Abstract
Objective
The objective of this study is to evaluate racial/ethnic differences in disease manifestations and survival in a US cohort of patients with systemic sclerosis (SSc), with a focus on Asian patients.
Methods
A retrospective cohort study was conducted among Kaiser Permanente Northern California adults with an incident SSc diagnosis by a rheumatologist from 2007 to 2016, confirmed by a chart review to fulfill 2013 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) classification criteria. Self‐reported race/ethnicity was categorized as non‐Hispanic white, Asian, Hispanic, and black. Disease manifestations and survival were compared, using white patients as the reference.
Results
A total of 609 patients with incident SSc were identified: 89% were women, and 81% had limited cutaneous SSc, with a mean age at diagnosis of 55.4 ± 14.8 years. The racial/ethnic distribution was 51% non‐Hispanic white (n = 310), 25% Hispanic (n = 154), 16% Asian (n = 96), and 8% black (n = 49). Compared with white patients, black patients had a greater prevalence of diffuse disease (14.5% vs. 44.9%; P < 0.001), and Asians had higher rates of anti‐U1‐RNP antibodies (32.1% vs. 11.9%; P = 0.005). Nine‐year overall survival rates following SSc diagnosis were lower in Asian (52.3%), black (52.2%), and Hispanic patients (68.2%) compared with white patients (75.8%). Pulmonary hypertension and infections were the leading causes of death in Asian patients. Asian race was associated with higher mortality on univariable (hazard ratio [HR] 1.83 [95% confidence interval (CI) 1.08‐2.99]; P = 0.020) and multivariable analyses (HR 1.80 [95% CI 0.99‐3.16]; P = 0.047) when adjusting for age, sex, body mass index, cutaneous subtype, smoking status, interstitial lung disease, pulmonary hypertension, renal crisis, and malabsorption syndrome.
Conclusion
Asian patients with SSc in this US cohort had increased mortality compared with white patients. These patients warrant close monitoring for disease progression.
Significance & Innovations.
This is the first study to compare racial/ethnic differences in patients with systemic sclerosis (SSc) using the Kaiser Permanente Northern California population, with a focus on Asian patients.
Compared with white patients, Asian patients were found to have increased mortality and higher rates of anti‐U1‐RNP antibodies.
Pulmonary hypertension and infections were the leading causes of death in Asian patients. Further research is needed to externally validate these findings and to identify both genetic and environmental determinants of SSc disease expression in different racial groups.
INTRODUCTION
Systemic sclerosis (SSc) is an autoimmune connective tissue disease characterized by excessive collagen deposition, vascular damage, inflammation, and progressive fibrosis of the skin and visceral organs 1. Race has a significant influence on the epidemiology, clinical manifestations, survival, autoantibody frequencies, and genetic factors in SSc 2.
Multiple studies in black and white patients have implicated the interaction between racial background, autoantibody subtype, and genetic factors in determining disease manifestations, severity, and progression in SSc 3, 4. Black patients have been reported to have a more severe clinical phenotype 5, with a younger age at SSc onset 6, 7, a higher frequency of diffuse skin involvement 6, more extensive pulmonary disease 7, 8, a higher risk for scleroderma renal crisis 9, and an overall worse prognosis (including higher mortality), compared with white patients 10, 11, 12.
Studies examining clinical characteristics and outcomes in Asian patients with SSc living in Asian countries suggest a more severe clinical phenotype in this racial group as well. Chinese patients with SSc had unique serological and clinical features, with higher frequencies of anti‐Scl‐70 antibodies, diffuse cutaneous disease, and pulmonary fibrosis but lower anti‐RNA polymerase III antibody frequency compared with US white patients from a prospective registry 13. In the Singapore SSc Cohort, Chinese patients with SSc had higher rates of the diffuse cutaneous subtype, an earlier age at diagnosis, more severe pulmonary arterial hypertension, and a different autoantibody profile (characterized by higher frequencies of anti‐Scl‐70, anti‐U1‐RNP, and anti‐SSA/Ro but lower anticentromere antibodies) compared with historical cohorts of US white patients 14. Thai patients have been shown to have higher frequencies of diffuse skin disease and positive anti‐Scl‐70 antibodies compared with white Australians from the same clinic 15. However, most of these studies were limited by the use of indirect racial comparisons, their cross‐sectional design, and lack of outcome assessment. One small study of North American white (n = 47), black (n = 15), and Japanese (n = 43) patients with SSc and anti‐Scl‐70 antibodies found that Japanese and black patients were more likely to develop severe lung disease and had higher mortality rates compared with white patients 3.
To date, limited research has been published on Asian patients with SSc living in the United States compared with patients of other racial/ethnic groups. In 2009, Schmajuk et al 16 performed a pilot cross‐sectional study to compare disease features by race in patients evaluated at Stanford University and affiliated hospitals. Asian Americans were less likely to have digital ulcers or anemia than white patients, but there were no other significant differences in disease manifestations. This study lacked follow‐up to evaluate survival and long‐term disease outcomes, and similar to the studies discussed previously, it was underpowered to capture important clinical differences between racial groups.
The purpose of this study was to evaluate racial/ethnic disparities in clinical characteristics, autoantibody profiles, and survival in patients with SSc. We used data from Kaiser Permanente Northern California (KPNC), a large integrated health care system with a well‐defined diverse patient population representative of the Northern California geographic region 17. This study setting offered a robust Asian American population, follow‐up on outcomes, and direct comparisons across racial/ethnic groups. Such comparisons may provide insight into the relative importance of race and genetic background versus environmental factors in the onset and expression of SSc. Our findings may also guide appropriate treatment, monitoring, and prognostication among different groups in the context of an increasing racial diversity in the US population.
MATERIALS AND METHODS
Setting
KPNC is a large demographically diverse integrated health care delivery system caring for more than 4.2 million members in Northern California (including Medicare beneficiaries), composing one‐third of the insured population in the area 17. This insured population is socially, economically, and racially/ethnically diverse 17. Members within KPNC receive comprehensive health care services, including preventive, ambulatory, emergency, and inpatient care, as well as laboratory, radiology, and pharmacy services. Members are cared for by physicians from the Permanente Medical Group, a self‐governed, physician‐led, prepaid, multispecialty medical group. This study was approved by the Kaiser Foundation Research Institute and Stanford University Institutional Review Boards, with waiver of signed consent.
Study design and population
We performed a retrospective cohort study using electronic medical records for all eligible KPNC adults with an incident diagnosis of SSc by a rheumatologist between January 1, 2007, and December 31, 2016. The start date was chosen after the regional implementation of electronic medical records at KPNC in 2006. Participants were first identified using the International Classification of Diseases, Ninth Revision (ICD‐9) and International Classification of Diseases, 10th Revision (ICD‐10) diagnosis codes for SSc (ICD‐9: 710.1; ICD‐10: M34) during the study period and were required to have at least one clinic visit with a rheumatologist. A subsequent medical record review was performed to ensure that only patients who fulfilled 2013 American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) classification criteria for definite SSc 18 were included in this study. Self‐reported race/ethnicity was categorized as: non‐Hispanic white, Asian (including Pacific Islander), black, and Hispanic white (excluding Asian and black) 19. Multiracial and other racial subgroups were excluded from this study because we did not have ancestry data. Incident cases were defined as patients with a new diagnosis of SSc at KPNC from 2007 onward (confirmed by a medical record review), and patients were required to have KPNC membership for a minimum of 12 months prior to SSc diagnosis. Prevalent cases (defined as having an SSc diagnosis prior to 2007 or carrying a diagnosis from an outside rheumatologist prior to KPNC membership) were excluded to avoid immortal time bias.
Clinical variables and autoantibodies
Data were obtained by electronic database extraction from KPNC's clinical electronic health record. Demographic information was collected, including sex, race/ethnicity, smoking status, and body mass index (BMI) at diagnosis. Disease subtype based on the extent of skin involvement (limited vs. diffuse cutaneous SSc), age at diagnosis, and age at symptom onset were collected. Date of diagnosis was defined as the date of first SSc diagnosis definitively made by a KPNC rheumatologist, according to an electronic medical record review of physician documentation.
Clinical manifestations of SSc were collected by organ system through electronic medical record review. Given that disease features may vary with time, we characterized the involvement of each organ system by its most severe expression at any recorded visit:
Vascular involvement included telangiectasias, Raynaud's phenomenon, digital gangrene, digital ulcers, and calcinosis.
Musculoskeletal involvement included joint pain, joint contractures, and myositis.
Cardiac involvement included congestive heart failure (defined as diastolic heart failure or systolic heart failure with impaired left ventricular function with an ejection fraction less than 50% on echocardiogram).
Pulmonary disease included interstitial lung disease (diagnosed by a pulmonologist or radiologist with high‐resolution computed tomography [CT] of the chest), chronic oxygen dependence, or history of lung transplantation. Suspected pulmonary hypertension was denoted for all patients with an estimated right ventricular systolic pressure greater than or equal to 40 mm Hg on Doppler echocardiography, as previously reported 20. For the subset of patients whose physician pursued a right heart catheterization (RHC) evaluation, pulmonary arterial hypertension was defined by a resting mean pulmonary artery pressure greater than or equal to 25 mm Hg in the setting of a pulmonary capillary wedge pressure less than or equal to 15 mm Hg.
Gastrointestinal involvement was noted for symptomatic gastroesophageal reflux, malabsorption syndrome, or small intestinal bacterial overgrowth.
Renal involvement included scleroderma renal crisis, kidney transplant, or dialysis dependence.
We documented the following serologic autoantibody results for each study subject: antinuclear antibody (ANA) with immunofluorescent staining pattern (positive value defined as ≥1:80), anticentromere, anti‐Scl‐70 (topoisomerase I), anti‐SSA/Ro, anti‐SSB/La, and anti‐U1‐RNP (ribonucleoprotein). RNA polymerase III was excluded from analysis because of excessive missing data. Although serologic antibodies may vary with time, we considered an individual seropositive if antibodies exceeded standard laboratory ranges at any point in time. Data for the ANA pattern were collected when available, namely if indirect immunofluorescence was ordered at the discretion of the individual rheumatologist.
Ever usage of immunosuppressant medications for each patient with SSc was abstracted from physician notes and pharmacy data and included mycophenolate mofetil, methotrexate, cyclophosphamide, azathioprine, low‐dose and high‐dose prednisone (defined as less than or equal to 20 mg and greater than 20 mg, respectively), intravenous immunoglobulin (IVIG), and rituximab. Prescribing practices for immunosuppressant medications were up to the discretion of the treating rheumatologist and/or pulmonologist.
Outcomes
The primary outcome of interest was overall survival rates by racial/ethnic group, measured from the date of SSc diagnosis. Primary causes of death for all deceased patients were ascertained through an electronic chart review of hospital discharge summaries or physician death notes and categorized as previously defined 21.
Statistical analysis
Descriptive statistics (frequencies, proportions, means, and medians) were used to describe the demographic and clinical characteristics of the four racial/ethnic groups. χ2 tests or Fisher's exact tests were used to compare categorical variables, as appropriate. Analysis of variance was used to compare the means for continuous variables across the four racial/ethnic groups. When the overall comparisons among the four racial/ethnic groups were statistically significant, we performed multiple comparisons using the Fisher method of adjusting the P value to assess paired comparisons for significance (implemented using the PROC MULTTEST procedure in SAS [SAS Institute, Inc.]).
The Kaplan‐Meier method was used to compare 5‐year and 9‐year survival rates between white, Asian, black, and Hispanic patients. Participants were censored at the earliest of the following: lost to follow‐up (defined as end of KPNC membership), death, or end of study period (December 31, 2016). Differences in survival were assessed by the log‐rank test. Univariable and multivariable Cox regression analyses were performed to determine which demographic and clinical variables were predictive of death. The proportionality of hazards assumption was met. Predictors used in the model included those that were statistically significant in the univariable analysis as well as those generally accepted in the literature to be potential predictors of mortality. Cases that were missing variables in the multivariable model were dropped from analysis, leaving only cases with complete data available. All analyses were conducted using SAS software version 9.4 (SAS Institute Inc.), with the threshold of significance set at P is less than 0.05.
RESULTS
Study population characteristics
A total of 609 KPNC adult incident patients with SSc were identified (Table 1). Most were female (89%) and had limited cutaneous SSc (81%). Mean age at diagnosis was 55.4 ± 14.8 years and mean disease duration from first non‐Raynaud symptom to SSc diagnosis was 5 ± 5 years. The racial/ethnic distribution was composed of 51% white (n = 310), 25% Hispanic (n = 154), 16% Asian (n = 96), and 8% black (n = 49) patients. Self‐reported ancestry of Asian patients included the following in descending frequency: Filipino (n = 24; 25.0%), Chinese (n = 23; 24.0%), Indian (n = 13; 13.5%), Japanese (n = 5; 5.2%), Vietnamese (n = 5; 5.2%), Korean (n = 4; 4.2%), Pacific Islander (n = 3; 3.1%), and not specified (n = 19; 19.8%).
Table 1.
Baseline demographics of SSc cohort by race/ethnicity
| Characteristics (n = Patients With Complete Data) | Overall (N = 609) | Asian (n = 96; 16%) | Black (n = 49; 8%) | Hispanic (n = 154; 25%) | White (n = 310; 51%) | P a |
|---|---|---|---|---|---|---|
| Female sex (n = 609), n (%) | 544/609 (89.3) | 85/96 (88.5) | 39/49 (79.6) | 141/154 (91.6) | 279/310 (90.0) | 0.117 |
| Diffuse (n = 609), n (%) | 116/609 (19.0) | 21/96 (21.9) | 22/49 (44.9) ** | 28/154 (18.2) | 45/310 (14.5) | <0.001 |
| BMI (kg/ma) (n = 560), mean ± SD | 26.6 ± 5.9 | 23.9 ± 4.9 ** | 28.5 ± 6.2 | 27.7 ± 5.2 | 26.6 ± 5.3 | <0.01 |
| Ever smoker (n = 568), n (%) | 187/568 (32.9) | 16/92 (17.4) ** | 20/48 (41.7) | 29/146 (19.9) ** | 122/282 (43.3) | <0.001 |
| Age at diagnosis of SSc by rheumatologist (n = 609), y, mean ± SD | 55.4 ± 14.8 | 51.7 ± 16.3 ** | 53.1 ± 12.9 * | 52.8 ± 14.5 ** | 58.9 ± 14.1 | <0.001 |
| Time from first non‐Raynaud symptom to SSc diagnosis (n = 529), y, mean ± SD | 4.9 ± 3.8 | 4.2 ± 3.5 | 3.9 ± 1.8 | 4.2 ± 2.8 | 5.5 ± 4.5 | 0.168 |
| Positive ANA (n = 578), n (%)b | 508/572 (88.9) | 81/93 (87.1) | 40/44 (90.9) | 143/152 (94.1) | 244/283 (86.2) | 0.131 |
| Nuclear ANA (n = 379), n (%) | 54/379 (14.2) | 8/57 (14.0) | 10/36 (27.8) | 10/99 (10.1) | 26/187 (13.9) | 0.078 |
| Anticentromere (n = 543), n (%) | 291/543 (53.6) | 42/92 (45.7) | 12/42 (28.6) | 92/143 (64.3) | 145/266 (54.5) | <0.001 |
| Anti‐SCL‐70 (n = 510), n (%) | 98/510 (19.2) | 23/90 (25.6) | 10/45 (22.2) | 27/138 (19.6) | 38/237 (16.0) | 0.246 |
| Anti‐SSA/Ro (n = 467), n (%) | 69/467 (14.8) | 18/81 (22.2) | 5/36 (13.9) | 19/126 (15.1) | 27/224 (12.1) | 0.178 |
| Anti‐SSB/La (n = 463), n (%) | 19/463 (4.1) | 7/79 (8.9) | 1/37 (2.7) | 4/126 (3.2) | 7/221 (3.2) | 0.139 |
| Anti‐U1RNP (n = 460), n (%) | 77/460 (16.7) | 26/81 (32.1) ** | 4/38 (10.5) | 21/123 (17.1) | 26/218 (11.9) | <0.001 |
| SLE overlap, n (%) | 21/77 (27.3) | 6/26 (23.1) | 2/4 (50.0) | 10/21 (47.6) | 3/26 (11.5) | 0.031 |
Abbreviation: ANA, antinuclear antibody; BMI, body mass index; SLE, systemic lupus erythematosus; SSc, systemic sclerosis.
χ2 test or Fisher's exact test for categorical variables; one‐way analysis of variance for continuous variables.
Positive ANA by immunofluorescence (≥1:80) (n = 379) or BioPlex 2200 ANA screen (n = 199). For nuclear ANA, denominators represent patients with available data.
*P < 0.05 when comparing with white patients; **P < 0.01 when comparing with white patients.
Statistically significant values (defined as P < 0.05) are in bold.
The mean age at SSc diagnosis for white patients (58.9 ± 14.1 years) was significantly older compared with all other groups. There were no significant differences in disease duration by race/ethnicity. In pairwise comparisons, black patients had a higher prevalence of diffuse cutaneous SSc (44.9% vs. 14.5%; P < 0.001) compared with white patients. Asian (17.4% vs. 43.3%; P < 0.001) and Hispanic patients (19.9% vs. 43.3%; P < 0.001) were significantly less likely to have ever smoked compared with white patients. Asian patients had the lowest mean BMI of all groups.
The majority of the cohort had at least one echocardiogram to evaluate for pulmonary hypertension (n = 508; 83.4%) and/or at least one high‐resolution chest CT scan to evaluate for interstitial lung disease (n = 570; 93.6%). Of patients with suspected pulmonary hypertension on echocardiogram (n = 154), approximately one‐third (n = 50) underwent RHC to evaluate for pulmonary arterial hypertension. Thirty patients had a confirmed diagnosis of pulmonary arterial hypertension on RHC (4.9% of the entire cohort). Decision to proceed with RHC was up to the discretion of the physician team and patient. Common reasons for declining RHC included borderline echocardiogram results, invasive interventions declined by patient, medical comorbidities, and/or potential risks of the procedure.
Autoantibodies
Asian patients had a significantly higher prevalence of anti‐U1‐RNP antibodies compared with white patients (32.1% vs. 11.9%; P = 0.005). Across all racial/ethnic groups, Asian patients had the highest prevalence of anti‐Scl‐70 (25.6%) and anti‐SSA/Ro antibodies (22.2%), whereas black patients had the lowest prevalence of anticentromere antibodies (28.6%), but these results were not statistically significant when adjusting for pairwise comparisons.
Organ involvement
Clinical manifestations in Asian, black, and Hispanic patients were not significantly different from those in white patients after adjusting for multiple pairwise comparisons (Table 2).
Table 2.
Organ involvement of systemic sclerosis cohort by race/ethnicity
| Characteristics (n = Patients With Complete Data) | All (N = 609), n (%) | Asian (n = 96), n (%) | Black (n = 49), n (%) | Hispanic (n = 154), n (%) | White (n = 310), n (%) | P a |
|---|---|---|---|---|---|---|
| Vascular | ||||||
| Telangiectasias (n = 496) | 376/496 (75.8) | 48/71 (67.6) | 20/36 (55.6)c | 97/129 (75.2) | 211/260 (81.2) | 0.002 |
| Raynaud phenomenon (n = 603) | 583/603 (96.7) | 93/96 (96.9) | 42/48 (87.5)c | 144/150 (96.0) | 304/309 (98.4) | 0.004 |
| Digital gangrene (n = 609) | 48/609 (7.9) | 4/96 (4.2) | 6/49 (12.8) | 15/154 (9.8) | 23/310 (7.4) | 0.239 |
| Digital ulcers (n = 437) | 205/437 (46.9) | 33/62 (53.2) | 22/35 (62.9) | 59/106 (55.7) | 91/234 (38.9) | 0.003 |
| Calcinosis (n = 608) | 140/608 (23.0) | 11/96 (11.5) | 9/48 (18.8) | 39/154 (25.3) | 81/310 (26.1) | 0.020 |
| Musculoskeletal | ||||||
| Joint pain (n = 593) | 460/593 (77.6) | 71/94 (75.5) | 42/49 (85.7) | 122/152 (80.3) | 225/298 (75.5) | 0.326 |
| Joint contractures (n = 608) | 154/608 (25.3) | 22/96 (22.9) | 13/49 (26.5) | 39/153 (25.5) | 80/310 (25.8) | 0.946 |
| Myositis (n = 608) | 48/608 (7.9) | 6/96 (6.3) | 5/48 (10.4) | 11/154 (7.1) | 26/310 (8.4) | 0.801 |
| Cardiopulmonary | ||||||
| Congestive heart failure (n=515) | 97/515 (18.8) | 12/84 (14.3) | 8/37 (21.6) | 20/133 (15.0) | 57/261 (21.8) | 0.249 |
| Suspected pulmonary hypertension on echocardiogram (n = 508) | 154/508 (25.3) | 26/81 (32.1) | 16/40 (40.0) | 34/125 (27.2) | 78/262 (29.7) | 0.474 |
| Pulmonary arterial hypertension (n = 50)b | 30/50 (60.0) | 4/8 (50.0) | 3/5 (60.0) | 7/10 (70.0) | 16/26 (61.5) | 0.867 |
| Interstitial lung disease (n = 570) | 219/570 (38.4) | 36/85 (42.4) | 26/46 (56.5) | 56/144 (38.9) | 101/295 (34.2) | 0.028 |
| Oxygen dependence (n = 608) | 83/608 (13.7) | 13/96 (13.5) | 10/49 (20.4) | 22/154 (14.3) | 38/309 (12.3) | 0.488 |
| Gastrointestinal | ||||||
| Gastroesophageal reflux (n = 574) | 500/574 (87.1) | 74/85 (87.1) | 41/44 (93.2) | 133/148 (89.9) | 252/297 (84.9) | 0.284 |
| Malabsorption syndrome or small intestinal bacterial overgrowth (n = 603) | 44/603 (7.3) | 7/94 (7.5) | 3/49 (6.1) | 16/152 (10.5) | 18/308 (5.8) | 0.333 |
| Renal | ||||||
| Renal crisis (n = 609) | 19/609 (3.1) | 5/96 (5.2) | 1/49 (2.0) | 5/153 (3.3) | 8/310 (2.6) | 0.598 |
| Dialysis dependence (n = 603) | 14/603 (2.3) | 1/95 (1.1) | 4/48 (8.3) | 5/153 (3.3) | 4/307 (1.3) | 0.016 |
χ2 test or Fisher's exact test for categorical variables.
Subset of patients who pursued right heart catheterization.
P < 0.10 when comparing with white patients.
Statistically significant values (defined as P < 0.05) are in bold.
Survival analyses
Cumulative death for the entire cohort was 17.2% (105 deaths/609 patients) over a median follow‐up time of 5 years. Cumulative death by race for white, Hispanic, Asian, and black patients was 48/310 (15.5%), 23/154 (14.9%), 22/96 (22.9%), and 12/49 (24.5%), respectively. The total proportion of patients censored because of KPNC membership ending prior to study end was 14.5%, whereas 68.3% of patients were alive at study end. There were no statistically significant differences in types of censorship by race/ethnicity (Supplementary Table 1). The median (interquartile range) follow‐up time for white, Hispanic, Asian, and black patients was 5.1 (2.3‐8.4), 4.5 (1.8‐8.1), 4.7 (2.0‐7.9), and 5.4 (2.4‐9.9) years, respectively.
Kaplan‐Meier survival analyses showed that Asian and black patients had increased overall mortality compared with white and Hispanic patients (Figure 1). Five‐year overall survival rates from the date of SSc diagnosis for white, Hispanic, Asian, and black patients were 85.3%, 86.4%, 76.4%, and 76.2%, respectively. Nine‐year overall survival rates from the date of diagnosis for white, Hispanic, Asian, and black patients were 75.8%, 68.2%, 52.3%, 52.2%, respectively (P = 0.057).
Figure 1.
Kaplan‐Meier Survival Curve by Race/Ethnicity from Diagnosis of Systemic Sclerosis.
Asian (hazard ratio [HR] 1.83 [95% confidence interval (CI) 1.08‐2.99]; P = 0.020) and black race (HR 1.69 [95% CI 0.86‐3.08]; P = 0.104) were associated with increased mortality compared with white race in unadjusted analysis (N = 609; Table 3). A multivariable Cox regression model adjusting for demographic variables including age at diagnosis, race/ethnicity, and smoking (n = 568) showed that Asian (HR 2.42 [95% CI 1.38‐4.14]; P = 0.001) and black patients (HR 2.1 [95% CI 1.05‐3.89]; P = 0.026) still had an increased risk of mortality compared with white patients.
Table 3.
Univariable and multivariable Cox regression analysis of variables associated with overall mortalitya
| Variables | Univariable Analysis | Multivariable Analysis | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P | HR | 95% CI | P | |
| Age at diagnosis, y (continuous) | 1.05 | 1.03‐1.07 | <0.001 | 1.05 | 1.03‐1.07 | <0.001 |
| Sex, female vs. male | 0.45 | 0.28‐0.76 | 0.002 | 0.55 | 0.31‐1.02 | 0.045 |
| BMI < 25 kg/m2 | 1.17 | 0.73‐1.90 | 0.426 | 1.05 | 0.65‐1.71 | 0.843 |
| BMI ≥ 30 kg/m2 | 1.03 | 0.56‐1.82 | 0.007 | 0.80 | 0.43‐1.46 | 0.474 |
| SSc subtype, diffuse vs. limited | 2.30 | 1.51‐3.43 | <0.001 | 1.81 | 1.05‐3.06 | 0.029 |
| Ever smoker | 1.46 | 0.97‐2.16 | 0.063 | 1.09 | 0.69‐1.68 | 0.131 |
| Asianb | 1.83 | 1.08‐2.99 | 0.020 | 1.80 | 0.99‐3.16 | 0.047 |
| Blackb | 1.69 | 0.86‐3.08 | 0.104 | 1.21 | 0.55‐2.46 | 0.619 |
| Hispanicb | 1.03 | 0.62‐1.68 | 0.895 | 1.06 | 0.60‐1.81 | 0.846 |
| Pulmonary hypertension | 3.74 | 2.51‐5.63 | <0.001 | 2.33 | 1.48‐3.68 | <0.001 |
| Interstitial lung disease | 2.29 | 1.56‐3.39 | <0.001 | 1.68 | 1.07‐2.65 | 0.025 |
| Scleroderma renal crisis | 3.58 | 1.45‐7.21 | 0.001 | 1.58 | 0.69‐3.53 | 0.309 |
| Malabsorption syndrome | 2.59 | 1.53‐4.17 | <0.001 | 2.11 | 1.17‐3.60 | 0.009 |
Abbreviation: BMI, body mass index; CI, confidence interval; HR, hazard ratio; SSc, systemic sclerosis.
Statistically significant values are in bold.
White patients as the reference group.
After adjusting for the aforementioned variables and SSc‐related covariates (including cutaneous subtype [diffuse vs. limited], BMI, pulmonary hypertension on echocardiogram, interstitial lung disease, scleroderma renal crisis, and malabsorption syndrome) (n = 505), Asian patients still had an increased risk of mortality (HR 1.80 [95% CI 0.99‐3.16]; P = 0.047) (Table 3) on multivariable Cox regression. Other variables associated with higher mortality included older age at SSc diagnosis, male sex, diffuse cutaneous disease, suspected pulmonary hypertension on echocardiogram, interstitial lung disease, and malabsorption syndrome.
Leading overall causes of death (as determined by the treating physician) included highly suspected or confirmed pulmonary hypertension, multiorgan failure, infection, and cancer (Table 4). Of the 16 patients who had a primary cause of pulmonary hypertension listed as the cause of death, half had an RHC that confirmed a diagnosis of pulmonary arterial hypertension and the other half had strong clinical suspicion for pulmonary hypertension, such as a reduced global right ventricular systolic function or right ventricular hypertrophy on echocardiogram and/or clinical signs of right heart failure requiring medical management. Asian patients were most likely to die from pulmonary hypertension and infections; black patients, multiorgan failure; Hispanic patients, pulmonary hypertension; and white patients, pulmonary fibrosis. However, the numbers of deaths in each category were small.
Table 4.
Causes of death by race/ethnicity for systemic sclerosis cohort
| Overall (N = 105) | Asian (n = 22) | Black (n = 12) | Hispanic (n = 23) | White (n = 48) | |
|---|---|---|---|---|---|
| Causes of death, n (%) | |||||
| Atherosclerotic cardiovascular or cerebrovascular disease | 7 (6.7) | 1 (4.5) | … | 1 (4.3) | 5 (10.4) |
| Cancer | 12 (11.4) | 2 (9.1) | 3 (25.0) | 2 (8.7) | 5 (10.4) |
| Gastrointestinal | 4 (3.8) | … | 1 (8.3) | 2 (8.7) | 1 (2.1) |
| Infection | 13 (12.4) | 5 (22.7) | … | 2 (8.7) | 6 (12.5) |
| Multiorgan failure | 16 (15.2) | 4 (18.2) | 4 (33.3) | 3 (13.0) | 5 (10.4) |
| Pulmonary fibrosis | 11 (10.5) | 1 (4.5) | 1 (8.3) | 2 (8.7) | 7 (14.6) |
| Pulmonary hypertension | 16 (15.2) | 5 (22.7) | 1 (8.3) | 4 (17.4) | 6 (12.5) |
| Renal crisis | 1 (1.0) | … | … | 1 (4.3) | … |
| Sudden death | 3 (2.9) | 1 (4.5) | … | 1 (4.3) | 1 (2.1) |
| Other | 11 (10.5) | 2 (9.1) | … | 3 (13.0) | 6 (12.5) |
| Unknown | 11 (10.5) | 1 (4.5) | 2 (16.7) | 2 (8.7) | 6 (12.5) |
| Cumulative death (N = 609), n (%) | 105/609 (17.2) | 22/96 (22.9) | 12/49 (24.5) | 23/154 (14.9) | 48/310 (15.5) |
Medication use
There were no statistically significant differences in exposure to medication by race/ethnicity for mycophenolate mofetil, methotrexate, cyclophosphamide, prednisone, or rituximab (Table 5). Compared with white patients, black patients had increased azathioprine exposure (16% vs. 9%) and Asian patients had increased IVIG use (7% vs. 2%), but the results were not statistically significant after adjusting for pairwise comparisons, and the frequencies were small.
Table 5.
Exposure to immunosuppressant medications in systemic sclerosis cohort by race/ethnicity (N = 609)
| Overall (N = 609), % | Asian (n = 96), % | Black (n = 49), % | Hispanic (n = 154), % | White (n = 310), % | P | |
|---|---|---|---|---|---|---|
| Azathioprine | 11 | 12 | 16 | 14 | 9 | 0.035 |
| Cyclophosphamide | 7 | 8 | 10 | 7 | 6 | 0.651 |
| IVIG | 3 | 7 | 2 | 3 | 2 | 0.056 |
| Methotrexate | 27 | 28 | 30 | 27 | 25 | 0.870 |
| Mycophenolate mofetil | 24 | 26 | 32 | 29 | 20 | 0.077 |
| Prednisone (maximum) | ||||||
| Low dose (≤20 mg) | 43 | 35 | 52 | 46 | 42 | 0.187 |
| High dose (>20 mg) | 26 | 29 | 24 | 22 | 27 | 0.594 |
| Rituximab | 4 | 6 | 9 | 5 | 3 | 0.170 |
Abbreviation: IVIG, intravenous immunoglobulin.
Statistically significant values (defined as P < 0.05) are in bold.
Missing data
There were no statistically significant differences in missingness of variables assessing baseline demographics, disease duration, internal organ involvement, or medication use by race/ethnicity. Overall, there were more missing data on autoantibodies, but the data still exceeded 70% completeness in each racial/ethnic group, and autoantibodies were not included in the Cox regression models.
DISCUSSION
Racial/ethnic differences in clinical features, autoantibody profiles, and survival among KPNC patients with SSc were demonstrated in this study. Importantly, we found that Asian patients had a more severe clinical phenotype than non‐Hispanic white patients. Similar to prior studies, Asian patients in our cohort had a younger age at disease diagnosis 14, 22 and higher prevalence of diffuse cutaneous disease 13, 15, 22, interstitial lung disease 3, 13, 22, 23, and anti‐Scl‐70 antibodies 13, 15, 22 compared with white patients. Consistent with prior studies, Asian patients were more likely to test positive for anti‐U1‐RNP and anti‐SSA/Ro antibodies 14, 24. In this cohort, we found that Asian patients had an increased mortality compared with white and Hispanic patients, with survival rates similar to those of black patients. Asian race was associated with a higher mortality even after adjusting for multiple confounders, suggesting that these patients warrant close monitoring for progressive disease and potentially more aggressive management. Pulmonary hypertension was one of the leading causes of death in Asian patients, indicating a severe vascular phenotype in this group as well. Immunosuppressant medication use was generally similar among all groups, except for slightly higher rates of azathioprine and rituximab exposure in black patients and of IVIG exposure in Asian patients. Given the variations in practice among different providers, we are unable to comment on reasons for treatment differences. However, these differences were not statistically significant on pairwise comparisons, and the numbers were small.
The variability of disease expression among racial groups in our study suggests that multiple factors linked to race, including genetic and environmental factors, may influence clinical manifestations, disease course, autoantibody status, and mortality in SSc. There is likely a complex interplay between multiple genetic and environmental factors contributing to the increased disease severity and mortality that we observed in Asian patients in our cohort. The causes for population variation in health outcomes in our study are likely multifactorial and may include biological, environmental, and social determinants.
Genetic heterogeneity among races may significantly impact the complex phenotype of SSc and provides a biologic basis for variations in disease severity. Multiple studies have shown that genetic differences between racial groups play an important role in modulating clinical manifestations, disease progression, and autoantibody profiles in SSc. In the multicenter Genome Research in African American Scleroderma Patients (GRASP) cohort, the largest genetic study in black patients with SSc to date, there was no association between increased SSc susceptibility and genes previously implicated in European Americans (including the ATP8B4 gene) 25. Multiple HLA antigen alleles have also been strongly associated with SSc and its clinical or serological subsets, and these associations have been found to vary in different racial/ethnic populations. For example, Han Chinese patients with SSc were more likely to have HLA‐DPB1*03:01 and to develop pulmonary fibrosis 26; Thai SSc patients had increased allele frequencies of HLA‐DR*15 associated with positive anti‐Scl‐70 antibodies 27; and DRB1*15 was associated with anti‐Scl‐70–positive Chinese patients with SSc, consistent with South African and Korean patients with SSc 28. Genetic studies have also identified small nucleotide polymorphism haplotypes in the fibrillarin‐1 gene that strongly associate with SSc susceptibility in Choctaw (a homogeneous group with a high prevalence and severe phenotype of SSc) and Japanese patients 2. These studies highlight the identification of genetic factors that may contribute to the biologic pathways affecting SSc susceptibility and expression. Further genomic research through large multicenter, multiracial cohorts is needed to investigate why certain populations have a higher susceptibility for disease and predisposition for more severe clinical manifestations. A better understanding of the biologic basis for genetic differences observed in SSc will hopefully lead to the development of targeted therapies for particular subpopulations.
The impact of environmental factors and social determinants on mortality in SSc warrants further investigation because prior studies have shown conflicting results. Some studies have demonstrated that socioeconomic status, including education level and income, has no impact on incidence or mortality in SSc 29, 30, and the relative risk of death remained higher in black patients compared with white patients in a US study even after adjusting for multiple covariates, including insurance status 5. In contrast, Moore et al 31 recently reported that black patients with SSc had more severe pulmonary disease and higher unadjusted mortality compared with matched non–African American patients, but race was not a significant risk factor for mortality after adjusting for socioeconomic factors. In a recent study evaluating a large and diverse cohort of Canadian patients with SSc, there was no difference in short‐term survival across ethnicities (European‐descended white [n = 745], Afro‐Caribbean [n = 58], South Asian [n = 70], East Asian [n = 80], Hispanic [n = 30], Arab [n = 9], First Nations [n = 7], and Persian [n = 6]), but East Asian patients interestingly had the longest median survival 32. Unlike most studies (including ours), their cohort occurred in the setting of a publicly funded universal health care system, where all citizens and permanent residents are eligible for public health insurance.
The same Canadian group also found no ethnic differences in survival in systemic lupus erythematosus (SLE) at their center, in contrast to differences reported at other centers without universal health care 33. The authors suggested that previously observed racial/ethnic disparities in survival may be partially attributed to access‐to‐care issues or health inequities. It is certainly possible that universal health care may improve access to care and decrease health inequities/disparities, resulting in overall more homogeneous outcomes. Our cohort at KPNC was unique because all members were insured, including those covered by Medicaid and Medicare, and had established care with a general rheumatologist, thereby allowing relatively equal access to medical services. Furthermore, as an integrated health care system, Kaiser Permanente has been cited as a leader in the field of health equity initiatives, with several formalized quality improvement programs aimed at reducing health disparities in the general population 34. However, we acknowledge that even within the KPNC insured population, it is still possible that differential access to care across races/ethnicities could still occur and potentially influence SSc disease course and mortality.
The impact of immigration patterns, generational status (foreign‐born vs. native‐born), and acculturation of immigrants on the clinical expression of SSc is also not known. Long‐term survival could potentially be influenced by cultural variations in health‐seeking behaviors, adherence, and cultural medical practices. In the general population, data from the US Department of Health and Human Services suggest that health, life expectancy, mortality, and morbidity patterns for immigrants and native‐born individuals vary considerably in the United States, with health patterns varying across racial/ethnic groups 35. A comparative study of Hispanic patients with SLE residing in the United States and Mexico found that Mexican patients were more educated, had better health‐related quality of life, and had overall less systemic SLE manifestations 36. The discrepant findings observed between these two groups of Hispanic patients with SLE in different geographic locations may reflect socioeconomic or biological factors, but this has not been studied extensively in SSc. Given the global phenomenon of immigration, further research in SSc is needed to understand the overall course and outcome of various racial/ethnic groups to prevent undesirable outcomes.
Our study has many strengths. First, our study population at KPNC is racially/ethnically diverse and closely approximates the general population of Northern California 17. This setting offers a unique opportunity to study the impact of race on SSc, allowing direct racial/ethnic comparisons from the same cohort instead of using historical controls. Second, we reviewed all individual electronic medical records to confirm incident SSc diagnoses and to obtain patient‐level data, including organ involvement and medication usage. Finally, because KPNC is an integrated health care system, we had longitudinal follow‐up survival data from both inpatient and outpatient settings.
Our study has some limitations. Our study population was cared for by multiple rheumatologists in a community‐based health care organization; thus, our results may not generalize to other health care settings. In particular, this study did not include the uninsured population, patients who receive health care in a fee‐for‐service or nonintegrated setting, or those referred to academic medical centers in the United States specializing in the care of patients SSc. It is possible that patients with milder disease expression were more likely to be seen within KPNC compared to medically complex cases referred to academic medical centers. However, even in a relatively homogeneous population of insured patients, we found significant racial/ethnic differences in disease severity and mortality; these disparities could potentially be amplified in other health care settings but requires further investigation. Given the large number of variables per patient in this study, we only collected organ complications as binary categorical variables (ever or never); thus, we were unable to run the Cox regression analysis using time‐varying covariates. As with the nature of retrospective studies, missing data were unavoidable, including data on RNA polymerase III and assessment of the severity of skin disease. Insufficient power precluded comparisons among individual Asian subgroups based on analysis of country and ethnic origin. Finally, it was not within the scope of this study to ascertain types of insurance coverage for each member of the study cohort, which also may have changed over time regarding deductibles and copayments. Our study did not collect information on socioeconomic status, including education level, income, or primary language spoken in the household.
In conclusion, clinical characteristics and outcomes in patients with SSc varied by racial/ethnic background within the KPNC population, with poor survival in Asian and black patients. Asian race was independently associated with a higher mortality, with pulmonary hypertension being a leading cause of death in this group. Close surveillance for disease progression and screening for pulmonary hypertension should be considered in this high‐risk population. Further research is needed to validate our findings in other cohorts and to identify both genetic and environmental determinants of SSc disease expression in different racial/ethnic groups. Understanding racial/ethnic variations in disease outcomes is critical because this will improve our ability to provide personalized medicine by informing patient counseling, appropriate baseline screening, and monitoring of disease progression.
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design
M. P. Chung, Dontsi, Postlethwaite, Simard, Zaba, L. Chung.
Acquisition of data
M. P. Chung, Dontsi, Postlethwaite, Kesh, Zaba, L. Chung.
Analysis and interpretation of data
M. P. Chung, Dontsi, Postlethwaite, Kesh, Simard, Fiorentino, Zaba, L. Chung.
Supporting information
ACKNOWLEDGMENTS
We thank Kristin Sainani, PhD, for reviewing a draft of the manuscript and Chi‐Mei Liu, PhD, for her research assistance with the KPNC Institutional Review Board.
Funding support was provided by Kaiser Permanente Northern California Graduate Medical Education Research. Dr. M. P. Chung's work was supported by the Training Program in Adult and Pediatric Rheumatology (5T32AR050942‐12). Dr. L. Chung's work was supported by the Scleroderma Research Foundation.
Melody P. Chung, MD: Kaiser Permanente Santa Clara, Santa Clara, California, and Stanford University School of Medicine, Palo Alto, California; 2Makdine Dontsi, MS, Debbie Postlethwaite, RNP, MPH: Kaiser Permanente Northern California, Oakland; 3Sumana Kesh, MD: Kaiser Permanente Santa Clara, Santa Clara, California; 4Julia F. Simard, ScD, David Fiorentino, MD, PhD, Lisa C. Zaba, MD, PhD: Stanford University School of Medicine, Palo Alto, California; 5Lorinda Chung, MD, MS: Stanford University School of Medicine and Department of Veterans Affairs Palo Alto Health Care System, Palo Alto, California.
Ms. Postlethwaite is an investigator on an unrelated US Food and Drug Administration–mandated safety study funded by Bayer Global and administered through the Kaiser Foundation Research Institute. No other disclosures relevant to this article were reported.
REFERENCES
- 1. Chung L, Lin J, Furst DE, Fiorentino D. Systemic and localized scleroderma. Clin Dermatol 2006;24:374–92. [DOI] [PubMed] [Google Scholar]
- 2. Reveille JD. Ethnicity and race and systemic sclerosis: how it affects susceptibility, severity, antibody genetics, and clinical manifestations. Curr Rheumatol Rep 2003;5:160–7. [DOI] [PubMed] [Google Scholar]
- 3. Kuwana M, Kaburaki J, Arnett FC, Howard RF, Medsger TA Jr, Wright TM. Influence of ethnic background on clinical and serologic features in patients with systemic sclerosis and anti‐DNA topoisomerase I antibody. Arthritis Rheum 1999;42:465–74. [DOI] [PubMed] [Google Scholar]
- 4. Arnett FC, Gourh P, Shete S, Ahn CW, Honey RE, Agarwal SK, et al. Major histocompatibility Complex (MHC) class II alleles, haplotypes and epitopes which confer susceptibility or protection in systemic sclerosis: analyses in 1300 Caucasian, African‐American and Hispanic cases and 1000 controls. Ann Rheum Dis. 2010;69:822–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Gelber AC, Manno RL, Shah AA, Woods A, Le EN, Boin F, et al. Race and association with disease manifestations and mortality in scleroderma: a 20‐year experience at the Johns Hopkins Scleroderma Center and review of the literature. Medicine (Baltimore) 2013;92:191–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Laing TJ, Gillespie BW, Toth MB, Mayes MD, Gallavan RH Jr, Burns CJ, et al. Racial differences in scleroderma among women in Michigan. Arthritis Rheum 1997;40:734–42. [DOI] [PubMed] [Google Scholar]
- 7. Greidinger EL, Flaherty KT, White B, Rosen A, Wigley FM, Wise RA. African‐American race and antibodies to topoisomerase I are associated with increased severity of scleroderma lung disease. Chest 1998;114:801–7. [DOI] [PubMed] [Google Scholar]
- 8. Beall AD, Nietert PJ, Taylor MH, Mitchell HC, Shaftman SR, Silver RM, et al. Ethnic disparities among patients with pulmonary hypertension associated with systemic sclerosis. J Rheumatol 2007;34:1277–82. [PubMed] [Google Scholar]
- 9. Morgan ND, Shah AA, Mayes MD, Domsic RT, Medsger TA Jr, Steen VD, et al. Clinical and serological features of systemic sclerosis in a multicenter African American cohort: analysis of the genome research in African American scleroderma patients clinical database. Medicine (Baltimore) 2017;96:e8980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Nietert PJ, Silver RM. Patterns of hospital admissions and emergency room visits among patients with scleroderma in South Carolina, USA. J Rheumatol 2003;30:1238–43. [PubMed] [Google Scholar]
- 11. Nietert PJ, Silver RM, Mitchell HC, Shaftman SR, Tilley BC. Demographic and clinical factors associated with in‐hospital death among patients with systemic sclerosis. J Rheumatol 2005;32:1888–92. [PubMed] [Google Scholar]
- 12. Silver RM, Bogatkevich G, Tourkina E, Nietert PJ, Hoffman S. Racial differences between blacks and whites with systemic sclerosis. Curr Opin Rheumatol 2012;24:642–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Wang J, Assassi S, Guo G, Tu W, Wu W, Yang L, et al. Clinical and serological features of systemic sclerosis in a Chinese cohort. Clin Rheumatol 2013;32:617–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Ling AL, Gee TG, Giap LW, Cheng NS, Santosa A, Chan G, et al. Disease characteristics of the Singapore systemic sclerosis cohort. Proceedings of Singapore Healthcare 2013;22:8–14. [Google Scholar]
- 15. McNeilage LJ, Youngchaiyud U, Whittingham S. Racial differences in antinuclear antibody patterns and clinical manifestations of scleroderma. Arthritis Rheum 1989;32:54–60. [DOI] [PubMed] [Google Scholar]
- 16. Schmajuk G, Bush TM, Burkham J, Krishnan E, Chung L. Characterizing systemic sclerosis in Northern California: focus on Asian and Hispanic patients. Clin Exp Rheumatol 2009;27 Suppl 54:22–5. [PubMed] [Google Scholar]
- 17. Gordon NP. Similarity of the adult Kaiser Permanente membership in Northern California to the insured and general population in Northern California: statistics from the 2009 California Health Interview Survey. 2012. URL: https://divisionofresearch.kaiserpermanente.org/projects/memberhealthsurvey/SiteCollectionDocuments/chis_non_kp_2009.pdf.
- 18. Van den Hoogen F, Khanna D, Fransen J, Johnson SR, Baron M, Tyndall A, et al. 2013 classification criteria for systemic sclerosis: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum 2013;65:2737–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Evans L. Racial and ethnic categories and definitions for NIH diversity programs and for other reporting purposes. 2015. URL: https://grants.nih.gov/grants/guide/notice-files/not-od-15-089.html.
- 20. Manno RL, Wigley FM, Gelber AC, Hummers LK. Late‐age onset systemic sclerosis. J Rheumatol 2011;38:1317–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Steen VD, Medsger TA. Changes in causes of death in systemic sclerosis, 1972‐2002. Ann Rheum Dis 2007;66:940–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Janardana R, Nair AM, Surin AK, Prakash JA, Gowri M, Danda D. Unique clinical and autoantibody profile of a large Asian Indian cohort of scleroderma: do South Asians have a more aggressive disease? [original article]. Clin Rheumatol 2019;38:3179–87. [DOI] [PubMed] [Google Scholar]
- 23. Hu S, Hou Y, Wang Q, Li M, Xu D, Zeng X. Prognostic profile of systemic sclerosis: analysis of the clinical EUSTAR cohort in China. Arthritis Res Ther 2018;20:235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Low AH, Johnson SR, Lee P. Ethnic influence on disease manifestations and autoantibodies in Chinese‐descent patients with systemic sclerosis. J Rheumatol 2009;36:787–93. [DOI] [PubMed] [Google Scholar]
- 25. Gourh P, Remmers EF, Boyden SE, et al. Brief report: whole‐exome sequencing to identify rare variants and gene networks that increase susceptibility to scleroderma in African Americans. Arthritis Rheumatol 2018;70:1654–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Wang J, Guo X, Yi L, Guo G, Tu W, Wu W, et al. Association of HLA‐DPB1 with scleroderma and its clinical features in Chinese population. PLoS One 2014;9:e87363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Louthrenoo W, Kasitanon N, Wichainun R, Wangkaew S, Sukitawut W, Ohnogi Y, et al. Association of HLA‐DRB1*15:02 and DRB5*01:02 allele with the susceptibility to systemic sclerosis in Thai patients. Rheumatol Int 2013;33:2069–77. [DOI] [PubMed] [Google Scholar]
- 28. He D, Wang J, Yi L, Guo X, Guo S, Guo G, et al. Association of the HLA‐DRB1 with scleroderma in Chinese population. PLoS One 2014;9:e106939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Royle JG, Lanyon PC, Grainge MJ, Abhishek A, Pearce FA. The incidence, prevalence, and survival of systemic sclerosis in the UK Clinical Practice Research Datalink [published erratum appears in Clin Rheumatol 2018;37:2601]. Clin Rheumatol 2018;37:2103–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Mansour S, Bonner A, Muangchan C, Hudson M, Baron M, Pope JE, et al. Low socioeconomic status (measured by education) and outcomes in systemic sclerosis: data from the Canadian Scleroderma Research Group. J Rheumatol 2013;40:447–54. [DOI] [PubMed] [Google Scholar]
- 31. Moore DF, Kramer E, Eltaraboulsi R, Steen VD. Increased morbidity and mortality of scleroderma in African Americans compared to non‐African Americans. Arthritis Care Res (Hoboken) 2019;71:1154–63. [DOI] [PubMed] [Google Scholar]
- 32. Al‐Sheikh H, Ahmad Z, Johnson SR. Ethnic variations in systemic sclerosis disease manifestations, internal organ involvement, and mortality. J Rheumatol 2019;46:1103–8. [DOI] [PubMed] [Google Scholar]
- 33. Johnson SR, Urowitz MB, Ibañez D, Gladman DD. Ethnic variation in disease patterns and health outcomes in systemic lupus erythematosus. J Rheumatol 2006;33:1990–1995. [PubMed] [Google Scholar]
- 34. Mayberry RM, Nicewander DA, Qin H, Ballard DJ. Improving quality and reducing inequities: a challenge in achieving best care [article retracted in World Hosp Health Serv 2009;45:1]. World Hosp Health Serv 2008;44:16–31. [PubMed] [Google Scholar]
- 35. Singh GK, Rodriguez‐Lainz A, Kogan MD. Immigrant health inequalities in the United States: use of eight major national data systems. ScientificWorldJournal 2013;2013:512313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Uribe AG, Romero‐Díaz J, Apte M, Fernández M, Burgos PI, Reveille J, et al. Impact of immigration on the clinical expression of systemic lupus erythematosus: a comparative study of Hispanic patients residing in the USA and Mexico. Rheumatology (Oxford) 2009;48:1392–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials