Epidemiology of juvenile idiopathic arthritis in a multiethnic cohortEthnicity as a risk factor.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 56, No. 6, June 2007, pp 1974–1984 DOI 10.1002/art.22709 © 2007, American College of Rheumatology Epidemiology of Juvenile Idiopathic Arthritis in a Multiethnic Cohort Ethnicity as a Risk Factor R. K. Saurenmann,1 J. B. Rose,2 P. Tyrrell,2 B. M. Feldman,2 R. M. Laxer,2 R. Schneider,2 and E. D. Silverman2 types. A higher frequency of enthesitis-related JIA was observed among patients of Asian origin, while those of black origin or native North American origin were more likely to develop polyarticular RF-positive JIA. Conclusion. In this multiethnic cohort, European descent was associated with a significantly increased risk of developing JIA, and the distribution of JIA subtypes differed significantly across ethnic groups. Objective. To study the influence of ethnicity on the risk of developing juvenile idiopathic arthritis (JIA) in a multiethnic community of patients with unrestricted access to health care. Methods. A questionnaire on ethnicity was distributed to all patients with JIA being followed up at the Hospital for Sick Children in Toronto, Ontario, Canada. Of 1,082 patients, 859 (79.4%) responded to the questionnaire. To calculate the relative risk (RR) of developing JIA in this study cohort, the results were compared with data from the age-matched general population of the Toronto metropolitan area (TMA) as provided in the 2001 census from Statistics Canada. Results. European descent was reported by 69.7% of the patients with JIA compared with a frequency of 54.7% in the TMA general population, whereas a statistically significantly lower than expected percentage of the patients with JIA reported having black, Asian, or Indian subcontinent origin. Children of European origin had a higher RR for developing any of the JIA subtypes except polyarticular rheumatoid factor (RF)– positive JIA, and were particularly more likely to develop the extended oligoarticular and psoriatic sub- Juvenile idiopathic arthritis (JIA) is the most common rheumatic disease occurring in childhood in the Western world. It is characterized by persistent inflammation of the joints, with onset prior to age 16 years. Although there have been multiple names associated with this illness, including juvenile rheumatoid arthritis (JRA), juvenile chronic arthritis, and juvenile arthritis, the currently agreed-upon classification is the International League of Associations for Rheumatology (ILAR) revised classification, which refers to the illness as JIA (1). The major subtypes of JIA are oligoarticular JIA (up to 4 affected joints), which may be persistent or extended (involving more than 4 joints after the first 6 months of disease), polyarticular (more than 4 affected joints) rheumatoid factor (RF)–negative JIA, polyarticular RF-positive JIA, systemic JIA, enthesitis-related arthritis (ERA), psoriatic JIA, or a classfication of “other JIA” when the criteria for more than 1 subtype of JIA or none of the criteria were met. The first epidemiologic studies of JIA performed in populations from Western Europe and North America showed an incidence of JIA between 1.3 and 22.6 per 100,000 person-years (2–8) and a prevalence of JIA between 7 and 148 per 100,000 person-years (4,7–10). However, comparison of these studies is difficult, because different diagnostic criteria and different study designs were used. The few studies that were Supported by grants from the Swiss Arthritis Society (Rheumaliga Schweiz) and Stiftung Sanitas, Switzerland. Dr. Feldman holds a Canada Research Chair in Childhood Arthritis. 1 R. K. Saurenmann, MD: Zurich University Children’s Hospital, Zurich, Switzerland, and The Hospital for Sick Children and University of Toronto, Toronto, Ontario, Canada; 2J. B. Rose, HBSc, P. Tyrrell, MSc, B. M. Feldman, MD, MSc, FRCPC, R. M. Laxer, MD, FRCPC, R. Schneider, MBBCh, FRCPC, E. D. Silverman, MD, FRCPC: The Hospital for Sick Children and University of Toronto, Toronto, Ontario, Canada. Address correspondence and reprint requests to R. K. Saurenmann, MD, Zurich University Children’s Hospital, Steinwiesstrasse 75, 8032 Zurich, Switzerland. E-mail: traudel.saurenmann@ kispi.unizh.ch. Submitted for publication November 24, 2006; accepted in revised form March 7, 2007. 1974 ETHNICITY IN JIA population-based showed a higher prevalence of JIA, between 167 and 400 per 100,000 person-years; this difference is likely due to undiagnosed cases being identified in “healthy” children when they underwent examination by trained pediatric rheumatologists (11,12). More recently, epidemiologic studies from countries other than in Europe and North America showed a different pattern of prevalence of the JIA subtypes as well as a different rate of development of uveitis (a common complication of JIA) (13–23) as compared with the results obtained in the North American and European studies. A study from Japan showed a very low overall disease prevalence (18), and studies from India and Kuwait showed a much lower than expected rate of antinuclear antibody (ANA) positivity and uveitis (13,17). Studies of African American patients and black South African patients showed a higher rate of RFpositive polyarticular JIA and a lower rate of uveitis in these ethnic groups (14,19). Moreover, in studies of native North American patients, higher rates of the polyarticular subtypes, both RF positive and RF negative, were found (9,16,24,25). Possible reasons for the differences observed between ethnic groups include environmental and genetic factors but also underrepresentation of the milder forms of JIA, such as oligoarticular JIA, because of the potential for underdiagnosis in children, which could be attributed to restrictions in their access to health care facilities. The Canadian Health Care System provides unrestricted and universal access to health care facilities for all of its residents. Toronto is a city with a predominant immigrant component and high ethnic diversity within the population. Pediatric rheumatology clinical services in Toronto are provided by a single tertiary care center, which serves the population of 4.7 million inhabitants of the Toronto metropolitan area (TMA). In the present study, we used this background population to define the influence of ethnicity on the prevalence and presentation of JIA and the risk of developing JIArelated uveitis. PATIENTS AND METHODS Patients. A search of the database of the Division of Rheumatology at The Hospital for Sick Children in Toronto revealed that arthritis had been diagnosed in 1,605 patients between July 1, 1984 and June 30, 2002. Following a review of the charts of all 1,605 patients, we identified a total of 1,082 patients who had a confirmed diagnosis of JIA and whose minimum followup time was 1 month. The following data were collected: sex, date of birth, date of diagnosis of JIA, type of JIA, ANA status, RF status, HLA–B27 status, and date of 1975 diagnosis of uveitis. The study was approved by the Research Ethics Board of The Hospital for Sick Children. In autumn 2003, a questionnaire was either mailed or handed out at clinic visits to the cohort of 1,082 patients. The questionnaire required patients, or their parents, to provide information about their ancestry/country of origin. In addition, patients were contacted by telephone and asked to provide the same information verbally, after they had given their informed consent to participate in the study. The questionnaire contained a list of countries of origin (Figure 1), grouped by ethnic regions in accordance with the list of countries and ethnic regions used in the 2001 census by Statistics Canada. Patients were asked to provide a response for as many categories as applied. Questionnaires from 859 patients (79.4%) were returned. The answers from 101 patients were not useful in our study, because 6 patients had not given any answer, 36 had indicated Canadian as their ethnic group, and 59 had indicated more than 1 ethnicity. Therefore, our study cohort consisted of 758 patients with JIA. Diagnosis of JIA. The diagnosis of JIA and identification of subtypes of JIA were based on the ILAR 1997 revised criteria (26). Patients whose diagnosis was made prior to the use of the ILAR criteria were subsequently diagnosed as having JIA according to the ILAR criteria from a review of the medical records. If only 1 test of RF was performed, as was the case in many patients, then the results of this test were used to assign a JIA subtype rather than apply the subtype category of “other JIA.” ANAs were considered positive if a titer of ⱖ1:20 was obtained on at least 1 clinic visit during the disease course. This cutoff point for ANA positivity is in accordance with the normal values used in our laboratory. In the context of the present study, we chose to apply this cutoff value to identify patients with ANA-positive JIA, because we do not know of any evidence suggesting that a different cutoff level would be better to recognize the JIA subtype associated with ANAs. Ethnicity. According to the answers obtained from the questionnaire, the patients were categorized into 1 of the following ethnic groups: 1) European, comprising patients of European or west Asian origin; 2) black, comprising patients of black African, African American, or Caribbean origin; 3) native North American origin; 4) Latin American, comprising patients of Central or South American origin; 5) Asian, comprising patients of southeast Asian, Chinese, Japanese, or Pacific Islands origin; 6) Indian subcontinent origin; and 7) Arab, comprising patients of north African or Arabic origin. For some analyses, the patients were categorized into 1 of 2 groups, those of European origin and those of non-European origin. TMA general population data on ethnicity. Data from the 2001 census provided by Statistics Canada were used for comparison of the distribution of ethnicities found in our JIA study cohort with that in the general age-matched population of the TMA. Statistics Canada provided data for the percentage of children ages 0–14 years and those ages 15–24 years, rather than the desired single grouping of 0–18 years. In addition, the age distribution within each ethnic group was available for only the total Canadian population, and not specifically for the population of Toronto. Therefore, to account for these differences, the number of children/adolescents in each ethnic group within the TMA population was derived as follows: total TMA population of any ethnic group ⫻ (percentage of that ethnic group within the 0–14-year age 1976 SAURENMANN ET AL Figure 1. Ethnicity questionnaire administered to 1,082 patients with juvenile idiopathic arthritis. group [for all Canada] ⫹ 30% of the percentage of that ethnic group for ages 15–24 years [for all Canada]); the latter proportion of 30% was used because children ages 15–18 years account for 30% of the 15–24-year-olds in the total Canadian population. Statistical analysis. Statistical analyses were performed using the JMP IN 5.1 program (SAS Institute, Cary, NC). The chi-square test, one-way analysis of variance (ANOVA), Kaplan-Meier survival analysis, and Cox proportional hazards regression analyses were used to assess the 79 (10.4) 87.3 79.5 3.9 (0.6–13.8) 223 (29.4) 76.2 71.2 5.1 (1.1–16.4) 302 (39.8) 79.1 73.4 4.8 (0.6–16.4) 417 (38.6) 77.5 72.5 4.8 (0.6–16.4) All 167 (22.0) 78.4 51.9 6.2 (0.5–17.8) 223 (20.6) 79.8 50 6.3 (0.5–17.8) Polyarticular RF negative 27 (3.6) 88.9 56 11.4 (3.4–16.1) 34 (3.1) 88.2 56.3 11.1 (3.4–16.1) Polyarticular RF positive * JIA ⫽ juvenile idiopathic arthritis; RF ⫽ rheumatoid factor; ANA ⫽ antinuclear antibody. 99 (9.2) 88.9 80.6 3.8 (0.6–13.8) Extended 318 (29.4) 73.9 69.9 5.1 (0.6–16.4) Persistent Oligoarticular JIA Characteristics of the patients with JIA in the total hospital cohort and the study cohort* Total hospital cohort No. (%) of patients % female % ANA positive Age at diagnosis, mean (range) years Study cohort No. (%) of patients % female % ANA positive Age at diagnosis, mean (range) years Table 1. 103 (13.6) 51.5 19.3 7.4 (1.1–17.0) 157 (14.5) 47.8 19 7.0 (0.9–17.0) Systemic JIA 82 (10.8) 51.2 50.7 9.0 (1.1–17.2) 122 (11.3) 56.6 50 8.9 (0.8–17.2) Psoriatic JIA 67 (8.8) 25.4 32.5 11.4 (2.8–16.9) 115 (10.6) 22.6 30.8 11.7 (2.8–17.6) Enthesitisrelated JIA 10 (1.3) 50 10 7.4 (2.8–10.8) 14 (1.3) 64.3 8.3 8.3 (2.8–12.6) Other JIA 758 67.4 56.9 6.74 (0.5–17.8) 1,082 65.6 55.7 6.9 (0.5–17.8) Total ETHNICITY IN JIA 1977 1978 SAURENMANN ET AL differences between ethnic groups. The chi-square test was used to compare the distribution of ethnicities in our JIA cohort with the distribution of ethnicities in the age-matched TMA general population. In addition, the chi-square test was used to calculate the relative risk (RR) of developing JIA among the different ethnic groups in our study cohort (see the VassarStats Web site for Statistical Computation, at http:// faculty.vassar.edu/lowry/VassarStats.html). RESULTS Ethnic differences between the JIA study cohort and the TMA general population. We did not find any statistically significant differences in the characteristics of the patients in our JIA study cohort of 758 patients as compared with The Hospital for Sick Children total JIA cohort of 1,082 patients (Table 1). Using the responses from the questionnaire, we determined the distribution of ethnicities among our JIA study cohort and compared it with the data provided by Statistics Canada on the general population of children in the TMA. We found that there was an overrepresentation of patients of European origin and native North American origin and an underrepresentation of patients of black, Asian, or Indian subcontinent origin in our JIA cohort as compared with the TMA general population (Figures 2A and B). European descent was reported by 69.7% of the patients with JIA compared with a frequency of 54.7% in the TMA general population. The difference in the distribution of children according to European versus non-European ancestry in our JIA study cohort as compared with that in the TMA population of children was statistically significant (P ⫽ 0.005 by chi-square test). Overall, children of European ancestry had an increased RR for developing JIA (RR 1.26), whereas children of non-European ancestry had a decreased RR for developing JIA (RR 0.43); both of these risk values were statistically significant when compared with the predicted values (P ⬍ 0.0001). When we directly compared the risk of developing JIA in the group of children of European descent with that in the group of children of non-European descent, we found that those of European descent had a significantly increased risk of developing JIA (RR 3.12) as compared with those of non-European descent (P ⬍ 0.0001). Among the individual non-European ethnic groups of children, those of black origin (RR 0.33), Indian subcontinent origin (RR 0.38), and Asian origin (RR 0.41) each had a statistically significantly lower likelihood of developing JIA as compared with the predicted values in the total at-risk population of children in the TMA (each P ⬍ 0.0001 versus TMA population). Similarly, children of Latin American ori- Figure 2. Distribution of ethnicities among the population of children and adolescents in the Toronto metropolitan area as provided by Statistics Canada in the 2001 census (A) and in the study cohort of 758 patients with juvenile idiopathic arthritis (B). gin and those of Arab origin had a decreased RR for developing JIA (RRs of 0.72 and 0.72, respectively), but these estimates were not statistically significant when compared with the predicted values (P ⫽ 0.40 and P ⫽ 0.54, respectively). The only non-European ethnic group in whom the RR of developing JIA was increased was the children of native North American descent (RR 1.72), but this was not statistically significant when compared with the predicted values (P ⫽ 0.075). Of note, only 10 patients with JIA in our cohort were of native North American descent. Characteristics of the study patients with JIA. The mean age at diagnosis of JIA was significantly ETHNICITY IN JIA younger among patients of European origin, whose mean age at diagnosis was 6.5 years (95% confidence interval [95% CI] 6.1–6.8 years), as compared with those of non-European origin, whose mean age at diagnosis was 7.8 years (95% CI 7.1–8.4 years) (P ⫽ 0.005 by one-way ANOVA). Patients of European origin had a higher female:male ratio, at 7:3, as compared with a female:male ratio of 3:2 for those of non-European origin (P ⫽ 0.047). Patients of native North American origin and those of Latin American origin had the highest female:male ratios, both at 9:1, while patients of black origin and those of Indian subcontinent origin each had female:male ratios of 2:1. Patients of Asian origin and those of Arab origin had the lowest female: male ratios, at 1:1 and 2:3, respectively. The sex distributions were statistically significantly different among the ethnic groups (P ⫽ 0.002 by chi-square test). The frequency of RF positivity was 3.4% among patients of European origin (13 of 384 patients tested), which was much lower than the frequency of 14.4% (16 of 111 patients tested) among patients of non-European origin (P ⫽ 0.0003 by chi-square test). There was no statistically significant difference in the frequency of ANA positivity between children of European ancestry and those of non-European ancestry (P ⫽ 0.12). There was also no significant difference in the overall rate of uveitis between patients of European ancestry and those of non-European ancestry (14.4% and 12.8%, respectively, P ⫽ 0.6). HLA–B27 was positive in 28.1% of the patients of European descent (46 of 164 patients tested) and in 40.5% of the patients of non-European descent (17 of 42 patients tested); this difference in frequency of HLA–B27 between groups was not significant (P ⫽ 0.13), which may be attributable to the low number of patients tested. Distribution of JIA subtypes in the study cohort. Within the European and non-European ethnic groups, subtypes of JIA were not equally distributed (Table 2). As compared with the distribution of disease subtypes among the patients in the total JIA study cohort, children of European ancestry were overrepresented among patients with the subtypes of extended oligoarticular JIA and psoriatic JIA, whereas children of non-European ancestry were overrepresented among those with the subtypes of RF-positive polyarticular JIA, ERA, and systemic JIA (P ⫽ 0.001 versus total JIA study cohort, by chi-square test). We then subdivided the patients of nonEuropean ancestry into specific ethnic groups and examined the distribution of JIA subtypes by ethnicity. We found that RF-positive polyarticular JIA was overrepresented among children of native North American, black, 1979 Table 2. Distribution of JIA subtypes in children of European descent compared with those of non-European descent* Ethnic group JIA subtype All subtypes Systemic Oligoarticular Persistent Extended RF-negative polyarticular RF-positive polyarticular Enthesitis-related Psoriatic Young-onset ANA positive† European (n ⫽ 599) Non-European (n ⫽ 159) 79 75.7 21 24.3 80.3 88.6 82 48.1 67.2 87.8 83.3 19.7 11.4 18 51.9 32.8 12.2 16.7 * Values are the percent of patients within the European or nonEuropean group. See Table 1 for definitions. † Subset of patients with onset of ANA-positive arthritis at a younger age. Latin American, or Indian subcontinent ethnicity. ERA was overrepresented in children of Asian origin, whereas persistent oligoarticular JIA was underrepresented in native North American children. Extended oligoarticular JIA was not seen in any patient of Arab or Latin American origin, and psoriatic JIA was underrepresented in the Indian subcontinent, Latin American, and native North American ethnic groups (Table 3). The difference in the distribution of the JIA subtypes between the specific non-European ethnic groups was significant (P ⫽ 0.0005 by chi-square test). We then examined the RR values for the likelihood of developing any of the subtypes of JIA relative to the total at-risk population in the TMA, i.e., the number of children ages ⬍18 years in the ethnic groupings defined by Statistics Canada. We found that children of European origin had an increased RR for developing any of the subtypes of JIA except RF-positive polyarticular JIA (Table 4). Among the specific non-European ethnic categories, children of black origin and those of Indian subcontinent origin were at a lower risk of developing any of the JIA subtypes except RF-positive polyarticular JIA, while children of Asian origin were at a lower risk of developing any of the JIA subtypes except ERA (Table 4). Although there were only small numbers of patients of Arab origin or Latin American origin within each subtype, the distribution of risk of developing each subtype was similar to that seen in patients of European descent. Patients of native North American descent tended to develop polyarthritis, including both RF-positive and RF-negative polyarticular JIA and extended oligoarticular JIA. 1980 SAURENMANN ET AL Table 3. Distribution of JIA subtypes by specific non-European versus European ethnic group* Non-European JIA subtype Oligoarticular Persistent Extended RF-negative polyarticular RF-positive polyarticular Systemic Psoriatic Enthesitis-related Native P, JIA study Indian Latin North P, European between cohort European All Asian subcontinent Black Arab American American vs. nonethnic (n ⫽ 758) (n ⫽ 599) (n ⫽ 159) (n ⫽ 50) (n ⫽ 40) (n ⫽ 31) (n ⫽ 17) (n ⫽ 11) (n ⫽ 10) European groups 29.8 9.3 22.2 3.3 13.7 11.4 9.3 30.1 11.8 23.1 2.2 13.1 12.1 7.6 29.6 6.3 18.9 8.8 16.3 6.3 13.8 26.0 2.0 20.0 4.0 12.0 8.0 24.0 35.0 10.0 12.5 7.5 17.5 0.0 12.5 19.4 9.7 19.4 16.1 12.9 12.9 9.7 35.3 0.0 17.7 0.0 23.5 11.8 11.8 36.4 0.0 18.2 18.2 27.3 0.0 0.0 10.0 10.0 40.0 20.0 10.0 0.0 0.0 0.79 0.029 0.2 0.0004 0.36 0.03 0.024 0.5 0.048 0.57 0.0013 0.74 0.012 0.013 * Values are the percent of the 758 patients in the JIA study cohort. The subtype of “other JIA” is not shown; 5 patients of European descent, 2 patients of Asian descent, 2 patients of Indian subcontinent descent, and 1 patient of native North American descent were classified in the “other JIA” subtype. P values were calculated by chi-square test. See Table 1 for definitions. Risk of young-onset ANA-positive arthritis. It has been suggested that patients with onset of ANA-positive arthritis at a younger age constitute a homogeneous disease subset. We therefore examined whether this was true across different ethnicities (27). We found that the RR for developing young-onset ANA-positive arthritis in children of European origin as compared with children of non-European origin was 4.21 (P ⬍ 0.0001) (Table 4), and that patients in the young-onset ANApositive disease subset made up 31.6% of the patients of European origin but only 23.6% of the patients of non-European origin (P ⫽ 0.03 by chi-square test). Among the different ethnic groups, patients of Asian origin had the lowest rate of young-onset ANA-positive arthritis (18%), and those of native North American origin had the highest rate (40%), but these differences between the individual ethnic groups were not significant (P ⫽ 0.095). Among the non-European ethnic groups, the RR for young-onset ANA-positive arthritis as compared with that among the total at-risk TMA population of children ages ⬍18 years was 0.24 for children of Asian origin (P ⬍ 0.0001), 0.26 for children of black origin (P ⫽ 0.0003), 0.36 for children of Indian subcontinent origin (P ⫽ 0.0001), 0.52 for children of Latin American origin (in only 2 patients), and 0.75 for children of Arab origin (in only 5 patients) (Table 4). Risk of JIA-related uveitis. Within the JIA study cohort, the rate of uveitis was the same regardless of ethnicity. However, when compared with the total atrisk TMA population of children ages ⬍18 years, chil- Table 4. Relative risk of developing JIA (by subtype) and JIA-related uveitis, according to ethnic group* Non-European† Arab Indian subcontinent Native North American European vs. non-European‡ JIA subtype European† All Asian Black Latin American All JIA Oligoarticular Persistent Extended RF-negative polyarticular RF-positive polyarticular Psoriatic Enthesitis-related Systemic Young-onset ANA positive Uveitis 1.26§ 0.43§ 0.41§ 0.33§ 0.72 0.8 0.38§ 1.72 3.12§ 1.28§ 1.43§ 1.31§ 0.8 1.46§ 1.05 1.16 1.34§ 1.27§ 0.39§ 0.25§ 0.36§ 1 0.23§ 0.61§ 0.46§ 0.32§ 0.36§ 0.35§ 0.07§ 0.38§ 0.39 0.32§ 1.1 0.34§ 0.24§ 0.28§ 0.22§ 0.22 0.31§ 1.5 0.42 0.38 0.27§ 0.26§ 0.36 0.82 0 0.93 3 0 0 1.32 0.52 1.08 0.95 0 0.67 0 0.92 0.98 1.2 0.75 0.26 0.44§ 0.39§ 0.21§ 0.92 0 0.54 0.49§ 0.36§ 0.33§ 0.45 1.79 3.2 8.4 0 0 0.91 2.4 2.98 3.3§ 6.0§ 3.9§ 0.8 6.4§ 1.74§ 2.5§ 4.21§ 3.5§ * JIA ⫽ juvenile idiopathic arthritis; RF ⫽ rheumatoid factor; ANA ⫽ antinuclear antibody. See Figure 1 for definitions of ethnic groups. † Values are the relative risk versus the total at-risk population (the number of children ages ⬍18 years in each ethnic group as defined by Statistics Canada. ‡ Values are the relative risk in European versus non-European ethnic groups. § P ⬍ 0.05 versus other ethnic groups or between groups. ETHNICITY IN JIA dren of European ancestry had an RR of 1.27 for developing JIA-related uveitis (P ⫽ 0.036), whereas children of non-European ancestry had an RR of 0.36 for developing JIA-related uveitis (P ⬍ 0.0001) (Table 4); thus, those of European origin as compared with those of non-European origin had an increased risk of developing JIA-related uveitis (RR 3.5). The presence of uveitis was statistically significantly associated with ANA positivity among patients of European descent (72 of 86 patients tested; P ⬍ 0.0001) but not among patients of non-European descent (13 of 18 patients tested; P ⫽ 0.10). Among all specific ethnic groups of children of non-European descent, the majority of patients with uveitis were ANA positive, but the number of patients with uveitis was very small, and there was no statistically significant association between the presence of uveitis and ANA positivity. Each of the individual risk factors of ANA status, sex, JIA subtype, and age at diagnosis has been associated with the development of uveitis in patients of different ethnicities (28). However, our Cox regression analysis, which incorporated all of the available variables (ANA status, sex, JIA subtype, age at diagnosis, and ethnicity), demonstrated that the combined profile of young-onset ANA-positive arthritis was the only risk factor associated with a significant risk of developing JIA-related uveitis, with an RR of 1.7 (P ⬍ 0.00009). We also performed Kaplan-Meier survival analyses in the different ethnic groups and found that youngonset ANA-positive arthritis was a significant risk factor for the development of JIA-related uveitis both in patients of European origin (P ⬍ 0.0001) and in patients of non-European origin (P ⫽ 0.0023). However, among the non-European ethnic groups, young-onset ANApositive arthritis was a significant risk factor only in children of black origin (P ⫽ 0.027) and in those of Latin American origin (P ⫽ 0.034), and there was a tendency toward an association in those of native North American origin (P ⫽ 0.07). In contrast, young-onset ANA-positive arthritis was not a significant risk factor in children of Arab origin (P ⫽ 0.12), Asian origin (P ⫽ 0.39), or Indian subcontinent origin (P ⫽ 0.7). Because of the small numbers of patients, it was not possible to perform Cox proportional hazards regression analysis of multiple risk factors by ethnicity. DISCUSSION Most large epidemiologic studies in which the prevalence of subtypes of JIA has been examined were based on populations of patients who were mainly of European ancestry. In contrast, smaller studies from Hawaii (20), Asia (22,23,29), and Central America 1981 (21,30) showed that there may be ethnic differences in the prevalence of JIA subtypes. The rheumatology clinic at The Hospital for Sick Children in Toronto is a large clinic that serves a multiethnic population of more than 4 million people whose access to health care is universal and unrestricted, making it an ideal center at which to study the potential influence of ethnicity. In this study, we demonstrated that there is a difference in the prevalence of JIA and in the subtype distribution of JIA based on ethnic background. Results of previous studies have suggested that oligoarticular JIA, RF-positive polyarticular JIA, and ERA have different prevalence rates in different ethnic groups (14,19,31). Our results demonstrated that children of European ancestry, as compared with children of non-European ancestry, had an increased risk of developing all of the subtypes of JIA except RF-positive polyarticular JIA and systemic JIA. Children of black or Indian subcontinent origin were at a lower risk of developing any of the JIA subtypes except RF-positive polyarticular JIA, while children of Asian origin were less likely than children of European ancestry to develop any of the subtypes of JIA except ERA. Patients of native North American origin tended to have an increased risk of developing polyarthritis, including RFpositive and RF-negative polyarticular JIA and extended oligoarticular JIA, as compared with the children of European descent. It has been suggested in some studies, but not all studies, that patients with JIA from the Indian subcontinent as well as those from Asia and Costa Rica have lower rates of young-onset oligoarticular arthritis and lower rates of uveitis as compared with European patients (13,21–23,31–33). Overall, we did not find any differences in the percentage of patients with persistent oligoarticular JIA between the European group of children and the non-European group of children, nor did we find any differences in the percentage of patients with oligoarthritis between those of Indian subcontinent descent and either the total JIA cohort or the patients of European descent. These results suggest that the apparent underrepresentation of patients with oligoarthritis found in some studies from India may represent a referral bias (31,32). Consistent with previous findings, however, we did observe that patients of native North American origin, black origin, and Asian origin were less likely to have oligoarticular JIA than were patients in other ethnic groups (22,23,34,35). Our population of children of Latin American origin, comprising 11 patients, was too small to make meaningful comparisons with the results in the study from Costa Rica (21,30). Overall, 1982 patients of European ancestry were more likely to develop extended oligoarticular JIA than were patients of non-European ancestry, with the lowest frequency of extended disease seen in patients of Asian, Arab, or Latin American descent. Previous studies from Asia demonstrated that patients with oligoarticular JRA were older at disease onset and were less likely to be female than their European counterparts, and that spondylarthropathy was more frequent in Asian patients; however, ethnicityspecific prevalence or incidence rates of JRA were not given (23). It is likely that such patients would have been classified as having ERA in our study. If this assumption is true, then our findings further confirm that patients of non-European ancestry are more likely to have ERA than are patients of European ancestry, and that this is attributable to an overrepresentation of patients with ERA among patients of Asian, Indian subcontinent, or Arab ancestry. Although patients of Asian descent constituted a larger proportion of the total ERA population than was predicted based on the percentage of Asians in the total JIA cohort, these patients had the same risk of developing ERA as did patients of European descent. These findings are consistent with those reported in studies from Singapore and Taiwan, in which it was shown that the overall prevalence of JIA was lower in Asian patients as compared with European patients, and that ERA was the most common subtype (22,23). It has previously been suggested that patients with JRA who were of African American or native North American descent were more likely to have RF-positive polyarticular JRA (20% of African American patients and at least 50% of native North American patients) than were patients of European origin (4% with RF-positive polyarticular JRA) (9,14,24,34,36). Consistent with those results, we found that 16% of black patients and 20% of native North American patients had RF-positive polyarticular JIA as compared with 3% of the total JIA cohort and 2% of patients of European descent. There was a trend toward an increased RR for the development of RF-positive polyarticular JIA in black children as compared with children of European descent in the total cohort. There was also a trend toward an increased RR in native North American children; however, our at-risk population may have been too small to allow detection of the absolute increased risk reported in previous studies in larger populations of adult and pediatric native North American patients (9,24,34,37). In contrast to the results among patients with RF-positive polyarticular JIA, we did not find any statistically significant difference in the percentage of SAURENMANN ET AL European patients and percentage of non-European patients with RF-negative polyarticular JIA, although children/adolescents of European ancestry in the Toronto area had a higher risk of developing polyarticular JIA than children of non-European ancestry. The only exception was the native North American population, whose RR for developing polyarticular RF-negative JIA was the highest, at 3.2. When we examined patients with extended oligoarticular JIA, we found that children of European descent had a higher frequency of this JIA subtype than did non-Europeans. Within the non-European group of patients with JIA, those of black, Indian subcontinent, or native North American origin had a frequency of extended oligoarticular JIA that was similar to that in European patients. In addition, children of European descent as compared with children of non-European descent had a higher risk of developing this form of JIA. Within the non-European group of children, those of Asian, Arab, or Latin American origin had an exceptionally low risk of developing extended oligoarticular JIA. Only native North American children had a risk of developing this subtype that was comparable with that in European children. To our knowledge, this is the first study to examine whether the risk of extended arthritis in patients with oligoarticular JIA differed according to ethnicity. There was a relative overrepresentation of Europeans as compared with non-Europeans among the disease subtype of psoriatic JIA. This was true for children in all non-European ethnic groups, except for children of Asian descent, in whom the frequency of psoriatic JIA was similar to that in European patients. Of note, we did not find any cases of psoriatic arthritis among the 38 patients of Indian subcontinent origin. Previous studies in adults have shown that there is an overall lower prevalence of psoriasis in India as compared with Europe and North America (38), but to our knowledge, studies comparing the epidemiologic patterns of psoriatic arthritis among different ethnicities have not been performed. Most studies of pediatric patients have not differentiated psoriatic arthritis from other forms of spondylarthropathy. Thus, our report appears to be the first to comment on the prevalence of juvenile psoriatic arthritis in different ethnicities. Ravelli et al (27), in a study from Italy, suggested that patients with early-onset (diagnosis before age 6 years) ANA-positive, RF-negative JIA constitute a distinct disease subset, irrespective of the ILAR classification. When we examined patients with young-onset ANA-positive arthritis without regard to the number of involved joints or the presence of psoriasis, we found ETHNICITY IN JIA that this group contained a significantly higher than expected proportion of patients of European ancestry compared with those of non-European ancestry, and that the TMA pediatric population of patients of European ancestry had a higher RR of developing this particular disease entity than did patients of nonEuropean ancestry. Within the non-European ethnic groups of children, all groups had a decreased RR for developing young-onset ANA-positive arthritis except for children of native North American origin and children of Arab origin. These results suggest that a predisposing genetic factor is responsible for the development of this combination of JIA characteristics in patients of European ancestry. Further studies are required to assess whether the involvement of the joints differs among ethnicities in patients with young-onset ANApositive arthritis. We found that the percentage of patients of European ancestry who developed JIA-related uveitis and the percentage of patients in this ethnic group who were ANA positive were similar to the percentages of patients of non-European ancestry who had these disease characteristics. These results appear to be in contrast to the findings in studies from India (17), Costa Rica (21), South Africa (19), and Singapore (23). However, when we examined the RR values for developing uveitis in the TMA general population, we found that Europeans had a higher RR than did non-Europeans, which is likely associated with the higher risk of developing JIA in those of European ancestry; nevertheless, once JIA was established in either ethnic group, the risk of developing uveitis was the same. These results are consistent with the findings from Singapore and India, but not with the findings from Costa Rica (17,21,23); however, our number of Latin American patients was smaller than that in the study from Costa Rica (21), which may account for the discrepancy between studies. Our results suggest that the risk factors for JIAassociated uveitis differ between ethnicities, and that the commonly found risk factors, such as young age at onset and ANA positivity, may not apply in patients of Asian, Indian subcontinent, or Arab origin. We are aware of the possible shortcomings of our study, in particular the fact that the data on ethnicity were self-reported. However, since the ethnicity assessment used by Statistics Canada for the 2001 census was also self-reported and the same ethnic groups were used, we believe that our results are valid. Furthermore, as we have shown, our results are supported by a variety of previous studies. Despite the large size of the JIA cohort, our study was limited by the small number of patients of non-European origin. Although most of our 1983 findings are supported by a number of other reports, our study can, in many respects, only point out areas of interest for further research. In conclusion, our study was able to show, for the first time, in a multiethnic community, that European ancestry is an important predisposing factor for the development of JIA, but especially the disease subtypes of extended oligoarticular JIA and psoriatic JIA. European ancestry can also be a predisposing factor for a combination of JIA characteristics defined by young age at diagnosis and ANA positivity, and, in the presence of these characteristics, the development of JIA-related uveitis. Our study was further able to show that there were significant differences in JIA subtype distribution among the different ethnic groups, suggesting that predisposing genetic factors exist. We have shown that despite these ethnic differences in the distribution of JIA subtypes, children of European ancestry are at the highest risk of developing all types of JIA except RFpositive polyarticular JIA. The only other exception, as suggested by the findings of previous studies, may be that native North American children are at a higher risk than children of European descent of developing a polyarticular course of disease, such as extended oligoarticular JIA, RF-negative polyarticular JIA, and RFpositive polyarticular JIA. Finally, we were able to show that the known risk factors for development of JIA-related uveitis may not be sufficient to explain the development of uveitis in patients of all ethnic backgrounds. Additional country-specific prevalence studies and other studies in large multiethnic populations are required to confirm our findings. ACKNOWLEDGMENTS Special thanks go to David N. Fisman, MD, MPH, medical epidemiologist at The Hospital for Sick Children and the University of Toronto, for reviewing the manuscript from an epidemiologist’s perspective. This work would not have been possible without the many hours of work and hundreds of telephone calls performed by the rheumatology volunteers at The Hospital for Sick Children. We particularly thank Stephanie Silverman, Reto Saurenmann, and Max Silverman for their support with the data collection. AUTHOR CONTRIBUTIONS Dr. Saurenmann had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Saurenmann, Feldman, Silverman. Acquisition of data. Saurenmann, Rose, Tyrrell, Laxer, Schneider. Analysis and interpretation of data. Saurenmann, Tyrrell, Feldman, Laxer, Schneider, Silverman. Manuscript preparation. Saurenmann, Feldman, Laxer, Schneider, Silverman. Statistical analysis. Saurenmann, Feldman. 1984 SAURENMANN ET AL REFERENCES 1. 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;31:390–2. 2. Laaksonen AL. A prognostic study of juvenile rheumatoid arthritis: analysis of 544 cases. Acta Paediatr Scand 1966:1–163. 3. Malleson PN, Fung MY, Rosenberg AM. The incidence of pediatric rheumatic diseases: results from the Canadian Pediatric Rheumatology Association Disease Registry. J Rheumatol 1996; 23:1981–7. 4. Andersson Gare B, Fasth A, Andersson J, Berglund G, Ekstrom H, Eriksson M, et al. Incidence and prevalence of juvenile chronic arthritis: a population survey. Ann Rheum Dis 1987;46:277–81. 5. Kunnamo I, Kallio P, Pelkonen P. Incidence of arthritis in urban Finnish children: a prospective study. Arthritis Rheum 1986;29: 1232–8. 6. Towner SR, Michet CJ Jr, O’Fallon WM, Nelson AM. The epidemiology of juvenile arthritis in Rochester, Minnesota 1960–1979. Arthritis Rheum 1983;26:1208–13. 7. Moe N, Rygg M. Epidemiology of juvenile chronic arthritis in northern Norway: a ten-year retrospective study. Clin Exp Rheumatol 1998;16:99–101. 8. Prieur AM, Le Gall E, Karman F, Edan C, Lasserre O, Goujard J. Epidemiologic survey of juvenile chronic arthritis in France: comparison of data obtained from two different regions. Clin Exp Rheumatol 1987;5:217–23. 9. Rosenberg AM, Petty RE, Oen KG, Schroeder ML. Rheumatic diseases in Western Canadian Indian children. J Rheumatol 1982;9:589–92. 10. Gewanter HL, Roghmann KJ, Baum J. The prevalence of juvenile arthritis. Arthritis Rheum 1983;26:599–603. 11. Mielants H, Veys EM, Maertens M, Goemaere S, De Clercq L, Castro S, et al. Prevalence of inflammatory rheumatic diseases in an adolescent urban student population, age 12 to 18, in Belgium. Clin Exp Rheumatol 1993;11:563–7. 12. Manners PJ, Diepeveen DA. Prevalence of juvenile chronic arthritis in a population of 12-year-old children in urban Australia. Pediatrics 1996;98:84–90. 13. Khuffash FA, Majeed HA, Lubani MM, Najdi KN, Gunawardana SS, Bushnaq R. Epidemiology of juvenile chronic arthritis and other connective tissue diseases among children in Kuwait. Ann Trop Paediatr 1990;10:255–9. 14. Schwartz MM, Simpson P, Kerr KL, Jarvis JN. Juvenile rheumatoid arthritis in African Americans. J Rheumatol 1997;24:1826–9. 15. Ozen S, Karaaslan Y, Ozdemir O, Saatci U, Bakkaloglu A, Koroglu E, et al. Prevalence of juvenile chronic arthritis and familial Mediterranean fever in Turkey: a field study. J Rheumatol 1998;25:2445–9. 16. Mauldin J, Cameron HD, Jeanotte D, Solomon G, Jarvis JN. Chronic arthritis in children and adolescents in two Indian health service user populations. BMC Musculoskelet Disord 2004;5:30. 17. Aggarwal A, Misra RN. Juvenile rheumatoid arthritis in India: rarity of antinuclear antibody and uveitis. Indian J Pediatr 1996; 63:301–4. 18. Fujikawa S, Okuni M. A nationwide surveillance study of rheumatic diseases among Japanese children. Acta Paediatr Jpn 1997; 39:242–4. 19. Haffejee IE, Raga J, Coovadia HM. Juvenile chronic arthritis in black and Indian South African children. S Afr Med J 1984;65:510–4. 20. Kurahara D, Tokuda A, Grandinetti A, Najita J, Ho C, Yamamoto K, et al. Ethnic differences in risk for pediatric rheumatic illness in a culturally diverse population. J Rheumatol 2002;29:379–83. 21. Arguedas O, Fasth A, Andersson-Gare B, Porras O. Juvenile chronic arthritis in urban San Jose, Costa Rica: a 2 year prospective study. J Rheumatol 1998;25:1844–50. 22. Wu CJ, Huang JL, Yang MH, Yan DC, Ou LS, Ho HH. Clinical characteristics of juvenile rheumatoid arthritis in Taiwan. J Microbiol Immunol Infect 2001;34:211–4. 23. See Y, Koh ET, Boey ML. One hundred and seventy cases of childhood-onset rheumatological disease in Singapore. Ann Acad Med Singapore 1998;27:496–502. 24. Boyer GS, Templin DW, Lanier AP. Rheumatic diseases in Alaskan Indians of the southeast coast: high prevalence of rheumatoid arthritis and systemic lupus erythematosus. J Rheumatol 1991;18:1477–84. 25. Oen K, Schroeder M, Jacobson K, Anderson S, Wood S, Cheang M, et al. Juvenile rheumatoid arthritis in a Canadian First Nations (aboriginal) population: onset subtypes and HLA associations. J Rheumatol 1998;25:783–90. 26. Petty RE, Southwood TR, Baum J, Bhettay E, Glass DN, Manners P, et al. Revision of the proposed classification criteria for juvenile idiopathic arthritis: Durban, 1997. J Rheumatol 1998;25:1991–4. 27. Ravelli A, Felici E, Magni-Manzoni S, Pistorio A, Novarini C, Bozzola E, et al. Patients with antinuclear antibody–positive juvenile idiopathic arthritis constitute a homogeneous subgroup irrespective of the course of joint disease. Arthritis Rheum 2005; 52:826–32. 28. Kanski JJ. Juvenile arthritis and uveitis. Surv Ophthalmol 1990; 34:253–67. 29. Fujikawa S, Okuni M. Clinical analysis of 570 cases with juvenile rheumatoid arthritis: results of a nationwide retrospective survey in Japan. Acta Paediatr Jpn 1997;39:245–9. 30. Arguedas O, Fasth A, Andersson-Gare B. A prospective population based study on outcome of juvenile chronic arthritis in Costa Rica. J Rheumatol 2002;29:174–83. 31. Aggarwal A, Misra R. Juvenile chronic arthritis in India: is it different from that seen in Western countries? Rheumatol Int 1994;14:53–6. 32. Malaviya AN, Narayanan K. Juvenile chronic arthritis in India. Indian J Pediatr 1981;48:659–67. 33. Porkodi R, Subramaniam R, Krishnamurthy V, Madhavan R, Parthiban M, Chandrasekaran AN. Pattern of rheumatic diseases in south India. IV. Clinical profile of juvenile rheumatoid arthritis. J Assoc Physicians India 1990;38:771–3. 34. Oen K, Fast M, Postl B. Epidemiology of juvenile rheumatoid arthritis in Manitoba, Canada, 1975–92: cycles in incidence. J Rheumatol 1995;22:745–50. 35. Uotila E. The temporomandibular joint in adult rheumatoid arthritis: a clinical and roentgenologic study. Acta Odontol Scand 1964;22 Suppl 39:1–91. 36. Hill RH, Walters K. Juvenile rheumatoid arthritis: a medical and social profile of non-Indian and Indian children. Can Med Assoc J 1969;100:458–64. 37. Hill RH, Robinson HS. Rheumatoid arthritis and ankylosing spondylitis in British Columbia Indians: their prevalence and the challenge of management. Can Med Assoc J 1969;100:509–11. 38. Christophers E. Psoriasis: epidemiology and clinical spectrum. Clin Exp Dermatol 2001;26:314–20.