• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arthritis Rheum. Author manuscript; available in PMC Jul 5, 2009.
Published in final edited form as:
PMCID: PMC2705769
NIHMSID: NIHMS107127

Bone Density, Structure, and Strength in Juvenile Idiopathic Arthritis

Importance of Disease Severity and Muscle Deficits

Abstract

Objective

To identify determinants of musculoskeletal deficits (muscle cross-sectional area [mCSA], trabecular volumetric bone mineral density [vBMD], and cortical bone strength [section modulus]) in patients with juvenile idiopathic arthritis (JIA) and to determine if cortical bone strength is appropriately adapted to muscle forces.

Methods

Peripheral quantitative computed tomography (pQCT) of the tibia was performed in 101 patients with JIA (79% female; 24 with oligoarticular JIA, 40 with polyarticular JIA, 18 with systemic JIA, and 19 with spondylarthritis [SpA]) and 830 healthy control subjects; all were ages 5–22 years. Outcomes of pQCT were expressed as sex- and race-specific Z scores. Multivariable linear regression models assessed mCSA and bone status in JIA patients compared with controls and identified factors associated with musculoskeletal deficits in JIA.

Results

The median duration of JIA was 40 months; 29% of the JIA patients had active arthritis, and 28% had received glucocorticoid therapy during the previous year. Compared with the controls, the mCSA and section modulus Z scores were significantly lower in patients with polyarticular JIA and those with SpA. Trabecular vBMD Z scores were significantly lower in patients with polyarticular JIA, those with systemic JIA, and those with SpA. Significant predictors of musculoskeletal deficits included active arthritis in the previous 6 months (mCSA), temporomandibular joint disease (mCSA and section modulus), functional disability (mCSA and vBMD), short stature (vBMD), infliximab exposure (vBMD), and JIA duration (section modulus). The section modulus was significantly reduced relative to mCSA in patients with JIA after adjustment for age and limb length.

Conclusion

Marked deficits in vBMD and bone strength occur in JIA in association with severe and longstanding disease. Contrary to the findings of previous studies, bone deficits were greater than expected relative to the mCSA, which illustrates the importance of adjusting for age and bone length.

Juvenile idiopathic arthritis (JIA) is the most common pediatric rheumatic disease (1). Risk factors for low bone mass in JIA include chronic inflammation, delayed pubertal maturation, malnutrition, muscle weakness, physical inactivity, and glucocorticoid therapy (2). The impact of childhood arthritis on bone health may be immediate, resulting in childhood fractures, or it may be delayed, resulting from suboptimal peak bone mass attainment or persistent disease activity. In a population-based study, we demonstrated that childhood arthritis was associated with an elevated risk of fracture in patients older than 10 years (3). Recent investigations using peripheral quantitative computed tomography (pQCT) in JIA demonstrated significant deficits in trabecular volumetric bone mineral density (vBMD), cortical bone mass, and estimates of bone strength in the appendicular skeleton (46). Adjustment for low muscle cross-sectional area (mCSA) eliminated the bone deficits in JIA (5), or it paradoxically resulted in the appearance of greater bone mass (4) in children with JIA as compared with healthy controls.

The mechanostat theory states that bone adapts to the mechanical or muscle forces to which it is subjected in order to keep the strain on the bone at a constant set point (7). Investigators hypothesized that since muscle and bone deficits in JIA were highly correlated, interventions to improve muscle mass and strength will optimize bone health (46). In the study outlining an approach to assessing the “functional muscle–bone unit” in children, muscle mass was first assessed relative to height or age, and bone was then assessed relative to muscle mass (8). This strategy did not account for the strong, independent relationship between bone length and both bone and muscle mass (9). Without adjustment for bone length, the assessment of the functional muscle–bone unit may be confounded in JIA, a disease often complicated by muscle wasting and both systemic and local disturbances of limb growth.

In the previous studies of bone health in JIA (46), the study patients had high levels of disease activity. With the advent of tumor necrosis factor α (TNFα) inhibitor therapy, greater numbers of children with JIA achieve clinical remission (10) and, potentially, normalization of muscle and bone deficits. Therefore, we used pQCT of the tibia to assess muscle and bone deficits in a population of JIA patients with a spectrum of disease activity as compared with controls, to identify JIA disease characteristics and therapies associated with muscle and bone mass deficits, and to demonstrate that the interpretation of the functional muscle–bone unit in children and adolescents with JIA is dependent on the magnitude of muscle deficits and is sensitive to limb-length adjustment.

PATIENTS AND METHODS

Study subjects

All patients with JIA diagnosed according to the revised criteria of the International League of Associations for Rheumatology (ILAR) (11) who were between the ages of 5 years and 21 years and were being treated at Children’s Hospital of Philadelphia were eligible for enrollment in the study. Children were excluded if they had conditions or medication exposures known to affect growth or bone mass and not associated with JIA.

Healthy control subjects were recruited from urban and suburban pediatric practices and through local advertisements. Exclusion criteria were chronic diseases or medication exposures known to affect bone mass (e.g., renal or hepatic disease, endocrine disorders, or a history of glucocorticoid or anticonvulsant therapy). The control sample used in this study was not matched to the JIA patients; rather, the sample was developed to provide robust estimates of bone mass and body composition variability according to sex, age, body size, pubertal maturation, and race strata.

The protocol was approved by the Institutional Review Board at Children’s Hospital of Philadelphia. Informed consent was obtained from the young adults and from the parents or guardians of those who were younger than 18 years, from whom assent for study was obtained.

Disease characteristics, disease activity, and medication exposures

Patients with JIA were classified according to the ILAR revised criteria (11). JIA characteristics and rheumatoid factor (RF) status were documented by interview and by medical record abstraction. At the study visit, functional status was assessed using the Childhood Health Assessment Questionnaire (C-HAQ) (12), disease activity was assessed using the total joint count (13), and patients were categorized as having active or inactive arthritis and as being in clinical remission while on or off medication (14). The erythrocyte sedimentation rate (mm/hour) and serum C-reactive protein (mg/dl) concentration were also measured.

Glucocorticoid exposure during the previous year was abstracted from the clinic, hospital, and emergency department records, and these exposures were summed to yield a cumulative dose (mg/kg) over the previous 12 months. Previous intraarticular injections of glucocorticoids and prescriptions for disease-modifying antirheumatic drugs and biologic agents were noted.

Assessments of anthropometric features and pubertal development

Weight and height were measured using a digital scale to the nearest 0.1 kg (Scaltronix, White Plains, NY) and a stadiometer to the nearest 0.1 cm (Holtain, Crosswell, UK), respectively. Age- and sex-specific Z scores (i.e., standard deviation scores) for height and body mass index (BMI; kg/m2) were calculated using year 2000 growth chart data from the Centers for Disease Control and Prevention (15). Pubertal stage was assessed according to the method of Tanner (16), using a self-assessment questionnaire (17).

Peripheral QCT assessment of bone and muscle parameters

We performed pQCT scans (Stratec XCT-2000; Orthometrix, White Plains, NY) of the left tibia in all study subjects. Peripheral QCT provides accurate estimates of vBMD, bone structural parameters, and mCSA (18). The tibia was measured because the impact of impaired biomechanical stimulation and bone modeling may be more pronounced at a weight-bearing site, and cortical thickness is greater in the tibia than in the radius and is therefore less subject to partial volume effects (19). The Stratec XCT-2000 scanner has a 12-detector unit that uses a slice thickness of 2.3 mm. The standard voxel size is 0.4 mm.

Trabecular vBMD (mg/cm3) was assessed in the distal tibia (3% site; 3% of the tibia length proximal to the distal physis). Cortical bone was assessed at the site of maximal cortical thickness (38% site). Cortical measures included cortical vBMD (mg/cm3) and the structural parameters of periosteal and endosteal circumference (mm), and section modulus (mm3). The section modulus is a composite measure of bone strength in bending and in torsion (20), and it is proportional to the (Rp 4 − Re 4)/Rp, where Rp represents the periosteal radius and Re represents the endosteal radius (21). The mCSA (mm3), a proxy measure of muscle forces, was assessed at the site of maximum mCSA (66% site).

Quality control was monitored daily using the hydroxyapatite phantom provided by the manufacturer. The short-term precision of paired pQCT measurements in 60 healthy pediatric volunteers ranged from 0.5% to 1.6%.

Statistical analysis

Analyses were conducted using Stata 9.2 software (StataCorp, College Station, TX). Two-sided tests of hypotheses were used; P values less than 0.05 were considered statistically significant. Differences in means were assessed using Student’s t-test for normally distributed data and using Wilcoxon’s rank sum test for non–normally distributed data. Differences in proportions were assessed using chi-square test or Fisher’s exact test where appropriate.

Calculation of Z scores

Bone outcomes and mCSA data were converted to Z scores to facilitate comparisons between JIA patients and healthy controls and to compare outcomes based on JIA characteristics and medication exposures. Growth was characterized by sex- and race-specific increases in bone mass and body composition (22); therefore, all Z scores are sex- and race-specific. The mCSA and cortical bone structure measures were assessed relative to tibia length (all natural log–transformed). Trabecular and cortical vBMD were assessed relative to age. Z scores for cortical measurements and the mCSA were generated using the half-normal method, which adjusts for heteroscedasticity (23). The LMS technique was used to generate Z scores for trabecular vBMD given the nonlinear relationship between trabecular vBMD and age (24).

Analysis of muscle and bone measurements

Data from the JIA patients were displayed graphically, relative to the data from the healthy control subjects (Figure 1). Next, pQCT Z scores in all JIA patients and in each JIA subtype were compared with those in all control subjects using multivariable linear regression. All Z scores generated for tibia length were subsequently adjusted for age, adding polynomial terms where appropriate to improve the model fit. Disease characteristics and medication exposures associated with differences in the mean mCSA, section modulus, or trabecular vBMD were identified.

Figure 1
Distribution of data from the bone and muscle measurements in patients with juvenile idiopathic arthritis (JIA) (solid circles) and in healthy control subjects (shaded circles). The data are presented for non-black females, since 74% of the JIA patients ...

Assessment of the functional muscle–bone unit

We hypothesized that previous studies of the functional muscle–bone unit in children with JIA were confounded by age and tibia length, especially in patients with low muscle mass relative to tibia length. For example, in the assessment of section modulus in a JIA patient with low mCSA relative to tibia length, the control subject with comparable mCSA will be younger and will have a shorter tibia. Since the section modulus is highly correlated with the tibia length (r = 0.92, P < 0.001), the control subject with a shorter tibia length will have a lower section modulus, resulting in the appearance of a greater section modulus relative to mCSA in the patient with JIA. Therefore, adjustment for age and tibia length is necessary to adjust for group differences in these parameters.

In order to test this hypothesis, we developed models for section modulus relative to muscle area and examined the effect of adjusting the models for age and tibia length among the JIA patients with muscle deficits (mCSA Z score of −1 or less), and among the JIA patients without muscle deficits (mCSA Z score greater than −1). Four multivariable linear regression models comparing patients with JIA and controls were constructed. In model 1A, the section modulus was adjusted for sex, race, and mCSA in all controls and in JIA patients whose mCSA Z score was −1 or less. In model 1B, age and tibia length were added to the covariates in model 1A. The same sequence was followed for models 2A and 2B, except that the JIA patients were limited to those whose mCSA Z score was greater than −1.

RESULTS

Characteristics of the study patients

The demographic and anthropometric characteristics of the patients with JIA and the healthy control subjects are shown in Table 1. Of the 101 JIA patients, 24 were categorized as having oligoarticular JIA (15 with persistent and 9 with extended disease), 40 as having polyarticular JIA (10 RF positive), 18 as having systemic JIA, and 19 as having spondylarthritis (SpA; 11 with enthesitis-related arthritis, 4 with psoriatic arthritis, and 4 with undifferentiated arthritis). A greater proportion of the JIA patients were female and a smaller proportion of the patients were black, as compared with the healthy controls. Significantly greater pubertal delay as well as significantly lower Z scores for height (Z = −0.20, P = 0.002), lower Z scores for BMI (Z = −0.08, P = 0.01), and shorter tibia length (P = 0.02) were noted only in the polyarticular JIA group as compared with the control group. The BMI Z scores were significantly higher in patients with systemic JIA (Z = 0.86, P = 0.04) than in the healthy controls, which is likely due to the greater glucocorticoid exposure in this JIA subset.

Table 1
Demographic and anthropometric characteristics of JIA patients and healthy controls*

Disease characteristics

The features of JIA according to disease subtype are shown in Table 2. Active arthritis was present in 29% of JIA patients, in whom the median joint count was 2 (range 1–39). However, 57% of patients had evidence of active disease within the 6 months preceding the study visit. The median C-HAQ score of 0.1 (range 0–2.4) reflected the high proportion of patients who were free of active arthritis at the time of the study visit. In the previous 12 months, methotrexate therapy was prescribed for 75% of the JIA patients and TNFα inhibitors for 28%. Of the 9 patients receiving infliximab at the study visit, all but 3 had failed to respond to etanercept treatment. One patient with systemic JIA was receiving cyclosporin A, and 1 patient with SpA was receiving sulfasalazine. High-dose glucocorticoid exposure was limited to those who had systemic JIA.

Table 2
Disease characteristics, disease activity, and medication exposures in all JIA patients and in JIA patient subgroups*

Muscle and bone outcomes in JIA

Z scores for the pQCT parameters are shown in Table 3. The analyses were repeated including Tanner stage as a covariate, and the results were unchanged (data not shown). Compared with the controls, the mCSA Z scores were lower in children with polyarticular JIA and those with SpA, normal in those with oligoarticular JIA, and higher in those with systemic JIA. After adjustment for the BMI Z score, the augmented mCSA Z score in patients with systemic JIA was no longer significant (0.32 [95% CI −0.08, 0.72], P = 0.12). However, after BMI Z score adjustment, the mCSA Z score deficits persisted in patients with polyarticular JIA (−0.36 [95% CI −0.62, −0.09], P = 0.009) and were marginally significant in patients with SpA (−0.35 [95% CI −0.73, 0.03], P = 0.08).

Table 3
Mean Z scores and 95% CIs for bone density, bone geometry, bone strength, and muscle cross-sectional area in all JIA patients and in JIA patient subgroups*

Similar to the mCSA results, the assessment of cortical structural measures revealed subtype-specific deficits. Z score deficits in periosteal circumference in the polyarticular JIA and SpA subgroups in combination with normal endosteal circumference indicated significant cortical thinning. Since the section modulus relates the periosteal dimensions to the endosteal dimensions, the section modulus Z scores were low only in the polyarticular JIA and SpA subgroups.

Volumetric BMD of the cortical and trabecular compartments were assessed separately. Cortical vBMD Z scores at the 38% site were significantly higher among patients with polyarticular JIA, those with systemic JIA, and those with SpA as compared with the value in the controls. In contrast, trabecular vBMD scores at the 3% site were low in patients with polyarticular JIA, those with systemic JIA, and those with SpA as compared with the value in the controls.

Predictors of muscle and bone outcomes in JIA

Having established group differences in muscle and bone measurements in JIA patients, we identified predictors of the mean mCSA, section modulus, and trabecular vBMD Z scores using linear regression within the JIA patients. Significant (P < 0.05) predictors of the mCSA Z score included a deleterious effect of temporomandibular joint disease (Z = −0.59 versus Z = 0.01, P = 0.03), disease activity within the previous 6 months (Z = −0.41 versus Z = 0.13, P < 0.05), and a C-HAQ score greater than 0.5 (Z = −0.67 versus Z = −0.04, P = 0.04). Compared with systemic JIA, the mCSA Z scores were lower in those with polyarticular JIA and those with SpA (P < 0.01). A lower section modulus Z score was associated with temporomandibular joint involvement (Z = −1.03 versus Z = −0.36, P = 0.02), older age (Z = −0.10 per year of age, P = 0.005), and longer disease duration (Z = −0.10 per year of disease, P = 0.01). A lower trabecular vBMD Z score was associated with a C-HAQ score greater than 0.5 (Z = −1.15 versus Z = −0.43, P = 0.01), a height Z score less than −1.0 (Z = −1.39 versus Z = −0.47, P < 0.01), and current use of infliximab (Z = −1.46 versus Z = −0.51, P = 0.02). Although only marginally significant (P = 0.05–0.08), temporomandibular joint involvement and RF positivity were associated with lower trabecular vBMD Z scores. Muscle and bone outcomes were not associated with arthritis present in the left leg or with methotrexate, etanercept, or cumulative glucocorticoid use over the previous year.

Evaluation of the functional muscle–bone unit

We performed stratified analyses to determine whether deficits in cortical bone structural measures in patients with JIA were explained by deficits in mCSA, that is, whether bone was appropriately adapted to muscle (Table 4). Model 1A compared the ln(section modulus) in JIA patients with an mCSA Z score of −1 or less with all of the control subjects, adjusting for ln(mCSA), sex, and race. Consistent with our hypothesis, the JIA patients with muscle deficits were significantly older (3.0 years; P < 0.001), and their tibia length was significantly longer (3.7 cm; P < 0.001) on average as compared with the controls with comparable mCSA values. The β coefficient for JIA in model 1A indicated that the ln(section modulus) in JIA patients was 8% higher than expected for the ln(mCSA) (P = 0.04).

Table 4
Sensitivity of the section modulus assessment to the presence of low mCSA in JIA patients, with adjustment for age and tibia length*

In model 1B, the addition of age (with polynomial terms) and ln(tibia length) demonstrated a significant bone deficit in JIA (P < 0.001); JIA patients had a 10% lower section modulus compared with healthy children of the same sex, race, age, tibia length, and mCSA. Therefore, adjustment for age and tibia length converted the seemingly significant excess in the section modulus relative to the mCSA in JIA patients to a significant deficit. Model 1B (R2 = 0.92) explained a greater proportion of the variability in the section modulus than did model 1A (R2 = 0.86) and confirmed that tibia length was a strong and independent predictor of the section modulus.

Models 2A and 2B compared JIA patients with normal mCSA with all of the control subjects. For a given ln mCSA value, patients with JIA in these models were of similar age, but the tibia length was significantly shorter (1.5 cm; P < 0.001) on average as compared with the controls. In model 2A, adjustment of the ln(section modulus) for the ln(mCSA), sex, and race demonstrated a significant JIA deficit of 11% (P < 0.001). Since tibia length was shorter in JIA for a given mCSA, a JIA patient in model 2A was compared with a control subject with a longer tibia. Therefore, the section modulus deficit in the JIA patients in model 2A may be overestimated. With the addition of age and ln(tibia length) in model 2B, the section modulus deficit in JIA was attenuated to 4% but was still statistically significant (P = 0.009). Model 2B (R2 = 0.93) explained a greater proportion of the variability in the section modulus than did model 2A (R2 = 0.87).

Consistent with these models, Figure 2 demonstrates that at all levels of the mCSA Z scores, the section modulus Z scores were lower in JIA patients. The use of Z scores relative to tibia length (i.e., section modulus relative to tibia length and mCSA relative to tibia length) for this comparison adjusts for the confounding effects of tibia length. In JIA patients, those with normal mCSA Z scores had section modulus Z scores that were low (Z = −0.27 [95% CI −0.50, −0.05], P = 0.02) as compared with the controls. There was a significant interaction between JIA and low mCSA (P < 0.001), indicating that JIA patients with low mCSA values had even lower section modulus Z scores than were expected (section modulus Z = −2.19 [95% CI −2.56, −1.82], P < 0.001), compared with controls that had similarly low mCSA values (section modulus Z = −1.14 [95% CI −1.34, −0.95], P < 0.001).

Figure 2
Section modulus for the tibia length Z score relative to the muscle cross-sectional area (CSA)–for–tibia length Z score in patients with juvenile idiopathic arthritis (solid circles) and in healthy control subjects (shaded circles). The ...

DISCUSSION

In this study of a large cohort of children with JIA, we demonstrated that JIA is complicated by deficits in mCSA, cortical section modulus, and trabecular vBMD in the tibia. Significant mCSA and cortical section modulus deficits were limited to children with polyarticular JIA and those with SpA. The lower section modulus, a proxy for bone strength, was due to a smaller periosteal circumference in conjunction with normal endosteal dimensions, resulting in cortical bone thinning. Trabecular vBMD deficits were marginal in oligoarticular JIA, but in all other subtypes, the deficits were marked. The lower mCSA was associated with functional disability, disease activity, and temporomandibular joint disease, while low bone measurements were associated with lower mCSA, functional disability, greater disease severity (temporomandibular joint involvement, infliximab use), older age, and longer disease duration. We showed that section modulus deficits were greater than expected for the lower mCSA seen in selected JIA subtypes and that this determination was highly sensitive to adjustment for age and tibia length.

Our data demonstrated that patients with JIA have significant bone strength deficits relative to mCSA. In contrast, previous pQCT studies in JIA suggested that bone status was normal or even greater than expected relative to muscle. Roth et al (4) reported that forearm mCSA relative to height was 0.8–2.4 SD lower in JIA patients, depending on the disease subtype, as compared with controls. Within those study patients, cortical bone mineral content (BMC) relative to mCSA was 2.3–4.5 SD greater than what would be expected. Statistically, we were able to duplicate this finding in the present study (model 1A), demonstrating that the comparison of JIA patients with low mCSA to healthy controls without adjustment for group differences in age and tibia length resulted in the appearance that bone mass was paradoxically greater than expected relative to mCSA. This is because a JIA patient with muscle deficits is older and has a longer limb than a healthy control subject with a similar mCSA.

Bechtold et al (6) performed pQCT of the radius in a severely affected population of children with JIA and marked short stature. Cortical BMC–for-mCSA Z scores were markedly low in systemic JIA (Z = −1.7), the subtype in which mCSA Z scores were highest. Therefore, as in our analysis (model 2A), we would expect that BMC would appear to be lower than normal relative to muscle mass. However, the deficit may have been exaggerated, since the limb length was likely shorter in JIA patients as compared with healthy controls. Finally, in their study of a population of patients with JIA and more modest mCSA deficits of 0.33–0.87 SD, Felin et al (5) reported that the BMC in the midshaft of the tibia was appropriate relative to the mCSA. Thus, the prevailing conclusion in previous studies that the cortical bone response relative to the mCSA is normal or even greater than normal may be explained by the presence of low muscle mass relative to tibia length in the JIA patients and the lack of adjustment for group differences in age and limb length.

The structural alterations we observed likely have important implications for fracture risk in JIA. A case–control study of healthy girls with forearm fractures demonstrated that the bone cross-sectional area of the distal radius was 8% lower in children with fractures (25). Similarly, a lower metacarpal width and metacarpal index (cortical thickness/metacarpal width) were independently associated with wrist and forearm fractures in children (26). Our findings are consistent with our large population-based cohort study in 1,939 patients with childhood arthritis and 207,072 matched controls. We demonstrated that fracture rates in the children and adolescents with arthritis were markedly elevated (incidence rate ratio 1.75–3.13) (3).

Despite the alterations in bone geometry, cortical vBMD was found to be significantly higher in the patients with JIA. This finding is consistent with the only other pQCT study in JIA that was performed on the tibia (5), but is not consistent with studies performed on the radius (4,6), where partial volume effects may result in an underestimation of the vBMD (19). At the edges of a bone, partially filled voxels may spuriously lower the vBMD. When the cortical thickness is greater, as would be expected in the tibia versus the radius, partial volume effects are less pronounced. Therefore, since cortical bone was thinner in JIA, the significantly greater cortical vBMD we observed may even be an underestimation of the true increase in cortical vBMD in JIA. The lower bone turnover observed in JIA (27) may result in reduced intracortical porosity and greater secondary mineralization (28), leading to higher cortical density.

Interestingly, high cortical density may not improve bone quality. In a comparison of 2 inbred mouse models, smaller cross-sectional dimensions of the femur in the A/J strain compared with the C57BL/6J strain was associated with a significantly greater ash content. However, these thinner, harder bones were more brittle and more susceptible to damage accumulation (29). A followup study in cadaveric human tibias demonstrated similar findings (20). These data help explain why slender bones are more susceptible to stress fractures (30). Similarly, composite pQCT cortical bone measurements that incorporate structure and density are less accurate in the prediction of radius failure load than are measurements assessing only structure (31).

Osteoblast and osteoclast development are likely to be crucial in the pathogenesis of bone alterations in JIA and may be negatively influenced by inflammation, glucocorticoids, and methotrexate. TNFα and interleukin-6 (IL-6) are key regulators of bone formation and resorption (3234). The deleterious effects of TNFα on bone formation may be important in inflammatory diseases and in health. In adults with rheumatoid arthritis, TNFα inhibition may prevent bone loss at the hip and spine, even in the absence of a clinical response (35). In healthy mice null for TNFα or its p55 receptor, peak bone mass was significantly higher secondary to greater bone formation (36). Young IL-6–transgenic mice have diminished longitudinal growth, thin cortical bone, and reduced trabecular thickness and connectivity (37). These alterations were found to be due to lower osteoblast numbers and activity and increased osteoclastogenesis. The bone loss in glucocorticoid-induced osteoporosis is mediated by transient increases in bone resorption (38) and sustained reductions in bone formation due to inhibition of osteoblastogenesis and increased osteoblast apoptosis (39). Although we did not find an effect of recent glucocorticoid exposure in our cross-sectional study, the deleterious effect of glucocorticoids on bone may be accentuated in the context of a pediatric inflammatory disease (40). Furthermore, the beneficial effects of glucocorticoids on disease activity in JIA may have countered the adverse effects on bone accrual. Methotrexate strongly inhibits osteoblast differentiation when administered as cancer chemotherapy (41). However, the lower doses used to treat chronic arthritis may not adversely influence bone formation (42,43).

This study has limitations to consider. First, we presented tibia pQCT data only. Our findings may represent local arthritis effects that may not be relevant to unaffected skeletal regions. However, the deficits were marked in children with polyarticular JIA and those with SpA, who are likely to have widespread arthritis and, consequently, deleterious alterations in bone geometry and bone density at other skeletal sites (4,6). Second, we did not perform QCT of the spine. Since glucocorticoid-treated children with JIA are at risk of developing vertebral compression fractures (44), accurate characterization of vertebral vBMD and size is needed. Interestingly, we recently used paired posteroanterior and lateral spine scans to estimated vBMD in the spine and concluded that traditional areal BMD as determined by dual x-ray absorptiometry may systematically underestimate bone deficits in JIA (45). This finding suggested that the bone defect in JIA is systemic. Third, bone age was not assessed. Since wrist arthritis is associated with accelerated maturation of the carpal bones in growing children, adjustment for bone age using wrist radiographs may not be reliable in JIA. Finally, our cross-sectional study is limited in the ability to relate chronic exposures to a single measure of musculoskeletal outcomes. Longitudinal studies are needed to more accurately characterize these relationships.

In conclusion, children with JIA are at risk for deleterious alterations in cortical bone strength and trabecular bone density, placing them at greater risk of fracture. Unlike other studies, we demonstrated that the pronounced bone deficits in JIA are greater than would be expected for their reductions in muscle cross-sectional area. Thus, bone alterations in JIA likely represent a mixed defect of bone development and, in some cases, lower muscle forces. Future clinical trials assessing both bone-active therapies and mechanical loading interventions are required.

ACKNOWLEDGMENTS

We greatly appreciate the dedication and enthusiasm of the children and their families who participated in this study. Special thanks to Anne Close, and Drs. Terri H. Finkel, David D. Sherry, and Randy Q. Cron for their assistance with the study.

Supported by the Research and Education Foundation of the American College of Rheumatology and by the Clinical Translational Research Center at the University of Pennsylvania (grant UL1-RR-024134 from the National Center for Research Resources). Dr. Burnham’s work was supported by the NIH (grant 5K23-RR-021969).

REFERENCES

1. Bowyer S, Roettcher P. for the Pediatric Rheumatology Database Research Group. Pediatric rheumatology clinic populations in the United States: results of a 3 year survey. J Rheumatol. 1996;23:1968–1974. [PubMed]
2. Burnham JM, Leonard MB. Bone disease in pediatric rheumatologic disorders. Curr Rheumatol Rep. 2004;6:70–78. [PubMed]
3. Burnham JM, Shults J, Weinstein R, Lewis JD, Leonard MB. Childhood onset arthritis is associated with an elevated risk of fracture: a population-based study using the General Practice Research Database. Ann Rheum Dis. 2006:1074–1079. [PMC free article] [PubMed]
4. Roth J, Palm C, Scheunemann I, Ranke MB, Schweizer R, Dannecker GE. Musculoskeletal abnormalities of the forearm in patients with juvenile idiopathic arthritis relate mainly to bone geometry. Arthritis Rheum. 2004;50:1277–1285. [PubMed]
5. Felin EM, Prahalad S, Askew EW, Moyer-Mileur LJ. Musculoskeletal abnormalities of the tibia in juvenile rheumatoid arthritis. Arthritis Rheum. 2007;56:984–994. [PubMed]
6. Bechtold S, Ripperger P, Dalla Pozza R, Schmidt H, Hafner R, Schwarz HP. Musculoskeletal and functional muscle-bone analysis in children with rheumatic disease using peripheral quantitative computed tomography. Osteoporos Int. 2005;16:757–763. [PubMed]
7. Frost HM. Changing concepts in skeletal physiology: Wolff’s law, the mechanostat, and the “Utah paradigm.” Am J Hum Biol. 1998;10:599–605.
8. Schoenau E, Neu CM, Beck B, Manz F, Rauch F. Bone mineral content per muscle cross-sectional area as an index of the functional muscle-bone unit. J Bone Miner Res. 2002;17:1095–1101. [PubMed]
9. Macdonald H, Kontulainen S, Petit M, Janssen P, McKay H. Bone strength and its determinants in pre- and early pubertal boys and girls. Bone. 2006;39:598–608. [PubMed]
10. Lovell DJ, Reiff A, Jones OY, Schneider R, Nocton J, Stein LD, et al. for the Pediatric Rheumatology Collaborative Study Group. Long-term safety and efficacy of etanercept in children with polyarticular-course juvenile rheumatoid arthritis. Arthritis Rheum. 2006;54:1987–1994. [PubMed]
11. Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001. J Rheumatol. 2004;31:390–392. [PubMed]
12. Singh G, Athreya BH, Fries JF, Goldsmith DP. Measurement of health status in children with juvenile rheumatoid arthritis. Arthritis Rheum. 1994;37:1761–1769. [PubMed]
13. Kvien TK, Hoyeraal HM, Sandstad B. Assessment methods of disease activity in juvenile rheumatoid arthritis—evaluated in a prednisolone/placebo double blind study. J Rheumatol. 1982;9:696–702. [PubMed]
14. Wallace CA, Huang B, Bandeira M, Ravelli A, Giannini EH. Patterns of clinical remission in select categories of juvenile idiopathic arthritis. Arthritis Rheum. 2005;52:3554–3562. [PubMed]
15. Ogden CL, Kuczmarski RJ, Flegal KM, Mei Z, Guo S, Wei R, et al. Centers for Disease Control and Prevention 2000 growth charts for the United States: improvements to the 1977 National Center for Health Statistics version. Pediatrics. 2002. pp. 45–60. [PubMed]
16. Tanner JM. Growth at adolescence. 2nd ed. Oxford: Blackwell Scientific; 1962.
17. Morris NN, Udry JR. Validation of a self-administered instrument to assess stage of adolescent development. J Youth Adolesc. 1980:271–280. [PubMed]
18. Sievanen H, Koskue V, Rauhio A, Kannus P, Heinonen A, Vuori I. Peripheral quantitative computed tomography in human long bones: evaluation of in vitro and in vivo precision. J Bone Miner Res. 1998;13:871–882. [PubMed]
19. Schoenau E, Neu CM, Rauch F, Manz F. Gender-specific pubertal changes in volumetric cortical bone mineral density at the proximal radius. Bone. 2002;31:110–113. [PubMed]
20. Tommasini SM, Nasser P, Schaffler MB, Jepsen KJ. Relationship between bone morphology and bone quality in male tibias: implications for stress fracture risk. J Bone Miner Res. 2005;20:1372–1380. [PubMed]
21. Schoenau E, Neu CM, Rauch F, Manz F. The development of bone strength at the proximal radius during childhood and adolescence. J Clin Endocrinol Metab. 2001;86:613–618. [PubMed]
22. Gilsanz V, Kovanlikaya A, Costin G, Roe TF, Sayre J, Kaufman F. Differential effect of gender on the sizes of the bones in the axial and appendicular skeletons. J Clin Endocrinol Metab. 1997;82:1603–1607. [PubMed]
23. Altman DG. Construction of age-related reference centiles using absolute residuals. Stat Med. 1993;12:917–924. [PubMed]
24. Kalkwarf HJ, Zemel BS, Gilsanz V, Lappe JM, Horlick M, Oberfield S, et al. The bone mineral density in childhood study: bone mineral content and density according to age, sex, and race. J Clin Endocrinol Metab. 2007;92:2087–2099. [PubMed]
25. Skaggs DL, Loro ML, Pitukcheewanont P, Tolo V, Gilsanz V. Increased body weight and decreased radial cross-sectional dimensions in girls with forearm fractures. J Bone Miner Res. 2001;16:1337–1342. [PubMed]
26. Ma D, Jones G. The association between bone mineral density, metacarpal morphometry, and upper limb fractures in children: a population-based case-control study. J Clin Endocrinol Metab. 2003;88:1486–1491. [PubMed]
27. Lien G, Flato B, Haugen M, Vinje O, Sorskaar D, Dale K, et al. Frequency of osteopenia in adolescents with early-onset juvenile idiopathic arthritis: a long-term outcome study of one hundred five patients. Arthritis Rheum. 2003;48:2214–2223. [PubMed]
28. Parfitt AM. The two faces of growth: benefits and risks to bone integrity. Osteoporos Int. 1994;4:382–398. [PubMed]
29. Jepsen KJ, Pennington DE, Lee YL, Warman M, Nadeau J. Bone brittleness varies with genetic background in A/J and C57BL/6J inbred mice. J Bone Miner Res. 2001;16:1854–1862. [PubMed]
30. Beck TJ, Ruff CB, Mourtada FA, Shaffer RA, Maxwell-Williams K, Kao GL, et al. Dual-energy X-ray absorptiometry derived structural geometry for stress fracture prediction in male U.S. Marine Corps recruits. J Bone Miner Res. 1996;11:645–653. [PubMed]
31. Ashe MC, Khan KM, Kontulainen SA, Guy P, Liu D, Beck TJ, et al. Accuracy of pQCT for evaluating the aged human radius: an ashing, histomorphometry and failure load investigation. Osteoporos Int. 2006;17:1241–1251. [PubMed]
32. Kaneki H, Guo R, Chen D, Yao Z, Schwarz EM, Zhang YE, et al. Tumor necrosis factor promotes Runx2 degradation through up-regulation of Smurf1 and Smurf2 in osteoblasts. J Biol Chem. 2006;281:4326–4333. [PMC free article] [PubMed]
33. Gilbert L, He X, Farmer P, Rubin J, Drissi H, van Wijnen AJ, et al. Expression of the osteoblast differentiation factor RUNX2 (Cbfa1/AML3/Pebp2αA) is inhibited by tumor necrosis factor-α J Biol Chem. 2002;277:2695–2701. [PubMed]
34. Kudo O, Sabokbar A, Pocock A, Itonaga I, Fujikawa Y, Athanasou NA. Interleukin-6 and interleukin-11 support human osteoclast formation by a RANKL-independent mechanism. Bone. 2003;32:1–7. [PubMed]
35. Marotte H, Pallot-Prades B, Grange L, Gaudin P, Alexandre C, Miossec P. A 1-year case-control study in patients with rheumatoid arthritis indicates prevention of loss of bone mineral density in both responders and nonresponders to infliximab. Arthritis Res Ther. 2007;9:R61. [PMC free article] [PubMed]
36. Li Y, Li A, Strait K, Zhang H, Nanes MS, Weitzmann MN. Endogenous TNFα lowers maximum peak bone mass and inhibits osteoblastic Smad activation through NF-κB. J Bone Miner Res. 2007;22:646–655. [PubMed]
37. De Benedetti F, Rucci N, Del Fattore A, Peruzzi B, Paro R, Longo M, et al. Impaired skeletal development in interleukin-6-transgenic mice: a model for the impact of chronic inflammation on the growing skeletal system. Arthritis Rheum. 2006;54:3551–3563. [PubMed]
38. Weinstein RS, Chen JR, Powers CC, Stewart SA, Landes RD, Bellido T, et al. Promotion of osteoclast survival and antagonism of bisphosphonate-induced osteoclast apoptosis by glucocorticoids. J Clin Invest. 2002;109:1041–1048. [PMC free article] [PubMed]
39. Canalis E, Bilezikian JP, Angeli A, Giustina A. Perspectives on glucocorticoid-induced osteoporosis. Bone. 2004;34:593–598. [PubMed]
40. Burnham JM, Shults J, Petit MA, Semeao E, Beck TJ, Zemel BS, et al. Alterations in proximal femur geometry in children treated with glucocorticoids for Crohn disease or nephrotic syndrome: impact of the underlying disease. J Bone Miner Res. 2007;22:551–559. [PubMed]
41. Xian CJ, Cool JC, Scherer MA, Macsai CE, Fan C, Covino M, et al. Cellular mechanisms for methotrexate chemotherapy-induced bone growth defects. Bone. 2007;41:842–850. [PubMed]
42. Minaur NJ, Jefferiss C, Bhalla AK, Beresford JN. Methotrexate in the treatment of rheumatoid arthritis. I. In vitro effects on cells of the osteoblast lineage. Rheumatology (Oxford) 2002;41:735–740. [PubMed]
43. Bianchi ML, Cimaz R, Galbiati E, Corona F, Cherubini R, Bardare M. Bone mass change during methotrexate treatment in patients with juvenile rheumatoid arthritis. Osteoporos Int. 1999;10:20–25. [PubMed]
44. Valta H, Lahdenne P, Jalanko H, Aalto K, Makitie O. Bone health and growth in glucocorticoid-treated patients with juvenile idiopathic arthritis. J Rheumatol. 2007;34:831–836. [PubMed]
45. Dubner SE, Shults J, Leonard MB, Zemel BS, Sembhi H, Burnham JM. Assessment of spine bone mineral density in juvenile idiopathic arthritis: impact of scan projection. J Clin Densitom. 2007 E-pub ahead of print. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...