Radiographic osteoarthritis of the hip and bone mineral density. Arthritis Rheum 1995 ; 38 : 907-1002. Dequeker J, Mokassa L, Aerssens J. Bone density and OA. J Rheumatol 1995 ; 22 Suppl : 98. Dequeker J, Johnell O. The Medos Study Group. Osteoarthritis protects against femoral neck fractures. The MEDOS Study Experience. Bone 1993 ; 14 : 51. Coppens M, Dequeker J. Osteoarthritis, obesity and hormones. Current Opin Orthop 1991 ; 2 : 482. Dequeker J, Goris G, Uytterhoeven R. Osteoporosis and osteoarthritis: anthropometric distinctions. JAMA 1983 ; 249 : 1448-53. Welten DC, Kemper HCG, Post GB. Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J Bone Miner Res 1994 ; 9 : 1089-94. Heliövaara M, Mäkelä M, Impivaara O. Association of overweight trauma and workload with coxarthrosis. Acta Orthop Scand 1993 ; 64 : 513-7.
527
25 Cooper C, McAlindon T, Coggon D. Occupational activity and osteoarthritis of the knee. Ann Rheum Dis 1994 ; 53 : 90-5. 26 Felson DT, Anderson JJ, Naimerk A, Walker AM. Obesity and knee osteoarthritis. Ann Intem Med 1988 ; 109 : 18-23. 27 Quam JP, Michet CJ, Wilson MG, Rand JA. Total knee arthroplasty. A population-based study. Mayo Clin Proc 1991 ; 66 : 589-94. 28 Hartz AJ, Fischer ME, Brill G. The association of obesity with joint pain and osteoarthritis in the HANES data. J Chronic Dis 1986 ; 39 : 311-9. 29 Van Saase JLCM, Vandenbroucke JP, Van Romunde LKJ, Valkenburg HA. Osteoarthritis and obesity in the general population: a relationship calling for an explanation. J Rheumatol 1988 ; 15 : 1152-8. 30 Felson DT. The epidemiology of knee osteoarthritis: results from the Framingham Osteoarthritis Study. Semin Arthritis Rheum 1990 ; 20 Suppl 1 : 42-50. 31 Frumar A, Meldrum D, Geola F, Shamonki I, Tataryn I, Deftos L, et al. Relationship of fasting urinary calcium to circulating estrogen and body weight in postmenopausal women. J Clin Endocrinol Metab 1980 ; 50 : 70-5. 32 Dequeker J, Mohan S, Finkelman R, Aerssens J, Baylink D. Generalized osteoarthritis associated with increased insulin-like growth factor types I and II and transferring growth factor beta in cortical bone from the iliac crest. Arthritis Rheum 1993 ; 36 : 1702-8. 33 Arden NK, Nevitt MC, Lane NE, Gore R, Hochberg MC, Scott JC, et al. Osteoarthritis and risk of falls, rates of bone loss and osteoporotic fractures. Arthritis Rheum 1999 ; 42 : 137885. 34 Foss MV, Byers PD. Bone density, osteoarthritis of the hip and fracture of the upper end of the femur. Ann Rheum Dis 1972 ; 31 : 259-64. 35 Roh YS, Dequeker J, Mulier JC. Bone mass in osteoarthritis measured in vivo by photon absorption. J Bone Joint Surg 1974 ; 56 : 587-91. 36 Haddaway MJ, Davic MW, McCell IW. Bone mineral density in healthy normal women and reproducibility of measurements in spine and hip using DEXA. Br J Radiol 1992 ; 65 : 213-8. 37 Guglielmi G, Grimston SK, Frischerk C, Pacific R. Osteoporosis: diagnosis with lateral and posteroanterior DEXA compared with quantitative CT. Radiology 1994 ; 192 : 845-9.
ORIGINAL ARTICLE
Altered bone mineral metabolism in patients with osteoarthritis Yasser M. El Miedany1*, Annie N. Mehanna2, Mohammed A. El Baddini3 1 Rheumatology and Rehabilitation Department, Ain Shams University Faculty of Medicine, Cairo, Egypt ; 2Radiology Department, Ain Shams University Faculty of Medicine, Cairo, Egypt ; 3Clinical Pathology Department, Cairo University Cancer Institute, Cairo, Egypt
(Submitted for publication July 12, 1999; accepted in revised form June 28, 2000)
Summary – Objective. Investigation of the relationship between osteoarthritis (OA) and mineral density, and determination of any alteration in bone mineral, metabolism as assessed by biochemical markers of bone resorption and formation. Methods. Forty females and 20 males were included in the study. Spinal OA as well as knee OA were defined from radiographs and graded according to Lane et al.’s and Spector et al.’s scoring systems. Bone mineral density (BMD) of the lumbar spine was measured by osteo CT. Bone turnover rates were estimated by measuring biochemical markers of bone resorption (urinary deoxypyridinoline) and bone formation (bone-specific alkaline phosphatase). Forty females and 20 males of the same age were studied as a control group. Results. BMD was greater in women with spinal OA as compared to controls (P < 0.05). Also, males with OA had a non-significantly higher BMD than controls. The bone resorption markers were higher than normal values. However, they were lower than the control group. Similarly, the bone formation markers were lower as compared to the control group. Conclusion. Spinal OA is associated with higher BMD. This protective effect of spinal OA against osteoporosis may be mediated through decreased rate of bone turnover. Joint Bone Spine 2000 ; 67 : 521-7. © 2000 Éditions scientifiques et médicales Elsevier SAS bone mineral density / bone turnover / osteoarthirits
INTRODUCTION Although osteoarthritis (OA) and osteoporosis are both common conditions with high prevalence in the older age group, they are not purely due to simple aging and do not anthropometrically affect distinct populations. However, the coexistence of the two disorders in individual patients has been considered to be rare [1]. * Correspondence and reprints: Dr. Yasser El Miedany, 2 Italian Hospital Street Abbassia, Cairo, 11381, Egypt. E-mail address: [email protected] (Y.M. El Miedany).
Hence, an inverse relationship between OA and osteoporosis has been proposed. Evidence for such a relationship may give rise to etiologic clues; however, the studies that are presently available do not allow definite conclusions [2]. Studies that favour an inverse relationship [3-6] as well as studies in which there is no sufficient support for such a relationship [7, 8] have been published. An increase in measured bone mineral density (BMD) in a region affected by OA may result from artifactual error caused by the presence of osteophytes, joint space narrowing and sclerosis within the region of interest.
522
Y.M. El Miedany et al.
Alternatively, it may reflect a generalized effect of OA on the skeleton, or a combination of both. If OA has a generalized effect on the skeleton, the study of bone metabolism in patients with OA may be important in understanding the pathophysiology of osteoporosis. There has been little study of the underlying mechanism of the interaction between bone mass and OA. The development of noninvasive biochemical assays for markers of bone resorption and formation enable estimation of bone turnover in large cohorts of subjects. There have been few reports of biochemical markers in subjects with spinal OA [9]. Likewise, few studies have been done on men suffering from OA [6, 10]. In this work we performed a controlled study to determine whether the presence of spinal OA was associated with a generalized increase in BMD and if so, if there exists a sex difference. This work was done to study as well the underlying mechanism at the tissue level through assessment of biochemical markers of bone metabolism. SUBJECTS AND METHODS Subjects Forty postmenopausal Egyptian females and 20 Egyptian males diagnosed with osteoarthritic changes in the form of lumbar spondylosis and/or knee osteoarthritis were randomly collected from those attending the physical medicine outpatient clinic at Ain Shams University. Full medical history and thorough clinical examination were done for each patient with a stress on symptoms and signs of lumbar spondylosis, knee and hip OA. At initial evaluation the height and weight were measured and body mass index (BMI wt/height2) was calculated as an estimate of obesity. None of our patients had history of fractures. Patients who gave a history of renal affection were excluded from the study. Also, patients with a history of disease (e.g., diabetes mellitus, hyperthyroidism, endogenous or iatrogenic hypercorticism), malignancy or medications (steroids, thyroid hormone, diuretics, antihypertensive drugs containing thiazides, antimitotics, heparin and anticonvulsants) that may alter bone mineral metabolism were also excluded from the study. None of the postmenopausal women were on hormone replacement therapy or any other form of treatment for osteoporosis. None of our patients was a smoker or an alcoholic. None were bedridden.
Radiological assessment Each subject had anteroposterior and lateral radiographs of the cervical, thoracic and lumbar spine in addition to both knee and hip joints. Spinal OA was defined from the radiographs and scored according to the criteria of Lane et al. [11]. They were graded according to their score into: mild (≤ 2), moderate (3–5) and severe (≥ 6). Knee OA was defined and scored from the radiographs according to Spector et al. [12]. They were also graded into mild (≤ 3), moderate (4–7) and severe (≥ 8). The maximum of the left and right sides was used for the analysis. The radiographs were all reviewed by a single observer irrespective of bone mineral density result. Bone densitometry Bone mineral density (BMD) was measured using quantitative CT (Siemens) of L1–L3 lumbar vertebrae. The procedure was done according to the Cann and Gennant method [13]. Briefly, a phantom (QCT Bone Mineral Analysis System, San Francisco, CA) containing potassium phosphate standard was placed under the patient. Cursors were placed on the vertebra image to define a 100-mm thick transverse section through the center of L1, L2, L3. Cross-sectional images of each vertebra were obtained and used to position elliptical cursors in the trabecular area of each vertebral body. CT counts were then obtained for the selected vertebrae. Spinal measurements were referenced to a calibration curve obtained from the standard (Caucasian, European) and were expressed in milligrams per cubic centimeter. The results were expressed in a Z-score (number of standard deviation above or below the normal mean after comparison with age- and sexmatched normal control values supplied by the manufacturer). The T-score for females was calculated according to the equation given by the manufacturer: BMD (trabecular) – 159.1 (BMC of 35-year-old female) ± 27.6 (SD). The T-score for males was calculated according to the equation given by the manufacturer: BMD (trabecular) – 174.5 (BMC of 20-year-old male) ± 26.5 (SD). Biochemical markers of bone turnover The following assays were performed: serum Ca, phosphorous and alkaline phosphatase, in addition to bone resorption marker (urinary deoxypyridinoline) and
523
Altered bone mineral metabolism and OA Table I. Comparison of the osteoarthritic female and male patients with the control group in terms of the variables tested in this study. Variable Age (years) Years since menopause Parity Body mass index (weight/height2) Calcium (mmol/L) Phosphorus (mmol/L) Alkaline phosphatase (mmol/L) Corticol BMC (mg/cm3) Trabecular BMC (mg/cm3) Z-score T-score L-spine X-ray score Knee X-ray score Deoxypyridinoline (nmol/mmol) Bone alkaline phosphatase (U/L)
Female group
Male group
OA patients
Control group
P-value
OA patients
Control group
P-value
53.7 ± 5.1 6.2 ± 4.2 5.4 ± 3.4 33.0 ± 5.4 2.23 ± 0.14 3.74 ± 0.62 161.9 ± 59.9 241.0 ± 37.1 99.2 ± 29.6 –0.66 ± 0.89 –2.17 ± 1.15 4.8 ± 1.6 5.8 ± 1.6 6.74 ± 1.45 17.1 ± 5.9
52.9 ± 5.9 7.8 ± 5.6 5.3 ± 3.8 29.3 ± 3.4 2.19 ± 0.12 3.71 ± 0.48 157.0 ± 58.8 234.6 ± 42.1 81.0 ± 38.7 –1.29 ± 0.87 –2.83 ± 1.03 0.25 ± 0.44 0.30 ± 0.47 7.95 ± 1.97 22.1 ± 5.7
NS NS NS P < .001 NS NS NS NS P < .01 P < .01 P < .01 P < .01 P < .01 P < .01 P < .01
52.3 ± 6.4
51.5 ± 6.4
NS
bone formation marker (bone-specific alkaline phosphatase). Aliquots of early morning urine were stored frozen at –70° C until each assay was done in batches. Venous blood samples were obtained, centrifuged and stored in a similar manner. Deoxypyridinoline was measured in urine by estimation of Pyrilinks-D using competitive enzyme immunoassay method. Metrabiosystemst kits were used. Bone-specific alkaline phosphatase was assessed using the enzyme immunoassay method from Metrabiosystemst, which uses monoclonal anti-bone alkaline phosphatase antibody coated on micrometer strips to capture and quantitative bone alkaline phosphatase activity. Control group Forty Egyptian postmenopausal female and 20 Egyptian males of matched age were randomly collected and included as a control group. They had no history of any disease or medication known to affect bone metabolism. They were subjected to the same laboratory and radiological assessment. The study and aim of the work were explained to all subjects who participated in the study. All patients and controls provided their informed consent. Statistical analysis Data was analyzed using the Statistical Package for Social Sciences (SPSS) Software on an IBM PC. Different variables were described by the mean ± SD. Student’s t test was used to compare between cases and
31.6 ± 3.5 29.7 ± 2.7 2.16 ± 0.11 2.19 ± 0.09 3.69 ± 0.519 3.61 ± 0.386 129.4 ± 27.9 152.0 ± 42.2 318.3 ± 45.05 299.95 ± 58.10 153.5 ± 27.9 146.3 ± 31.3 0.87 ± 0.68 0.82 ± 0.66 –0.93 ± 1.113 –1.10 ± 1.074 4.7 ± 1.6 0.3 ± 0.47 5.6 ± 1.6 0.25 ± 0.44 4.41 ± 1.74 5.09 ± 1.89 16.2 ± 4.0 17.4 ± 3.6
NS NS NS NS NS NS NS NS P < .01 P < .01 NS NS
controls. The Mann-Whitney test was used to compare the variables that were not uniformly distributed. Spearman’s correlation coefficient was used to assess the correlation between trabecular and cortical bone mineral density, and knee and lumbar spine grading of osteoarthritis. Effect of obesity was assessed by Spearman’s correlation and multiple linear regression using Stata 6 for Windows 98. RESULTS Table I compares the female and male osteoarthritis patients with the control groups in terms of the variables tested in this study. It shows that osteoarthritic patients had higher bone mineral content compared to the control group. This difference was significant (P < 0.01) in the female group while it was not significant in the male group. Studying the bone markers revealed that bone resorption and formation markers were lower among osteoarthritic patients than the control group. Again, this difference was significant (P < 0.01) in the female patients while it was nonsignificant in the male group. Osteoarthritic patients had significantly higher BMI compared to the control group (P < 0.01 in the female group). Tables IIa and b are an analysis of variance showing the mean value and the correlation of trabecular and cortical bone mineral content among different grades of knee and lumbar spondylosis. It confirms the significant correlation between BMC and degree of osteoarthritis. Table III shows a comparison between obese osteoarthritic (28/40) (BMI > 30) versus obese non-
524
Y.M. El Miedany et al.
Table IIa. Analysis of variance showing the mean values and the correlation of the bone mineral content between the different grades of lumbar spondylosis (female group). Variables
Mild (Grade I)
Moderate (Grade II)
Severe (Grade III)
Significance
Cortical bone mineral content Trabecular bone mineral content
236.5 ± 40.4 92.7 ± 26.7
286.7 ± 55.1 121.0 ± 37.9
279.2 ± 53.8 139.6 ± 36.43
P < 0.001 P < 0.001
osteoarthritic females (18/40) regarding the Z-score as well as a comparison between non-obese osteoarthritic (BMI < 30) (12/40) versus non-obese nonosteoarthritic females (22/40). In both cases there was a significant difference (P < 0.03 and P < 0.02, respectively). Table IV shows that body weight as expressed by body mass index did not correlate significantly with the osteoarthritis X-ray score nor bone mineral density (r = 0.116 and 0.096, respectively). On the other hand, bone mineral density correlated significantly with the osteoarthritis X-ray score (r = 0.729, P < 0.001). Controlling for body mass index (table V), bone mineral density was still significantly correlated with osteoarthritis. Multiple regression analysis has been performed twice: in the first model the T-score was posed as the dependent variable while in the second bone mineral density was used as the dependent variable. In both models body mass index has been excluded. The regression coefficient of the osteoarthritis score was 0.681 and 0.636, respectively (table VI). DISCUSSION We found that elderly Egyptian women with moderate to severe radiographic features of OA had significantly (P < 0.02) higher BMD in the axial skeleton than did women with no or minimal radiographic findings of OA. BMD was increased in women who had not only spinal but also knee OA. The higher BMD of women
with OA was not explained by potential confounding variables such as age, weight, physical activity or use of medications that increase bone mineral density. Similarly, on the male side, although there was no significant difference, men with OA had also more bone mass than those without OA. Our findings are consistent with those of previous studies [6, 14-17] that showed that an increase in BMD was found not only in the spine but also at sites where there is no influence of osteophytes or subchondral sclerosis. In addition, our results showed significant correlation between BMD and X-ray scores of knee and lumbar spine osteoarthritic changes. This agrees with what has been reported by Nevitt et al. [18], who found that higher bone mineral density at both appendicular and axial sites was associated with the presence and size of hip osteophytes, while isolated narrowing without osteophytes was not associated with higher bone mineral density. Measurement of the biochemical markers of bone turnover enabled us to assess the mechanism of increased bone mineral density in patients with OA. In the female group bone resorption and formation markers were significantly (P < 0.01) lower in patients with OA compared to the control group. On the male side markers of bone resorption and formation were lower in the OA group compared to the control group although the differences were not significant. This might imply that patients with OA had a lower rate of bone turnover.
Table IIb. Analysis of variance showing the correlation of the bone mineral content between the different grades of knee OA (female group). Variables Trabecular bone mineral content Cortical bone mineral content
Mild (Grade I)
Moderate (Grade II)
Severe (Grade III)
Significance
91.2 ± 27.2 236.5 ± 41.6
112.9 ± 32.5 273.7 ± 51.9
144.5 ± 37.3 284.4 ± 57.0
P < 0.001 P < 0.001
Table III. Comparison of Z-score between obese, non-obese and control female groups.
Z-score
Obese OA (28/40)
Obese control (18/40)
P-value
Non-obese OA (12/40)
Non-obese control (22/40)
P-value
–0.33 ± 0.96
–1.07 ± 1.23
P < 0.03
–0.41 ± 1.23
–1.12 ± 2.25
P < 0.02
Altered bone mineral metabolism and OA Table IV. Spearman’s correlation showing the correlation between the variables included in the study. Variable
BMI
OA score
T-score
BMD
BMI
1.00
r = 0.116 NS 1.000
r = 0.151 NS r = 0.729 P < .001 1.00
r = 0.096 NS r = 0.720 P < .001 r = 0.951 P < .001 1.000
OA score T-score BMD
r = 0.116 NS r = 0.151 NS R = 0.96 NS
r = 0.729 P < .001 r = 0.720 P < .001
r = 0.951 P < .001
BMD: bone mineral density; OA: osteoathritis X-ray score.
Our results agree with what has been reported by Peel et al. and Gevers et al. [9, 14], who found in their studies a decrease in both resorption and formation markers in women with spinal OA. Generalized OA could be a good negative indicator for selecting patients at risk for osteoporosis. The generalized increase in bone mineral density in OA indicates that this disease might initially be a subchondral bone disease rather than a cartilage disorder [19]. There are several possible reasons for the inverse relationship between OA and osteoporosis. In our study as well as in others [3, 4, 6], the association exceeds the single joint and even correlates to the degree of OA. This suggests that generalized factors might play a role. Table V. Correlation between the variables tested in this study after controlling for body weight. Variable
BMD
OA score
T-score
BMI
1.000
r = 0.6315 P < .001 1.000
r = 0.9228 P < .001 r = 0.6756 P < .001 1.000
OA score T-score
r = 0.6315 P < .001 r = 0.9228 P < .001
r = 0.6756 P < .001
BMD: Bone Mineral Density; OA: Osteoarthritis X-ray score.
Table VI. Multiple regression analysis showing the correlation between osteoarthritis and bone mineral density in our group of patients. Dependent variable
Constant
Regression coefficient for OA score
P-value
T-score BMD
–4.344 37.681
0.681 0.636
P < .001 P < .001
BMD: Bone mineral density.
525
Elevated levels of insulin-like growth factors (IGF) I and II and transforming growth factor beta (TGFβ), the most abundant growth factors in human bone, have been shown to be increased in the matrix of the iliac crest in patients with OA. This finding is consistent with the hypothesis of a generally increased osteoblastic biosynthetic activity in these patients [20]. The anabolic properties of these growth factors might contribute to an increased stiffness of the subchondral bone and subsequent cartilage damage. Both IGFs and TGFβ are known to enhance osteoblastic replication and to promote the synthesis of different protein components, including procollagen I [21]. In addition, IGF I and II induce matrix opposition and decrease collagen degradation as well as the expression of interstitial collagenase, suggesting a role in the preservation of bone matrix. Besides obesity as a shared determinant [22], physical activity early in life may produce the association, since high levels of physical activity during youth have been shown to be of key importance in reaching peak bone mass [23] and may also lead to an increased risk of OA later in life [24, 25]. Gender difference in terms of the significant inverse relationship between OA and bone mineral density can be explained by the physiological changes that occur in woman at menopause leading to increased bone resorption, hence osteoporosis. The increased bone turnover might magnify the systemic effect of OA in females compared to males, who do have higher backgrounds of bone mass and do not suffer hormonal changes at that age. Moreover, in the work done by Felson et al. [26], the authors reported that women have a 1.5- to twofold higher incidence of knee OA compared to men. Quam et al. [27] reported that the age-adjusted frequency of total knee and hip arthroplasty was respectively 40% and 30% higher in women than in men. The association of obesity with the presence of OA has been examined through analysis of cross-sectional data and longitudinal cohort designs [26, 28-30]. Most studies used BMI as the measure of obesity. In the study done by Hochberg et al. [10], it was reported that body weight was associated with both definite and bilateral knee OA in both sexes, and this work supports contribution of mechanical as opposed to systemic factors to explain the association. Results of our study revealed that BMI was not significantly correlated with trabecular bone mineral content or T-score, nor with the X-ray score. However, the osteoarthritic group of patients studied had a significantly higher body mass index (P < 0.01) compared to the control group, whereas
526
Y.M. El Miedany et al.
after ruling out the role of body weight, osteoarthritis was significantly correlated with bone mineral density. Our results further extend the observation that body weight is associated with OA. Although the association of obesity with osteoporosis in patients with spinal OA is not frequently reported as compared to knee OA, our results support the notion that the systemic effect of obesity might have a protective effect against osteoporosis. But it is not the only factor involved. As conversion of androsterone to estrone in fat tissue represents the main source of estrogen in postmenopausal women [31] and estrogen-induced secretion of both IGFs and TGFβ by osteoblast-like cells is well documented, estrogen-mediated growth factor deposition in bone might be involved in the increased density and rigidity of the trabecular bone in OA, contributing to the pathogenesis of OA but also protecting from osteoporotic fracture [32]. However, in a recent study done on a cohort of 5,552 elderly women [33], subjects with OA did not have a significantly reduced risk of osteoporotic fracture. More interesting are the findings reported by Foss and Byers [34] and Roh et al. [35], who found that the negative association of OA with osteoporosis can only partially be accounted for by an absolute or relative estrogen excess. An increase in circulating growth hormone levels, both baseline growth hormone and growth hormone response to indirect stimuli, has been documented in OA compared to osteoporosis and may be of pathogenetic significance as growth hormone has a positive effect on bone and cartilage. Vertebral bone mineral density is usually measured in the anteroposterior plane though this method may give falsely high values. An increase in measured BMD in a region affected by OA may result from artifactual error caused by the presence of osteophytes, joint space narrowing and sclerosis within the region of interest. Aortic calcification is another potential confounder. Moreover, the anteroposterior BMD measurement using DEXA machines includes the posterior spinal element rich in cortical bone, while the main concern is the trabecular bone present in the vertebral body. In addition, the lower reproducibility and technical difficulties associated with lateral DEXA serve to limit the application of this method [36]. These were the reasons why we preferred to assess BMD using osteo CT rather than anteroposterior DEXA. Osteo CT has the advantage of selectively measuring the trabecular and cortical compartments of the vertebrae and has therefore been recognized as a sensitive method for assessment of bone
mineral density in patients with osteoporosis. Osteo CT also has the advantages of excluding the posterior compact bone elements of the vertebrae, any hypertrophic and degenerative change and/or vascular calcification, which commonly occur in aged people like our group of patients. Lastly, in the study done by Guglielmi et al. [37], the authors reported that although anteroposterior DEXA is a highly precise technique, with a 1% variation in short-term reproducibility studies, quantitative CT has been shown to help discriminate between healthy women and those with osteoporosis better than anteroposterior DEXA. In conclusion, the present study demonstrates that radiological OA is associated with a generalized increase in BMD and a decrease rate of bone turnover. These results are consistent with the hypothesis that OA has a protective effect against bone loss, mediated by a lower rate of bone turnover. This is more pronounced in women than in men. Alternatively, our results may suggest the existence of a subgroup of women with increased bone turnover, who are susceptible to the development of osteoporosis and who do not develop physiological degenerative changes of the skeleton with aging. REFERENCES 1 Foss MVL, Byers PD. Bone density, osteoarthrosis of the hip and fracture of the upper end of the femur. Ann Rheum Dis 1972 ; 31 : 259-64. 2 Hodon LD, Wright V, Smith NI. A. Bone mass in osteoarthritis. Ann Rheum Dis 1992 ; 51 : 823-8. 3 Gevers G, Dequeker J, Geusens P. Physical and histomorphological characteristics of iliac crest bone differ according to the grade of osteoarthritis at the hand. Bone 1989 ; 10 : 173-7. 4 Reid IR, Evans MC, Ames R, Wattie DJ. The influence of osteophytes and aortic calcification on spinal mineral density in postmenopausal women. J Clin Endocrinol Metab 1991 ; 72 : 1372-6. 5 Äström J, Beertema J. Reduced risk of hip fractures in the mothers of patients with osteoarthritis of the hip. J Bone Joint Surg Br 1992 ; 74 : 270-5. 6 Hannan MT, Anderson JJ, Zhang Y. Bone mineral density and knee osteoarthritis in elderly men and women. The Framingham Study. Arthritis Rheum 1993 ; 36 : 1671-6. 7 Price T, Hesp R, Mitchell R. Bone density in generalized osteoarthritis. J Rheumatol 1987 ; 14 : 560-5. 8 Orwall ES, Oviatt SK, Mann T. The impact of osteophytic and vascular calcifications on vertebral mineral density measurements in men. J Clin Endocrinol Metab 1989 ; 70 : 1202-6. 9 Peel NF, Barrington NA, Blumsohn A. Bone mineral density and bone turnover in spinal osteoarthritis. Ann Rheum Dis 1995 ; 54 : 867-71. 10 Hochberg MC, Lethbridge-Cejku M, Scott WW. Upper extremity bone mass and osteoarthritis of the knees. Data from the Baltimore Longitudinal Study of Aging. J Bone Miner Res 1995 ; 10 : 432-6. 11 Lane NE, Nevitt MC, Genant HK, Hochberg MC. Reliability
Altered bone mineral metabolism and OA
12 13 14
15
16
17
18 19 20 21 22 23 24
of new indices of radiographic osteoarthritis of the hand and hip and lumbar disc degeneration. J Rheumatol 1993 ; 20 : 1911-6. Spector TD, Cooper C, Cushanghan DJ, Dieppe P. A radiographic atlas of knee OA. London: Springer-Verlag; 1992. Cann CG, Gennant HK. Precise measurement of vertebral mineral content using computed tomography. J Comput Assist Tomogr 1980 ; 4 : 493-7. Gevers G, Dequeker J, Geusens P, Verstraeten A, Vanderschueren D. Serum bone Gla protein and other markers of bone mineral metabolism in postmenopausal osteoporosis and osteoarthritis. J Orthop Rheum 1988 ; 1 : 22-7. Hart DJ, Mootoosamy I, Doyle DV, Spector TD. The relationship between osteoarthritis and osteoporosis in the general population. The Chingford Study. Ann Rheum Dis 1994 ; 53 : 158. Marcelli C, Favier F, Kotzki PO. The relationship between osteoarthritis of the hands, bone mineral density, and osteoporotic fractures in elderly women. Osteoporos Int 1995 ; 5 : 382-6. Burger H, Van Daele PLA, Odding E. Association of radiographically evident osteoarthritis with higher bone mineral density: an increased bone loss with age. The Rotterdam Study. Arthritis Rheum 1996 ; 39 : 81-6. Nevitt MC, Lane NE, Scott JC. Radiographic osteoarthritis of the hip and bone mineral density. Arthritis Rheum 1995 ; 38 : 907-1002. Dequeker J, Mokassa L, Aerssens J. Bone density and OA. J Rheumatol 1995 ; 22 Suppl : 98. Dequeker J, Johnell O. The Medos Study Group. Osteoarthritis protects against femoral neck fractures. The MEDOS Study Experience. Bone 1993 ; 14 : 51. Coppens M, Dequeker J. Osteoarthritis, obesity and hormones. Current Opin Orthop 1991 ; 2 : 482. Dequeker J, Goris G, Uytterhoeven R. Osteoporosis and osteoarthritis: anthropometric distinctions. JAMA 1983 ; 249 : 1448-53. Welten DC, Kemper HCG, Post GB. Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J Bone Miner Res 1994 ; 9 : 1089-94. Heliövaara M, Mäkelä M, Impivaara O. Association of overweight trauma and workload with coxarthrosis. Acta Orthop Scand 1993 ; 64 : 513-7.
527
25 Cooper C, McAlindon T, Coggon D. Occupational activity and osteoarthritis of the knee. Ann Rheum Dis 1994 ; 53 : 90-5. 26 Felson DT, Anderson JJ, Naimerk A, Walker AM. Obesity and knee osteoarthritis. Ann Intem Med 1988 ; 109 : 18-23. 27 Quam JP, Michet CJ, Wilson MG, Rand JA. Total knee arthroplasty. A population-based study. Mayo Clin Proc 1991 ; 66 : 589-94. 28 Hartz AJ, Fischer ME, Brill G. The association of obesity with joint pain and osteoarthritis in the HANES data. J Chronic Dis 1986 ; 39 : 311-9. 29 Van Saase JLCM, Vandenbroucke JP, Van Romunde LKJ, Valkenburg HA. Osteoarthritis and obesity in the general population: a relationship calling for an explanation. J Rheumatol 1988 ; 15 : 1152-8. 30 Felson DT. The epidemiology of knee osteoarthritis: results from the Framingham Osteoarthritis Study. Semin Arthritis Rheum 1990 ; 20 Suppl 1 : 42-50. 31 Frumar A, Meldrum D, Geola F, Shamonki I, Tataryn I, Deftos L, et al. Relationship of fasting urinary calcium to circulating estrogen and body weight in postmenopausal women. J Clin Endocrinol Metab 1980 ; 50 : 70-5. 32 Dequeker J, Mohan S, Finkelman R, Aerssens J, Baylink D. Generalized osteoarthritis associated with increased insulin-like growth factor types I and II and transferring growth factor beta in cortical bone from the iliac crest. Arthritis Rheum 1993 ; 36 : 1702-8. 33 Arden NK, Nevitt MC, Lane NE, Gore R, Hochberg MC, Scott JC, et al. Osteoarthritis and risk of falls, rates of bone loss and osteoporotic fractures. Arthritis Rheum 1999 ; 42 : 137885. 34 Foss MV, Byers PD. Bone density, osteoarthritis of the hip and fracture of the upper end of the femur. Ann Rheum Dis 1972 ; 31 : 259-64. 35 Roh YS, Dequeker J, Mulier JC. Bone mass in osteoarthritis measured in vivo by photon absorption. J Bone Joint Surg 1974 ; 56 : 587-91. 36 Haddaway MJ, Davic MW, McCell IW. Bone mineral density in healthy normal women and reproducibility of measurements in spine and hip using DEXA. Br J Radiol 1992 ; 65 : 213-8. 37 Guglielmi G, Grimston SK, Frischerk C, Pacific R. Osteoporosis: diagnosis with lateral and posteroanterior DEXA compared with quantitative CT. Radiology 1994 ; 192 : 845-9.