AMERICAN INDIANS HAVE A HIGHER RISK OF SJÖGREN’S SYNDROME AND MORE DISEASE ACTIVITY THAN CAUCASIANS AND AFRICAN-AMERICANS.

Abstract

Objective:

To describe the clinical and serological manifestations of Sjögren’s syndrome (SS) in ethnic groups of the United States.

Methods:

Cross-sectional study of 648 patients with primary SS: 20 African-American (AA), 164 American Indian (AI), 426 European-American (EA) and 38 of other races evaluated in a multi-disciplinary sicca research clinic.

Results:

AA subjects comprised 3.1% of the SS cohort, much lower than the percentage of AA in the State of Oklahoma (p=3.01xE-05), the United States (p=2.24E-13) or a lupus cohort at the same institution (p=4.26×10E-27). In contrast, the percentage of AI in the SS cohort (25.3%) was much higher than expected (p=3.17E-09 vs. SLE; p=6.36-26 vs. Oklahoma and p=8.14E-96 vs. USA population). The SS classification criteria were similar between AA and EA, but subjects of AI ancestry had lower rates of abnormal tear and salivary flow as well as anti-Ro/SSA and anti-La/SSB antibodies. Paradoxically, AI had higher levels of disease activity (mean±SD ESSDAI 3.77±4.78) in comparison to whites (2.90±4.12; p=0.011) and more extraglandular manifestations affecting mainly the articular and glandular domains. Meanwhile, AA patients were characterized by higher rates of hypergammaglobulinemia (OR 1.39, 95%CI 1.39-8.65, p=0.01), elevated ESR (OR 3.95, 95%CI 1.46-9.95, p=0.009), and parotid enlargement (OR 4.40, 95%CI 1.49-13.07, p=0.02).

Conclusions:

American Indians are affected at high rates with SS but present with few classical features, potentially preventing timely diagnosis. In contrast to SLE, SS is infrequent and not more severe amongst AA, but the triad of hypergammaglobulinemia, increased ESR and parotid enlargement warrants extra vigilance for lymphomagenesis.

Sjögren’s syndrome (SS) is a chronic autoimmune disease characterized by salivary and lacrimal gland dysfunction mediated by lymphocytic infiltration and/or anti-Ro/SSA and anti-La/SSB antibodies (1). The glandular involvement may extend to other exocrine glands like those of nasal and tracheobronchial mucosa, pancreas, gastrointestinal and genitourinary tracts (2). Extraglandular manifestations are present in about a third of SS patients. These include arthralgia, arthritis, Raynaud’s phenomenon, lymphadenopathy, pulmonary small airway disease, systemic vasculitis, interstitial nephritis, myositis, peripheral neuropathy, and increased risk of lymphomas (2-4).

Confirmation of SS diagnosis or research classification in a patient with sicca symptoms is complicated. The most widely accepted criteria for research classification are the highly correlated 2002 American-European Consensus Group (AECG) Classification Criteria (5) and the more recent 2016 American College of Rheumatology/European League Against Rheumatism Classification Criteria (ACR/EULAR) for Primary Sjögren’s Syndrome (6, 7). In the clinical setting, the diagnosis of SS is difficult to establish because it is based on clinical judgement, there is no single diagnostic gold standard test, and often a multi-disciplinary team and invasive procedures are required. Consequently, SS is frequently misdiagnosed, underdiagnosed or diagnosed at late stages of the disease (8, 9).

In spite of SS being second only to rheumatoid arthritis in frequency among inflammatory rheumatic diseases (10, 11), the epidemiology, racial distribution and ethnicity-specific phenotype has only recently been addressed. Furthermore, the vast majority of the descriptive studies, clinical trials and genetic studies have centered around populations of white, European ancestry. Given that racial minorities are disproportionately affected and have worse outcomes in other autoimmune diseases related to SS, in particular systemic lupus erythematosus (SLE), it is especially relevant to fill this knowledge gap (12). Thus, we undertook this study to describe the phenotypic manifestations of SS in Black/African-Americans (AA) and American Indians (AI) in comparison to White/European-Americans (EA) and contrasted the findings with SLE.

PATIENTS AND METHODS

Study Participants and Ethical Considerations

The Oklahoma Sjögren’s Syndrome Center of Research Translation cohort includes 1682 voluntary participants evaluated for sicca symptoms between 2005 and 2017 at the Oklahoma Medical Research Foundation (n=1230), the University of Minnesota (n=372), Cedars-Sinai Medical Center (n=61), and the Carolinas Medical Center (n=19). A total of 648 participants meet AECG classification criteria for primary SS; for the analyses of the clinical and serological features, we included a subset of 610 subjects self-identified with the three major races in our cohort: 426 EA, 164 AI and 20 AA. Other racial and ethnical categories, namely Asian, Native Hawaiian or Other Pacific Islander and Hispanic, were not included in the clinical and serological analysis because they included too few subjects (n=38 combined).

Each institution’s Institutional Review Board approved all procedures and the participants provided two-tiered informed consent prior to entering the study. Initial verbal consent was obtained before the screening phone interview and the mailing of study paperwork and questionnaires; on the day of the clinic visit and before any procedure took place, additional written informed consent was obtained.

Comparison Groups.

We compared the SS cohort to a SLE cohort with a similar geographic distribution and housed in the same institution, the Oklahoma Lupus Cohort. Each subject satisfied the 1997 revised ACR classification criteria for Systemic Lupus Erythematosus (13, 14).

Classification of Race and Ethnicity.

We followed the NIH guidelines in determination of race and ethnicity of participants in both the SS and SLE cohorts using self-reporting or self-identification. The 1997 Office of Management and Budget (OMB, Executive Office of the President) revised minimum standards include two ethnic categories (Hispanic or Latino, and Not Hispanic or Latino) and five racial categories (American Indian or Alaska Native, Asian, Black or African-American, Native Hawaiian or Other Pacific Islander, and White or European-American) (15). Data for racial distribution in the United States and the state of Oklahoma were obtained from the 2010 US Census (16). The categories in this classification are social-political constructs and should not be interpreted as being anthropological in nature. In the AI racial category, we included individuals self-reporting as two races when those were American Indian/White because more than half of all AI self-identify in such way in the US Census data (16).

Clinical Procedures

The protocols for patient recruitment, assessment, and data collection have been described previously (17) and include all the tests necessary for SS classification based on the AECG 2002 revised criteria (5) and the 2016 ACR/EULAR criteria (6, 7). The cohort and this study are based on a cross-sectional, single visit assessment of each participant.

Briefly, candidates for evaluation at the four Sjögren’s research clinics were referred by health care providers or responded to public advertisement and included patients already clinically diagnosed with SS as well as subjects with subjective dry eyes and dry mouth without a prior diagnosis of the disease. Exclusion criteria followed the AECG classification (5) and included past head and neck radiation therapy, hepatitis C infection, acquired immunodeficiency syndrome, pre-existing lymphoma, sarcoidosis, and graft-versus-host disease as well as pregnancy and inability to provide informed consent. At the time of the design and inception of this sub-study, additional inclusion criteria were absence of concurrent autoimmune conditions and confirmed AECG classification as SS (5); the vast majority of these subjects (EA 95.3%; AI 95.7%; and AA 95.0%) also meet the ACR/EULAR criteria for SS classification (6, 7).

The study subjects participated in multi-specialty clinical evaluation that included measurement of whole unstimulated salivary flow (WUSF), minor labial salivary gland biopsy, Schirmer’s I testing, calculation of the van Bijsterveld (vBS) (18) and ocular staining scores (OSS) (19), determination of relevant antibodies (anti-Ro/SSA, anti-La/SSB, antinuclear antibodies (ANA), rheumatoid factor, serum immunoglobulins, anti-cardiolipins and anti-ENA), and full physical examination.

Disease Activity and Extraglandular Manifestations.

Disease activity was assessed using the EULAR Sjögren’s syndrome disease activity index (ESSDAI) and EULAR Sjögren’s Syndrome Patient Reported Index (ESSPRI) on the day of clinical evaluation (20, 21) Historical extraglandular manifestations, based on the ESSDAI definitions (22), were documented from medical records, which were available for 13 AA, 112 AI, and 209 EA participants.

Statistical analysis

Count data were analyzed using Fisher’s exact, Pearson’s χ², or Mann Whitney tests, as appropriate. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. The comparison of SS or SLE prevalence based on race to the population data was done using a proportions (z) test. Where no specific a priori hypothesis was being tested, correction for multiple testing was done using the false discovery rate (FDR) two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli (23). Hypothesis tests were performed in GraphPad Prism version 7.0e for Mac OS X (GraphPad Software, La Jolla, California, USA) and VassarStats (vassarstats.net). A p value <0.05 was considered statistically significant.

RESULTS

The evaluation for classification as primary SS of 1682 patients with sicca in our cohort resulted in 648 subjects that met the classification criteria; we included in this study 610 belonging to the three main racial subgroups previously described (Methods). Non-Hispanic EAs represented 65.7% (n=426) of all SS, 25.3% (n=164) were AI, and AA were 3.1% (n=20). The proportion of SS vs. non-Sjögren’s sicca was similar across the three groups (46.4% EA, 41.7% AA, 40.2% AI, p=0.11) (Supplemental Table 1). The rates of AA and AI in the cohort raised the question of the racial distribution, predisposition, and predilection of SS in comparison to the general population (Table 1). Based on 2010 US Census data (16), AA constitute 7.3% of the population of the state of Oklahoma and 12.6% of the population of the United States of America. Among SS patients, AAs are represented at less than half the population rate (OR 0.41, 95% CI 0.26–0.63, p=3.01E-05 vs. Oklahoma population and OR 0.22, 95% CI 0.14–0.34, p=2.24E-13 vs. the USA population). The contrary is the case for AIs, who are over-represented amongst the SS patients (25.3%) in comparison to the State of Oklahoma (11.9%; OR 2.51, 95%CI 2.10–2.99, p=6.36E-26) and the AI population of the United States (1.41%; OR 23.63, 95% CI 19.8–28.21, p=8.14E-96). Of interest, when considering the subjects in the cohort who did not meet research classification criteria for SS (non-Sjögren’s sicca), only 28 of 715 (3.9%) were AA (p=0.0007 vs. Oklahoma population and p=1.91E-12 vs. USA population), while 34.1% were American Indian (p=6.48E-75 vs. Oklahoma population and p=8.89E-100). Thus, amongst subjects with sicca, both those meeting and not meeting SS criteria are much more likely to be AI and much less likely to be AA than the general population.

Table 1.

Racial distribution of subjects classified as Sjögren’s syndrome in comparison with the racial distribution of the general population in the State of Oklahoma and the United States of America

	OMRF cohort SS	Oklahoma population	OR (95%CI)	P value	United States population	OR (95%CI)	P value
European-American	65.7%	72.2%	0.74 (0.63-0.87)	0.0003	72.4%	0.73 (0.62-0.86)	0.0002
African-American	3.1%	7.27%	0.41 (0.26-0.63)	3.01E-05	12.61%	0.22 (0.14-0.34)	2.24E-13
American Indian	25.3%	11.9%	2.51 (2.10-2.99)	6.36E-26	1.41%	23.6 (19.8-28.2)	8.14E-96

^aSjögren’s Syndrome: Subjects are classified as primary Sjögren’s Syndrome based on the AECG criteria.(5). Racial distribution based on data from the 2010 US Census (16).

This difference in the ethnic/racial make-up of SS in comparison to the population racial distribution, prompted us to look at an SLE cohort recruited under similar conditions. The AA representation in the SS cohort (3.1%) was significantly lower than the AA representation of the SLE comparison cohort (23%) (OR 0.11, 95% CI 0.07–0.17, p=4.62E-27) (Table 2). Thus, AA were 9 times more likely to have SLE than SS. The opposite was true for AI, who represented 13% of the SLE cohort but ~25% of the SS subjects (OR 2.28, 95% CI 1.72–3.0, p=3.17E-09) (Table 2). Given that the SLE cohort was recruited exclusively in Oklahoma, we then constrained our analysis to include only those SS subjects recruited from Oklahoma (n=414) and the difference was even more striking: AI represented 36.7% and AA only 4.1% of SS (AI vs. SLE OR 3.90, 95%CI 2.91–5.22, p=5.95E-21; AA vs. SLE OR 0.14, 95%CI 0.09–0.24, p = 4.75E-17) (Supplemental Table 2).

Table 2.

Impact of race on risk of SS in comparison to the risk of SLE

	European- American (EA)	African- American (AA)	American Indian (AI)	Other*	EA vs. non-EA		AA vs. non-AA		AI vs. non-AI
					OR (95%CI)	p value	OR (95%CI)	p value	OR (95%CI)	p value
SS	426	20	164	38
N=648	(65.7%)	(3.1%)	(25.3%)	(5.9%)	1.31 (1.05 -1.64)	0.02	0.11 (0.07- 0.17)	4.62E- 27	2.28 (1.72- 3.0)	3.17E-09
SLE	422	163	92	33
N=710	(59.4%)	(23.0%)	(13.0%)	(4.6%)

SS: primary Sjögrens’ syndrome based on AECG classification criteria. SLE: Systemic Lupus Erythematosus based on the 2002 ACR classification criteria.

^*Other races include Asian, Hispanic, Pacific Islander/Hawaiian Native and unknown.

Sociodemographic features of the three racial groups of SS participants are shown in Table 3 (Supplemental Table 3 shows the sociodemographic features of non-SS sicca participants by race). For those classified as SS, the gender distributions were similar in all races but AI were younger than EA. AA and AI lived in areas with lower median income and had significantly higher Body Mass Indices than EA but did not differ between each other for those measures. AI were the group with the lowest education level while the highest proportion of college-educated patients were AA.

Table 3.

Sociodemographic characteristics of participants with Sjögren’s syndrome^a in the Oklahoma Sjögren’s Syndrome Center of Research Translation Cohort. The three major races in the cohort are included: European-American, African-American and American Indian.

	European-American n = 426b	African-American n = 20b	Native- American n=164b	P value EA vs AA	P value EA vs AI	P value AA vs AI
Age (median years [IQR])c	58 [49-67]	55 [48.8-60.3]	55.5 [45.3-62]	0.15	0.006	0.838
Gender (female)d	391 (91.8%)	19 (95%)	156 (95.1%)	1.0	0.215	1.0
Education
12 years or lessd	92 (24.0%)	1 (6.3%)	53 (38.4%)	0.381	0.002	0.033
College or equivalentd	190 (49.6%)	13 (81.3%)	71 (51.5%)	0.019	0.766	0.032
Graduate/Professional or moree	101 (26.4%)	2 (12.5%)	13 (9.4%)	0.379	1.92E-05	0.694
Income (median USD [IQR])d	52,074 [42,759-70,901]	45,016 [33,953-53,982]	49,560 [39,306-57,241]	0.029	0.0007	0.277
BMI (median [IQR])d	26.5 [23.5-32.0]	34.8 [28.1-40.3]	29.4 [23.4-36.4]	0.0018	0.007	0.058

^aSjögren’s Syndrome: Subjects are classified as primary Sjögren’s Syndrome based on the AECG criteria.(5).

^bDenominators may vary due to some missing datapoints.

^cMann-Whitney Test;

^dFisher’s exact test;

^eχ² Test.

EA: European-American; AA: African-American; AI: American Indian. BMI: Body Mass Index

The evaluation of the criteria for SS classification (Figure 1, Supplemental Tables 4 and 5) did not demonstrate any significant differences between AA and EA patients. More EA individuals had reduced whole unstimulated salivary flow (WUSF) rates in comparison to AA (70.1% vs. 45%, respectively; p=0.025) but this difference did not retain statistical significance after adjustment for multiple testing (FDR q=0.06). In contrast, fewer AI patients had decreased salivary flow (WUSF <0.1 mL/min in 70.1% of EA vs. 61.4% of AI, p=0.049, FDR q=0.10), and tear production (Schirmer’s I test <5mm in 60.3% EA vs. 44.5% in AI, p=0.0008, FDR q=0.004) in comparison to EA patients. A similar pattern was observed for autoantibodies and focal lymphocytic sialadenitis. AA and EA participants had similar rates of focus score ≥1 in their minor salivary gland lip biopsy and of anti-Ro/SSA, anti-La/SSB, ANA, and rheumatoid factor in their serum. Meanwhile, AI SS patients had lower rates than EA of anti-Ro/SSA (p=0.005, FDR q=0.02), anti-La/SSB (p=8.41E-06, FDR q=8.49E-05), and rheumatoid factor (p=0.01, FDR q=0.04). Other serologic and hematologic markers showed that AA subjects more frequently had hypergammaglobulinemia (IgG) than EA subjects (52.6% vs. 23.6%; OR 3.60, 95%CI 1.39 to 8.65; p=0.01, FDR q=0.04) and AI subjects, who only had increased serum IgG in 14.6% of cases (p=0.0004 vs. AA and p=0.02 vs. EA) (Figure 2, Supplemental Tables 6 and 7). Additionally, the erythrocyte sedimentation rate (Westergren) was elevated in almost half (47.1%) of all AA in contrast to EA (18.4%; OR 3.95, 95%CI 1.46–9.95; p=0.009; FDR q=0.04) and AI (23.2%; OR 0.34, 95% CI 0.13–0.98; p=0.043; FDR q=0.13) (Figure 2, Supplemental Tables 6 and 7). No differences were found in rates of other, non-canonical auto-antibodies such as dsDNA, Sm, or anticardiolipin (data not shown).

Figure 1. Classification criteria in European-American, American Indian and African-American patients with Sjögren’s Syndrome.

The features of European-Americans are shown in orange, of American Indians in yellow and of African-Americans in green. Percentage of patients meeting each of the AECG classification criteria for primary Sjögren’s syndrome. FS=focus score in minor salivary gland lip biopsy; vBS=van Bijsterveld Score; WUSF=whole unstimulated salivary flow; ESR=erythrocyte sedimentation rate.

Figure 2. Serological findings in European-American, American Indian and African-American patients with primary Sjögren’s Syndrome.

The features of European-Americans are shown in orange, of American Indians in yellow and of African-Americans in green. Percentage of patients with serological abnormalities.

Extraglandular manifestations and disease activity were similar between all racial groups in most domains, however EA tended to have slightly lower ESSPRI and ESSDAI scores than both AA and AI (Figure 3 and Supplemental Table 8). The most active or frequently involved domains in both AA and AI were the biological, glandular domain (activity and/or history of parotid swelling) and articular domain. However, the perceived pain, fatigue and dryness, as evaluated on the ESSPRI were not different. Finally, there were no significant differences between races in the frequency of lymphoma or concurrent diseases (data not shown).

Figure 3. Disease activity in European-American, American Indian and African-American patients with Sjögren’s Syndrome.

The features of European-Americans are shown in orange, of American Indians in yellow and of African-Americans in green. Percentage of patients showing activity in ESSDAI (EULAR Sjögren’s Syndrome Disease Activity Index) domains (20). Only domains in which >2% of subjects had activity are shown. Peripheral nervous system and respiratory domains are not shown because confirmatory tests (i.e. skin biopsy for small nerve neuropathy and/or nerve conduction velocity and chest imaging studies) were not available for the vast majority of patients.

DISCUSSION

Systemic connective tissue diseases, including those closely related to SS such as SLE, systemic sclerosis and rheumatoid arthritis, have great variations in prevalence and disease severity in different populations. Both ethno-racial and geographical components contribute to these differences and provide a glimpse into potential divergent pathophysiological mechanisms that remain to be explored. Recognizing the phenotypic differences across population subsets should lead to improved diagnostic timeliness and accuracy, and to enhanced personalized follow up and therapeutic strategies. Thus, we aimed to establish whether such differences exist in SS by analyzing the clinical and serological characteristics of a large multi-racial cohort of patients with sicca from the United States of America who were directly evaluated using a standardized protocol. This study is unique in its detailed characterization of a large group of American Indians.

We found marked differences in the susceptibility to SS in the different races; AI are affected at a higher rate than expected based on their proportion in the general population and AA are under-represented both in comparison to the general population and to SLE. A similar pattern of susceptibility has been observed in RA; while the annual global incidence of RA is between 0.5 and 1.0% in Caucasians, it can be as low as 0.1% in rural Africans and up to 5% in some AI Tribes (Pima, Blackfeet, and Chippewa Indians) (24-26). In the case of SS, most of the knowledge is derived from studies in populations of white European ancestry. Reported incidence rates of SS have ranged from 0.11 to 0.53 per 10,000 and prevalence from 1 to 330 per 10,000. In the United States, a population-based study of majority white population in Minnesota (>90% EA) estimated the prevalence of physician-diagnosed SS between 2 and 10 per 10,000 (27). A population-based study in a multi-racial suburb of Paris estimated the prevalence of primary SS at 1 to 1.5 per 10,000 subjects, with people of non-European background (North-African Arabs, sub-Saharan Africans, Afro-Caribbeans, Asians and a small proportion of Pacific-Islanders) having an increased risk and unique clinical profile (28). In our cohort of patients with sicca, the majority of participants were EA, but AI who met criteria for SS were overrepresented at twice the population rate, while AA were represented at less than half the rate in the general population of Oklahoma. These data provide further evidence of skewed representation of autoimmune diseases across racial groups and geographical locations.

To explore whether these differences are mainly due to biological rather than socioeconomic factors, we chose as a natural comparison group the Oklahoma Lupus Cohort, which is housed in the same institution and has similar catchment areas and referral sources. The racial distribution of SLE in this local cohort, is consistent with the demographics of SLE in the United States, where the prevalence is higher amongst minorities (AA, African-Caribbean, Hispanic and Asian) compared with Caucasians (12, 29, 30). The proportion of AA SLE patients is ~9 times higher than the proportion of SS in our sicca cohort, ~3 times that of AA in Oklahoma and twice the rate in the U.S.; meanwhile, AI constitute 13–16% of the SLE cases, slightly higher than the ~12% in Oklahoma, but only half the percentage of SS AI subjects (personal communication, Dr. JT Merrill). Our study is not a population-based epidemiological study, nonetheless, in view of the absence of previous surveys in the United States of multi-racial cohorts of SS and/or sicca, our study population provides a unique insight into the matter.

Differences in socioeconomic status, access to care, and attitudes towards participation in research are known covariables of race and ethnicity (31-34); these are influences that we cannot completely rule out amongst the participants in our study. Racial minorities, in particular AA, AI and Hispanics in the U.S. have well-documented disparities in physician referrals to specialized centers such as ours and individuals belonging to lower socioeconomic strata often present late or not at all in spite of significant disease burden (35, 36). In our study, AA had indicators of lower socioeconomic status as compared to EA, a possible source of their underrepresentation. The contrary is the case for AI, who in spite of also having indicators of lower socioeconomic status, in many cases have access to Indian Health Services or tribally run facilities (37) and are thereby more likely to be referred for specialized attention and to be overrepresented. However, the racial distribution of the SS cohort is different from the SLE cohort, which draws from the same population and referral sources. Thus, we hypothesize that the observed differences in disease frequency and phenotype are at least in part due to bona fide biological factors (i.e. population-specific genotypes underlying dysregulation of differing pathways), rather than purely attributable to marginalization and social determinants.

Similarly, racial disparities are reflected in clinical features and disease outcomes. In SLE, AA and Hispanics have a poorer renal prognosis than Caucasians and lower response rates to conventional treatment; AA are more likely to have anti-Sm, anti-RNP, discoid skin lesions, proteinuria, psychosis, and serositis (38, 39). In the Paris study, as a group, non-Europeans were younger and more likely to show hypergammaglobulinemia and presence of anti-Ro/SSA and anti-La/SSB antibodies; the SS prevalence among the non-Europeans was approximately twice the rate observed amongst their white counterparts (28). Likewise, a recent report of the influence of geolocation and ethnicity on the clinical presentation at diagnosis of SS in the Big Data Sjögren Project Consortium (40) [(BDSPC), an international, multicenter registry], showed that individuals of African ancestry were younger at diagnosis and had the highest proportion of affected males (28, 40). Unfortunately, neither of these studies detailed the clinical findings of SS patients of Native-American ancestry as an independent group, thus preventing us from making direct comparisons with our population.

The most actionable, race-specific findings are that AA SS patients have clinical features associated with higher risk of lymphoma (4, 41-43), including hypergammaglobulinemia, elevated sedimentation rate and parotid swelling. We do not have follow up data to assess whether these participants go on to develop more severe disease or respond poorly to conventional treatment given the cross-sectional design of our study, but these findings warrant closer surveillance of AA patients with SS.

Indigenous populations in the United States and Canada have higher prevalence of inflammatory arthritis and other rheumatic diseases as well as higher hospitalization rates but lower use of specialized services for related conditions (35). A study of Oklahoma tribal populations showed that rheumatic diseases are a significant cause of morbidity and that unusual disease manifestations result in delayed or impaired diagnoses. For example, only 69% of AI subjects with SLE had a positive ANA at the time of evaluation and 32.5% of AI patients with RA were seronegative for rheumatoid factor and anti-CCP (37). In our study, SS participants of AI ancestry were younger and had less prominent objective salivary and lacrimal dysfunction, glandular lymphocytic infiltration, or canonical autoantibodies but, paradoxically, higher levels of disease activity affecting the glandular and articular domains. The paucity of classical SS features in AI patients may complicate or postpone the diagnosis and treatment in these populations.

Weaknesses of our study are that it is not a population-based study and that it is cross-sectional in nature. Neither generalizations about prevalence of SS in minority populations of the United States nor of the prognostic implications of the specific phenotypic characteristics of each racial group can be made. However, we propose increased surveillance for malignancy in AA patients with SS and provide the first description of the high rate of SS amongst AIs, raising awareness of the relatively few clinical clues that may prevent accurate and opportune diagnoses.

SIGNIFICANCE AND INNOVATIONS.

• Race and ethnicity influence the clinical manifestations and disease severity of Inflammatory rheumatic diseases.

• American Indians are overrepresented amongst Sjögren’s syndrome patients and have increased disease activity in spite of few classical clinical manifestations.

• African-Americans are affected at the same rate and severity as Caucasians but have a higher rate of risk factors for lymphoma.

• Awareness of the race-specific sub-phenotypes of Sjögren’s syndrome should lead to more accurate and opportune diagnosis and follow-up.