522

ARTHRITIS & RHEUMATISM Vol. 43, No. 3, March 2000, pp 522530 2000, American College of Rheumatology

BONE MINERAL DENSITY AND FREQUENCY OF OSTEOPOROSIS IN FEMALE PATIENTS WITH RHEUMATOID ARTHRITIS
Results from 394 Patients in the Oslo County Rheumatoid Arthritis Register

GLENN HAUGEBERG, TILL UHLIG, JAN A. FALCH, JOHAN I. HALSE, and TORE K. KVIEN Objective. To examine the bone mineral density (BMD), frequency of osteoporosis, and risk factors for BMD reduction in a representative population of female rheumatoid arthritis (RA) patients ages 2070 years. Methods. BMD in the femoral neck, total hip, and spine L24 (anterior-posterior view) was measured in 394 RA patients recruited from a validated county RA register (completeness 85%) comprising 721 women ages 2070 years. BMD was measured with dual-energy x-ray absorptiometry, and age-specific values were compared with pooled values from a European/US population of healthy subjects free from earlier fractures, chronic diseases, and medications influencing bone metabolism. A multiple linear regression model was used to determine individual predictors of BMD. Results. No statistically significant differences were found in demographic, disease activity, disease severity, or health status parameters between the RA register patients in whom BMD was measured and the remaining register patients. Femoral neck BMD was significantly reduced by 4.2% in the age group 5059 years, and by 5.0% in those ages 6070 years. For BMD in the total hip, the significant reductions were 3.7%, 6.0%, and 8.5% in the age groups 4049 years, 5059 years, and 6070 years, respectively. No significant
Supported in part by grants from the Research Council of Norway, Lions Clubs International MD 104 Norway, the Norwegian Rheumatism Association, the Norwegian Women Public Health Association, Trygve Gythfeldt and Wifes Legacy, Grethe Harbitzs Legacy, and Marie and Else Mustads Legacy. Glenn Haugeberg, MD, Till Uhlig, MD, Tore K. Kvien, MD, PhD: Diakonhjemmet Hospital, Oslo, Norway; Jan A. Falch, MD, PhD: Aker Hospital, Oslo, Norway; Johan I. Halse, MD, PhD: Osteoporosis Clinic, Oslo, Norway. Address reprint requests to Glenn Haugeberg, MD, Oslo City Department of Rheumatology, Diakonhjemmet Hospital, PO Box 23, Vinderen, N-0319 Oslo, Norway. Submitted for publication April 28, 1999; accepted in revised form November 2, 1999.

reduction in spine L24 BMD was found. A 2-fold increased frequency of osteoporosis was observed in all 4 age groups of RA patients compared with the reference population, ranging from 0% to 28.6% in the femoral neck, 0% to 29.9% in the total hip, and 1.8% to 31.5% in the spine. Predictors of reduced BMD were as follows: at the femoral neck, older age, low body weight, current use of corticosteroids, greater physical disability (as measured by the modified Health Assessment Questionnaire [M-HAQ]), and presence of rheumatoid factor; at the total hip, older age, low weight, current use of corticosteroids, and higher M-HAQ disability score; and at the lumbar spine, older age, low weight, and current use of corticosteroids. Conclusion. Register-based prevalence data on BMD reduction in female RA patients ages 2070 years are presented for the first time in this report, which demonstrates a 2-fold increase in osteoporosis in this representative population. Osteoporosis is recognized as an extraarticular complication in rheumatoid arthritis (RA) (1), and an increased risk of hip (2) and vertebral (3,4) fractures has been demonstrated in patients with RA. Several studies in RA patients have demonstrated decreased bone mineral density (BMD) in the hip, spine, and total body, but the data are limited due to either the small number of patients studied or the selection of patients (3,57). One population-based study of BMD reduction in female RA patients has been conducted, but that study solely examined perimenopausal patients (8). It is likely that examinations of BMD reduction in a series of patients who are referred to a rheumatology unit will overestimate the problem. Our approach to overcome the problem of selection bias has been to establish a register of RA patients in the county of Oslo, Norway. This validated register, with a completeness of 85%, contains both

BONE DENSITY IN A FEMALE RA POPULATION

523

mild and severe cases that are suggested to be representative of the total RA population in the county (9). The guidelines for the prevention and treatment of primary osteoporosis and corticosteroid-induced osteoporosis are mainly based on bone density measurements (10,11). It has been recommended that bone density measurements be performed in subgroups of RA patients with risk factors for osteoporosis, e.g., more severe disease activity, older age, and worse functional status (12). No data exist on the magnitude of BMD reduction and the frequency of osteoporosis in a representative sample of female RA patients, emphasizing the need for such data. The primary objective of this study was to examine BMD in the hip and spine, the frequency of osteoporosis (defined as a T score of 2.5 SD below the mean value of controls) (13), and the frequency of reduced bone mass (defined as a Z score of 1 SD below the mean value of controls) in a population of female RA patients ages 2070 years who were recruited from the validated Oslo County RA register. Second, we wanted to examine the frequency of osteoporosis and reduced bone mass in subgroups of patients according to age, menopause status, prednisolone use, disease duration, level of disability, and rheumatoid factor (RF) status, and to examine possible predictors of reduced BMD. PATIENTS AND METHODS
Patients. The inclusion criteria in the present study were enrollment in the Oslo RA register (9,14), female sex, white race, and age between 20 and 70 years. At the onset of this study in 1996, 721 women (ages 2070 years, born in 1926 or later) were enrolled in the register. This register, validated to have a completeness of 85% (9), contains patients whose disease fulfills the American College of Rheumatology (formerly, the American Rheumatism Association) 1987 revised classification criteria for RA (15) and whose residential address is in the county of Oslo (14). Of the 721 female patients, 503 underwent a clinical examination. The patients were examined according to an extensive protocol comprising commonly used measures of disease activity, disease severity, and health status. Of the 503 women, 412 received BMD measurements. A total of 394 female patients (attendance rate 55%) were finally included in the study for further bone density analyses (18 patients were excluded due to either nonwhite ethnic background or age over 70 years at the time of BMD measurement). The BMD examination was performed, on average, 2 months after the clinical data had been collected. Demographic and clinical variables. The demographic and clinical characteristics of the patients, listed in Table 1, were obtained partly by self-report questionnaires and partly by interview and clinical examinations that were performed by 2 specially trained research nurses in cooperation with a

Table 1. Demographic variables and measures of disease activity, disease severity, and health status in 394 white female rheumatoid arthritis patients ages 2070 years* Variable Demographic variables Age, years Body weight, kg Height, cm BMI, kg/m2 Menopause, % Menopause age, years (n 260) Age at menarche, years Number of children Oophorectomy, % Current user of estradiol, % (n 376) Smoking status, % (n 384) Nonsmoker Previous smoker Current smoker Disease variables Disease duration, years Age at disease onset, years Rheumatoid-factor positive, % Global assessment score (VAS 0100 mm) (n 375) M-HAQ score (range 14) (n 385) Pain score (VAS 0100 mm) (n 385) Fatigue score (VAS 0100 mm) (n 385) Swollen joint count (28 joints) (n 386) Tender joint count (28 joints) (n 386) Deformed joint count (18 joints) (n 386) ESR, mm/hour (n 378) Ever user of DMARDs, % (n 389) Prednisolone, % (n 383) Never user Previous user Current user Value 54.8 11.6 65.6 11.7 165.5 6.3 24.0 4.2 66.0 47.9 5.5 13.6 1.4 1.7 1.3 4.3 28.5 33.0 30.5 36.5 13.0 9.4 40.5 12.8 52.0 28.7 24.1 1.61 0.50 36.1 21.9 44.9 27.8 7.8 6.1 6.8 6.4 2.2 3.7 21.6 16.5 84.8 36.3 23.2 40.5

* Except where otherwise indicated, values are the mean SD. BMI body mass index; VAS visual analog scale; M-HAQ modified Health Assessment Questionnaire; ESR erythrocyte sedimentation rate; DMARDs disease-modifying antirheumatic drugs. The number varies due to missing values.

rheumatologist (TU). The following variables were collected by a combination of self-report questionnaires and interview: menopause status, age at menarche, number of children, prior oophorectomy, current use of estradiol, smoking status, current or previous use of disease-modifying antirheumatic drugs (DMARDs), and history of prednisolone use. Pain and fatigue were measured on a visual analog scale (VAS; 0100 mm). Self-reported physical disability was evaluated by using the modified Health Assessment Questionnaire (M-HAQ; 8 items, score range 14, with higher scores indicating worse disability) (16). Investigators global assessment of disease activity was measured on a VAS (0100 mm). The clinical examination included a swollen joint count (28 joints), tender joint count (28 joints), and deformed joint count (18 joints). The erythrocyte sedimentation rate (ESR; in mm/hour) was also measured at the visit. Patients having an RF titer 64 that was measured on at least 1 occasion during the disease course were considered to be RF positive.

524

HAUGEBERG ET AL

BMD measurements. The standardized BMD measurements in the hip (total hip and femoral neck) and the lumbar spine L24 (anterior-posterior view) were performed by 3 trained technicians who each used the same dual-energy x-ray absorptiometry (DEXA) equipment (Lunar Expert, Madison, WI). In 14 patients, no hip measurement was performed due to bilateral hip replacements, and in 14 patients, only the right hip was measured. BMD measurement in the spine was not performed in 1 patient. To ensure uniformity of the results, all data were reanalyzed by the experienced technicians after all data had been collected. The spine phantom precision error, calculated from daily measurements of the phantom as a coefficient of variation (CV), was 0.9% for the whole measurement period (15 months). The in vivo reproducibility of BMD measurements was assessed on the basis of duplicate measurements in 31 healthy female hospital workers (mean age 56.1 years [SD 4.9], range 5066 years). The CV was 1.5% for measurements at the femoral neck, 1.5% at the total hip, and 2.2% at the lumbar spine L24. The interobserver variation between the 3 trained technicians performing the measurements was assessed by analyzing the results from 30 randomly selected RA patients. The readings of each technician were compared in a pairwise manner. For BMD measurements at the femoral neck, the precision varied from 0.5% to 0.8%, for the total hip, from 0.4% to 0.5%, and for the spine L24, from 0.7% to 1.4%. Reference population. The BMD measurements in the RA population were compared with pooled data from a combined European/US reference population, which comprised different data on healthy subjects from Europe and the US. This reference population of women (all between the ages of 20 and 69 years) contained 9,505 who had received hip measurements and 10,281 with spine measurements (1725). The variation in the mean BMD values among the geographic sites in this pooled population was 1.3% (SD/mean) (17). The reference database was based on measures in ambulatory healthy subjects who were free from earlier fractures, chronic diseases, and medications influencing bone metabolism (e.g., corticosteroids, anticonvulsants, and thyroxine) (17). One small study (18) that was included in the database did not provide detailed information about the healthy subjects. The BMD data for the pooled European/US reference population were provided by Lunar (Madison, WI), including the data for T-score estimations and equations for estimating the BMD values adjusted for age and weight, which was required for Z-score estimations. For T-score estimations, the following mean BMD values for healthy young adults (ages 2039 years) were used: femoral neck 0.98 gm/cm2 (SD 0.12), total hip 1.00 gm/cm2 (SD 0.12), and spine L24 1.20 gm/cm2 (SD 0.12). For Z-score estimations, the following equations were used to compute age- and weight-adjusted BMD: in the femoral neck, for ages 2045 years, 0.98 (0.002 [age 30]) (0.003 [weight 65]), for ages 4665 years, 0.95 (0.007 [age 45]) (0.003 [weight 65]), and for ages 66100 years, 0.81 (0.004 [age 65]) (0.003 [weight 65]); in the total hip, for ages 2045 years, 1.00 (0.001 [age 30]) (0.003 [weight 65]), for ages 4665 years, 0.985 (0.006 [age 45]) (0.003 [weight 65]), and for ages 66100 years, 0.865 (0.0055 [age 65]) (0.003 [weight 65]); and in the spine L24, for ages 2045 years, 1.20 (0.000 [age 30]) (0.004 [weight 65]), for ages 4665

years, 1.20 (0.011 [age 45]) (0.004 [weight 65]), and for ages 66100 years, 0.98 (0.002 [age 65]) (0.004 [weight 65]). The proportion of patients having osteoporosis in the pooled European/US reference population was estimated by using the mean BMD for each decade and developing centiles and T scores based on the mean and SD values for young adults, assuming a normal distribution. To ensure the validity of the data from the pooled European/US reference population, we compared it with data (all obtained with Lunar machines) from populations in Sweden (18), Finland (19), and Norway (26). There were no substantial differences between the pooled European/US data and the Scandinavian reference data for the femoral neck and spine L24 (18,19,26), since the 95% confidence intervals (95% CI) of the mean BMD values were overlapping in each age decade or were close to the BMD values for the pooled European/US reference population. Ethics and legal aspects. The local ethics committee approved this study. The Data Inspectorate had previously approved the register of RA patients in Oslo. Statistical analysis. The RA patients as a whole and subgroups according to age, menopause, use of prednisolone, disease duration, level of physical disability (M-HAQ score), and RF status were compared with the pooled European/US reference population. Mean values and 95% CI were calculated by decades of age. The bone mass was expressed as the BMD (in gm/cm2), Z score, and T score. Osteoporosis was defined according to the conventional World Health Organization (WHO) definition (a T score of 2.5 SD below the mean value of controls) (13), while reduced bone mass was defined as a Z score of 1 SD below the mean value of controls (age- and weight-adjusted T score). The statistical analyses were carried out using the SPSS program, version 8.0 (SPSS, Chicago, IL). The 95% CI for the proportions of patients having a T score 2.5 SD below the mean of controls and a Z score 1 SD below the mean of controls were calculated using the equation for binomial distribution (27). Group comparison was performed by Students 2-tailed unpaired t-test (for continuous variables) or Pearsons chi-square test (for categorical variables). We used multiple linear regression analyses to explore the relationship between BMD and the demographic and disease-related risk factors listed in Table 1. P values less than or equal to 0.05 or no overlap between the 95% CI of comparative groups were considered statistically significant.

RESULTS Patient characteristics. Table 1 shows the clinical characteristics of the 394 female RA register patients finally included in the study. The group of RA register patients in whom BMD was measured (n 412) did not differ significantly from the RA register patients in whom BMD was not measured (n 309) with regard to age (mean 54.9 years versus 55.4 years; P 0.56), disease duration (mean 12.9 years versus 13.2 years; P 0.62), and RF positivity (51.6% versus 46.9%; P 0.44).

BONE DENSITY IN A FEMALE RA POPULATION

525

Table 2. Spine and hip BMD reduction by age decades in 394 female RA patients compared with the European/US reference population*
Spine L24 Age (years), group 2029 RA Ref. 3039 RA Ref. 4049 RA Ref. 5059 RA Ref. 6070 RA Ref. No. with spine measurements 11 707 45 1,006 67 2,302 108 3,589 162 2,677 No. with femur measurements 11 779 44 1,164 64 2,448 107 2,928 154 2,186 Mean gm/cm2 (95% CI) 1.210 (1.1571.263) 1.196 (1.1871.205) 1.247 (1.1951.299) 1.210 (1.2031.217) 1.203 (1.1621.244) 1.180 (1.1751.185) 1.097 (1.0621.132) 1.102 (1.0981.106) 1.014 (0.9831.045) 1.015 (1.0101.020) BMD reduction in RA, % None None None 0.5 0.1 Femoral neck Mean gm/cm2 (95% CI) 0.927 (0.8660.988) 0.996 (0.9881.004) 0.987 (0.9431.031) 0.972 (0.9650.979) 0.926 (0.8950.957) 0.948 (0.9430.953) 0.850 (0.8240.876) 0.887 (0.8830.891) 0.772 (0.7500.794) 0.813 (0.8080.818) BMD reduction in RA, % 6.9 None 2.3 4.2 5.0 Total hip Mean gm/cm2 (95% CI) 0.952 (0.8821.022) 1.006 (0.9981.014) 0.993 (0.9471.039) 0.997 (0.9901.004) 0.946 (0.9150.977) 0.982 (0.9770.987) 0.879 (0.8520.906) 0.935 (0.9310.939) 0.797 (0.7720.822) 0.871 (0.8660.876) BMD reduction in RA, % 5.4 0.4 3.7 6.0 8.5

pop. pop. pop. pop. pop.

* BMD bone mineral density; RA rheumatoid arthritis; 95% CI 95% confidence interval; Ref. pop. reference population. No overlap of 95% CI for patients and the European/US reference population.

Among the RA register patients who underwent a clinical examination (n 503), no statistically significant differences were found between the patients whose BMD was measured and the patients whose BMD was not measured with regard to the following clinical characteristics: age (mean 54.7 years versus 53.4 years; P 0.32), disease duration (mean 12.7 years versus 12.1 years; P 0.54), RF positivity (51.1% versus 51.6%; P 0.66), investigators global assessment of disease activity (mean VAS score 29.7 mm versus 28.9 mm; P 0.76), M-HAQ score (mean 1.61 versus 1.62; P 0.87), VAS score for pain (mean 36.4 mm versus 34.1 mm; P 0.38), VAS score for fatigue (mean 45.2 mm versus 42.8 mm; P 0.45), swollen joint count (28 joints) (mean 7.8 versus 7.6; P 0.70), tender joint count (28 joints) (mean 6.4 versus 7.2; P 0.98), deformed joint count (18 joints) (mean 2.2 versus 2.0; P 0.70), ESR (mean 22 mm/hour versus 24 mm/hour; P 0.13), ever use of DMARDs (84.7% versus 78.6%; P 0.32), and the use of prednisolone (never use 36.2% versus 44.7%, previous use 22.7% versus 21.3%, and current use 41.1% versus 34.0%; all P 0.29). Thus, it is considered that the final sample examined was fairly representative of the underlying population with regard to demographic and disease characteristics. Changes in BMD. Femoral neck BMD was significantly reduced by 4.2% in the age group 5059 years, and by 5.0% in the age group 6070 years (Table 2). The reduction in BMD in the total hip for the age groups 4049 years, 5059 years, and 6070 years was 3.7%, 6.0%, and 8.5%, respectively. No significant reduction in BMD in the spine L24 was found in the different age groups (Table 2).

BMD was generally lower in the RF-positive patients, comprising 204 women (52%), compared with the RFnegative patients. In the RF-positive patients, the reduction in BMD in the femoral neck was 6.1% for the age group 5059 years and 7.5% for the age group 6070 years. In the total hip, the reduction in BMD was 4.3%, 7.8%, and 10.7% for the age groups 4049 years, 5059 years, and 6070 years, respectively. In RF-negative patients, a significant BMD reduction was found only in the total hip, by 4.6% in the age group 5059 years and by 4.8% in the age group 6070 years (data not shown). Overall frequency of osteoporosis and reduced bone mass. The overall frequency of osteoporosis (T score of 2.5 SD below the mean of controls) in the female RA population was 14.7% (95% CI 11.118.3) in the femoral neck, 14.7% (95% CI 11.118.3) in the total hip, and 16.8% (95% CI 13.120.5) in the spine L24 (Table 3). Even though BMD was not reduced in the spine, on a group level, compared with the reference population, the proportion of patients having osteoporosis was similar to or higher than the frequency of osteoporosis found in the femoral neck and total hip. The proportion of all patients having reduced bone mass (Z score of 1 SD below the mean of controls; expected to be 16% according to normal distribution in the reference population) was 27.6% (95% CI 23.132.1) in the femoral neck, 31.6% (95% CI 26.936.3) in the total hip, and 19.6% (95% CI 15.7 23.5) in the spine L24 (Table 3). Frequency of osteoporosis and reduced bone mass in subgroups of patients. The highest proportion of patients having osteoporosis or reduced bone mass

526

Table 3. Frequency of osteoporosis and reduced bone mass in the spine L24 and hip (femoral neck and total hip) among the whole RA population (n 394) and patient subgroups* Osteoporosis Spine L24 16.8 (13.120.5) 1.8 (1.75.3) 3.0 (0.0026.2) 11.1 (5.217.0) 31.5 (22.740.3) 0 3.1 (07.3) 9.3 (3.814.8) 28.6 (21.535.7) 2.3 (0.34.9) 21.2 (16.126.3) 6.5 (2.410.6) 10.6 (4.117.1) 24.7 (17.731.7) 10.5 (3.617.4) 8.9 (3.314.5) 17.1 (10.626.6) 23.0 (13.432.6) 9.2 (4.413.6) 17.7 (11.024.4) 19.0 (10.627.4) 9.8 (5.414.2) 18.9 (13.324.5) 7.9 (1.814.0) 9.9 (4.115.7) 15.5 (9.321.7) 27.0 (16.937.1) 8.0 (3.812.2) 16.1 (9.622.6) 23.8 (14.732.9) 9.2 (4.913.5) 18.9 (13.324.5) 5.8 (1.99.7) 12.9 (5.820.0) 24.0 (17.130.9) 1.5 (0.63.6) 21.6 (16.526.7) 15.7 (9.521.9) 21.6 (16.626.6) 11.5 (6.216.8) 11.2 (4.617.8) 31.2 (23.938.5) 26.0 (16.235.8) 14.9 (8.021.8) 18.8 (12.225.4) 20.7 (11.929.5) 15.2 (9.720.7) 23.1 (15.930.3) 20.0 (11.728.3) 15.3 (10.020.6) 24.6 (18.530.7) 0 1.6 (0.024.7) 8.4 (3.113.7) 29.9 (22.737.1) 12.5 (3.821.2) 13.4 (5.221.6) 26.9 (18.535.3) 19.8 (13.725.9) 6.0 (2.010.0) 22.4 (17.327.5) 8.6 (3.913.3) 10.1 (3.816.4) 26.6 (19.633.6) 10.4 (3.617.2) 12.9 (6.419.4) 17.3 (10.923.7) 26.8 (17.236.4) 12.2 (7.217.2) 18.5 (11.825.2) 18.9 (10.827.0) 14.1 (9.019.2) 18.3 (12.823.8) 14.7 (11.118.3) 14.7 (11.118.3) 19.6 (15.723.5) Femoral neck Total hip Spine L24 Femoral neck 27.6 (23.132.1) 16.4 (6.626.2) 26.6 (15.837.4) 29.9 (21.238.6) 30.5 (23.237.8) 18.5 (11.825.2) 32.4 (26.638.2) 17.4 (11.723.7) 23.5 (14.532.5) 41.1 (33.149.1) 27.6 (17.537.7) 29.7 (20.838.6) 24.0 (16.631.4) 31.1 (20.641.6) 22.1 (15.728.5) 30.6 (22.538.7) 34.5 (24.344.7) 20.8 (14.826.8) 34.6 (27.741.5) Reduced bone mass Total hip 31.6 (26.936.3) 20.0 (9.430.6) 28.1 (17.139.1) 29.0 (20.437.6) 39.0 (31.346.7) 20.8 (13.827.8) 37.2 (31.243.2) 18.8 (12.325.3) 27.1 (17.736.5) 47.3 (39.255.4) 31.6 (21.142.1) 29.7 (20.838.6) 29.5 (21.637.4) 37.8 (26.848.8) 25.2 (18.531.9) 30.6 (22.538.7) 45.2 (34.655.8) 23.7 (17.430.0) 37.8 (30.844.8)

No. with spine/femur measurements 393/380 56/55 67/64 108/107 162/154 134/130 259/250 139/138 89/85 154/146 77/76 101/101 133/129 82/74 164/163 130/124 90/84 177/173 191/185

All RA patients Age group, years 2039 4049 5059 6070 Menopause Premenopausal Postmenopausal Corticosteroids Never user Previous user Current user Disease duration, years 04 59 1019 20 M-HAQ score 11.49 1.501.99 2.00 and above RF status Negative Positive

* Osteoporosis was defined as a T score 2.5 SD below the mean value of controls, and reduced bone mass as a Z score 1 SD below the mean value of controls. Pearsons chi-square test was used to compare rheumatoid arthritis (RA) patients with the pooled European/US reference population having reduced bone mass. Values are the percentage of patients (95% confidence intervals). M-HAQ modified Health Assessment Questionnaire; RF rheumatoid factor. The summations of patients in the different subgroups diverge from the total number of spine and hip measurements due to missing values. P 0.001. P 0.05.

HAUGEBERG ET AL

BONE DENSITY IN A FEMALE RA POPULATION

527

Figure 1. Frequency of spine and hip osteoporosis in female rheumatoid arthritis (RA) patients compared with the estimated frequency in the pooled European/US reference population in different age groups. Values above the bars are percentages. P 0.05; P 0.001 (by Pearsons chi-square test).

was, in general, found in subgroups of patients who were older in age, were postmenopausal, were currently receiving prednisolone, had longer disease duration, had higher disability, and were positive for RF. As expected, the highest frequency of osteoporosis was found in the age group 6070 years for all 3 measurement sites (Table 3 and Figure 1). The percentage having osteoporosis in this age group was 28.6% (95% CI 21.535.7) in the femoral neck, 29.9% (95% CI 22.737.1) in the total hip, and 31.5% (95% CI 22.740.3) in the spine L24. The proportion of patients having osteoporosis compared with the European/US reference population is illustrated in Figure 1 by age groups. The highest frequency of reduced bone mass in RA patients (Z score of 1 SD below the mean of controls) was found among current users of prednisolone. In the femoral neck, the frequency of reduced bone mass among current users of

prednisolone was 41.1% (95% CI 33.149.1), in the total hip, 47.3% (95% CI 39.255.4), and in the spine L24, 31.2% (95% CI 23.938.5) (Table 3). The frequencies of osteoporosis and/or reduced bone mass in the subgroups of RA patients according to age, menopause status, prednisolone use, disease duration, level of disability (M-HAQ score), and RF status are shown in Table 3. Risk factors for BMD reduction as determined by multivariate analyses. The demographic and diseaserelated variables listed in Table 1 were statistically tested in a linear regression model as independent predictors of BMD reduction at different measurement sites. In the final model, older age, low body weight, current use of corticosteroids, higher M-HAQ score, and RF positivity were identified as statistically significant independent predictors of BMD reduction in the femoral neck. Older age, low body weight, current use of corticosteroids, and higher M-HAQ scores predicted total hip BMD, and older age, low body weight, and current use of corticosteroids were found as statistically significant predictors of BMD reduction at the spine L24 (Table 4). DISCUSSION As far as we know, this is the first study examining BMD reduction, frequency of osteoporosis, and reduced bone mass in a large sample of female RA patients who were representative of the entire RA population in an area. The major findings in our study were the demonstration of a moderate overall reduction in BMD at the femoral neck and total hip, and no BMD reduction in the spine L24. The overall frequency of osteoporosis was doubled in the female RA population, and reduced BMD was associated with high age, low body weight, and current use of corticosteroids, in

Table 4. Multiple regression analysis of bone mineral density at different sites of measurement (dependent variables) and demographic and disease variables (independent variables)* Spine L24 Beta Age (years) Body weight (kg) Current use of corticosteroids (no/yes) Disease duration (years) M-HAQ score (range 14) RF positive (no/yes) R2 6.872 5.5203 9.0272 1.1393 2.3512 1.6382
3

Femoral neck P 0.001 0.001 0.001 0.28 0.24 0.38 32 Beta 5.923 5.0943 6.2412 4.5624 3.7082 2.8692
3

Total hip P 0.001 0.001 0.001 0.53 0.008 0.026 43 Beta 5.095 5.8173 7.6562 3.2994 4.5152 2.3392
3

SE 0.001 0.001 0.019 0.001 0.020 0.018

SE 0.001 0.001 0.014 0.001 0.014 0.013

SE 0.001 0.001 0.014 0.001 0.015 0.013

P 0.001 0.001 0.001 0.67 0.002 0.10 43

* Beta values are standardized coefficients. The R2 is the total variance explained in the model. SE standard error; M-HAQ modified Health Assessment Questionnaire; RF rheumatoid factor.

528

HAUGEBERG ET AL

addition to physical disability and RF positivity at some measurement sites. Several studies have shown an association between RA and bone density reduction in the spine and hip (3,57). Depending on the menopause status and use of corticosteroids among the women studied, the BMD reduction in the femoral neck ranged from 2.4% to 17.6%, and in the spine, changes in BMD ranged from no reduction to 10.7% (3,57). Most of these studies have limitations due either to the small number of patients or to the selection of patients examined. One previous population-based study examined only perimenopausal women and found a BMD reduction of 5.5% in the lumbar spine and 8.7% in the femoral neck (8). The crude BMD reduction in our study was not impressive, but the large SD reflects major interindividual variations. This explains the more than 2-fold increase in frequency of RA patients being classified as osteoporotic or as having reduced bone mass (Table 3), in spite of having similar or only slightly decreased mean BMD values compared with those in the reference population. This supports the view that some individual patients in the RA population have an increased risk for developing osteoporosis and reduced bone mass (12). The lack of BMD reduction in the spine in our study could be due to increased spine osteoarthritis and/or vertebral fractures, increasing with age and disease duration, and the spine results should therefore be interpreted with caution. This would probably explain why the proportion of patients with a Z score 1 SD below the mean of controls was lower for the age group 6070 years than for the age group 5059 years, and why the highest proportion of patients with a Z score 1 SD below the mean of controls was found in patients with the shortest disease duration (04 years). One other factor to consider in the interpretation of the results is the absence of raw weight data in the historical reference population. Weight-adjusted Z scores in the spine L24 were not significantly different from those of the controls (Table 3), whereas the T scores showed a statistically significant 2-fold increased prevalence of osteoporosis. A body weight difference between patients and controls may, if present, contribute to the different results regarding T scores and Z scores in the spine L24. In contrast, consistent results regarding increased prevalences of T and Z scores were found for both hip site measurements. The frequency of osteoporosis (as measured with Lunar DEXA equipment) was higher in our RA population than in populations from England (28) and the US (29). In the age group 5059 years, the frequency of

femoral neck osteoporosis in our RA patients was 9.3% compared with 2.0% in the English population (28) and 5.3% in the US population (29), and estimated to be 4.4% in our pooled European/US reference population. For the age group 6070 years, the frequencies were 28.6%, 5.9%, 16.7%, and 13.8%, respectively, in each of these populations. The validity of our findings depends on the representativeness of the RA patients whose BMD was measured compared with the underlying RA population, and on the BMD comparability of the European/US reference population with the background population in Norway. It is likely that patients examined in our study were representative of the whole validated RA register population for 2 reasons. First, investigators in our group have previously shown that the Oslo RA register comprises 85% of the RA patients in the county (9). Second, no obvious differences could be demonstrated with regard to demographic, disease activity, disease severity, or health status characteristics between the register patients whose BMD was measured and those whose BMD was not measured. Appropriate reference values are critical for the clinical evaluation of normality (30,31). The BMD values for the European/US reference population showed, however, an overall good agreement with those of populations from Finland (19), and also with the data from 2 smaller studies, 1 conducted in Sweden (18) and 1 conducted in the county of Oslo (Norway) (26). When femoral neck BMD data from the Third National Health and Nutrition Examination Survey (QDR; Hologic, Waltham, MA) (30) are converted to DPX-equivalent (Lunar) values (29) using the conversion equations of Genant et al (32), these are also comparable with the values in the European/US reference population. The WHO definition of osteoporosis, defined as a T score 2.5 SD below the mean of controls (13), is relative and dependent on the mean and SD of the reference population. Even minor differences in the mean BMD and SD between reference data can have important effects on patient classification (3336). This was demonstrated by Ahmed et al (37) using 2 reference values with differences of 5% in the mean spine BMD (0.994 versus 1.047 gm/cm2) and 10% in the SD (0.12 gm/cm2 versus 0.11 gm/cm2) among young, healthy individuals. In that study, 48% versus 64% of subjects were classified as normal, 30% versus 37% as osteopenic, and 6% versus 15% as osteoporotic on the basis of the 2 reference values (37). Fractures are the only known clinically relevant consequence of osteoporosis. In RA, an increased risk of

BONE DENSITY IN A FEMALE RA POPULATION

529

fractures has been demonstrated in both the hip (2) and spine (3,4). In our study using multivariate analysis, the use of corticosteroids, higher M-HAQ scores, and RF status, in addition to low body weight and high age, were found to be independent predictors of reduced BMD, depending on the measurement site. The highest frequency of osteoporosis and reduced bone mass (Z score 1 SD below the mean of controls) was found among current users of corticosteroids, patients being most disabled, or those being RF positive, indicating that these patients were at the highest risk for fractures. The use of corticosteroids and disease-related measurements in RA patients should be taken into account when making the decision as to whether a DEXA measurement should be performed in the assessment of an individual fracture risk in the RA patient. However, the extent to which BMD reduction can be directly translated into a risk of fracture in RA patients is still an open question. In conclusion, the 2-fold increase in osteoporosis demonstrated in this study suggests the magnitude of the osteoporosis problem in the whole female RA population. This result, taken together with the association between BMD and disease and demographic risk factors, gives clues to which patients should be referred for BMD measurement. The results are also of importance in health care planning for patients with RA. ACKNOWLEDGMENTS
We thank Lunar Corporation (President Richard B. Mazess and Howard S. Barden) for advice and help in providing us with the normative data for the pooled European/US reference population, and Petter Mowinckel for statistical advice. We gratefully appreciate our technicians Ingerid Mu ller, Sidsel Arnkvrn, and Espen Haavardsholm for expert technical assistance.

REFERENCES
1. Deodhar AA, Woolf AD. Bone mass measurement and bone metabolism in rheumatoid arthritis: a review. Br J Rheumatol 1996;35:30922. 2. Hooyman JR, Melton LJ III, Nelson AM, OFallon WM, Riggs BL. Fractures after rheumatoid arthritis: a population-based study. Arthritis Rheum 1984;27:135361. 3. Peel NF, Moore DJ, Barrington NA, Bax DE, Eastell R. Risk of vertebral fracture and relationship to bone mineral density in steroid treated rheumatoid arthritis. Ann Rheum Dis 1995;54: 8016. 4. Spector TD, Hall GM, McCloskey EV, Kanis JA. Risk of vertebral fracture in women with rheumatoid arthritis. BMJ 1993;306:558. 5. Martin JC, Munro R, Campbell MK, Reid DM. Effects of disease and corticosteroids on appendicular bone mass in postmenopausal women with rheumatoid arthritis: comparison with axial measurements. Br J Rheumatol 1997;36:439.

6. Sambrook PN, Eisman JA, Yeates MG, Pocock NA, Eberl S, Champion GD. Osteoporosis in rheumatoid arthritis: safety of low dose corticosteroids. Ann Rheum Dis 1986;45:9503. 7. Hall GM, Spector TD, Griffin AJ, Jawad ASM, Hall ML, Doyle DV. The effect of rheumatoid arthritis and steroid therapy on bone density in postmenopausal women. Arthritis Rheum 1993; 36:15106. 8. Kroger H, Honkanen R, Saarikoski S, Alhava E. Decreased axial bone mineral density in perimenopausal women with rheumatoid arthritisa population based study. Ann Rheum Dis 1994;53: 1823. 9. Kvien TK, Glenns A, Knudsrd OG, Smedstad LM, Mowinckel P, Frre . The prevalence and severity of rheumatoid arthritis in Oslo: results from a county register and a population survey. Scand J Rheumatol 1997;26:4128. 10. Consensus Development Conference. Prophylaxis and treatment of osteoporosis. Am J Med 1991;90:10710. 11. American College of Rheumatology Task Force on Osteoporosis Guidelines. Recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis. Arthritis Rheum 1996;39: 1791801. 12. Lems WF, Dijkmans BA. Should we look for osteoporosis in patients with rheumatoid arthritis? Ann Rheum Dis 1998;57: 3257. 13. Kanis JA, Melton LJ III, Christiansen C, Johnston CC, Khaltaev N. The diagnosis of osteoporosis. J Bone Miner Res 1994;9: 113741. 14. Uhlig T, Kvien TK, Glenns A, Smedstad LM, Frre . The incidence and severity of rheumatoid arthritis: results from a county register in Oslo, Norway. J Rheumatol 1998;25:107884. 15. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:31524. 16. Pincus T, Summey JA, Soraci SA Jr, Wallston KA, Hummon NP. Assessment of patient satisfaction in activities of daily living using a modified Stanford Health Assessment Questionnaire. Arthritis Rheum 1983;26:134653. 17. Lunar Corporation. Operators manual, expert-XL, software version 1.7. Madison, WI: Lunar Corporation; 1998. 18. Karlsson MK, Gardsell P, Johnell O, Nilsson BE, Akesson K, Obrant KJ. Bone mineral normative data in Malmo, Sweden: comparison with reference data and hip fracture incidence in other ethnic groups. Acta Orthop Scand 1993;64:16872. 19. Kroger H, Heikkinen J, Laitinen K, Kotaniemi A. Dual-energy x-ray absorptiometry in normal women: a cross-sectional study of 717 Finnish volunteers. Osteoporos Int 1992;2:13540. 20. Laitinen K, Valimaki M, Keto P. Bone mineral density measured by dual-energy x-ray absorptiometry in healthy Finnish women. Calcif Tissue Int 1991;48:22431. 21. Wetzel R, Pfandl S, Bodenburg R, Puhl W. Knochendichte Referenzwerte von deutschen FrauenUntersuchung der LWS mit dem Lunar DPX-Densitometer. Osteologie 1996;5:7181. 22. Burger H, van Daele PL, Algra D, van den Ouweland FA, Grobbee DE, Hofman A, et al. The association between age and bone mineral density in men and women aged 55 years and over: the Rotterdam Study. Bone Miner 1994;25:113. 23. Krall EA, Dawson-Hughes B, Hirst K, Gallagher JC, Sherman SS, Dalsky G. Bone mineral density and biochemical markers of bone turnover in healthy elderly men and women. J Gerontol 1997;52: 617. 24. Mazess RB, Barden HS, Eberle RW, Denton MD. Age changes of spine density in posterior-anterior and lateral projections in normal women. Calcif Tissue Int 1995;56:2015. 25. Truscott JG, Simpson D, Fordham JN. Compilation of national bone densitometry reference data. In: Ring EFS, Elvins DM, Ghalla AK, editors. Current research in osteoporosis and bone

530

HAUGEBERG ET AL

26. 27. 28.

29. 30. 31.

32.

mineral measurement IV. London: The British Institute of Radiology; 1996. p. 778. Falch JA, Meyer HE. Bone mineral density measured by dual x-ray absorptiometry: a reference material from Oslo. Tidsskr Nor Laegeforen 1996;116:2299302. Altman DG. Comparing groupscategorical data. In: Altman DG, editor. Practical statistics for medical research. London: Chapman & Hall; 1997. p. 230. Petley GW, Cotton AM, Murrills AJ, Taylor PA, Cooper C, Cawley MI, et al. Reference ranges of bone mineral density for women in southern England: the impact of local data on the diagnosis of osteoporosis. Br J Radiol 1996;69:65560. Mazess RB, Barden H. Bone density of the spine and femur in adult white females. Calcif Tissue Int 1999;65:919. Looker AC, Johnston CCJ, Wahner HW, Dunn WL, Calvo MS, Harris TB, et al. Prevalence of low femoral bone density in older U.S. women from NHANES III. J Bone Miner Res 1995;10:796802. Simmons A, Simpson DE, ODoherty MJ, Barrington S, Coakley AJ. The effects of standardization and reference values on patient classification for spine and femur dual-energy x-ray absorptiometry. Osteoporos Int 1997;7:2006. Genant HK, Grampp S, Gluer CC, Faulkner KG, Jergas M,

33.

34.

35.

36.

37.

Engelke K, et al. Universal standardization for dual x-ray absorptiometry: patient and phantom cross-calibration results. J Bone Miner Res 1994;9:150314. Laskey MA, Crisp AJ, Cole TJ, Compston JE. Comparison of the effect of different reference data on Lunar DPX and Hologic QDR-1000 dual-energy x-ray absorptiometers. Br J Radiol 1992; 65:11249. Pocock NA, Sambrook PN, Nguyen T, Kelly P, Freund J, Eisman JA. Assessment of spinal and femoral bone density by dual x-ray absorptiometry: comparison of lunar and hologic instruments. J Bone Miner Res 1992;7:10814. Simmons A, Barrington S, ODoherty MJ, Coakley AJ. Dual energy x-ray absorptiometry normal reference range use within the UK and the effect of different normal ranges on the assessment of bone density. Br J Radiol 1995;68:9039. Simmons A, ODoherty MJ, Barrington SF, Coakley AJ. A survey of dual-energy x-ray absorptiometry (DEXA) normal reference ranges used within the UK and their effect on patient classification. Nucl Med Commun 1995;16:104153. Ahmed AI, Blake GM, Rymer JM, Fogelman I. Screening for osteopenia and osteoporosis: do the accepted normal ranges lead to overdiagnosis? Osteoporos Int 1997;7:4328.