Abstract
Objective. Although multiple diseases associated with primary Sjögren’s syndrome (pSS) have been reported, reliable data regarding the prevalence of specific medical comorbidities among patients with pSS remain sparse. We investigated the prevalence and risk for a broad spectrum of medical conditions among patients with pSS in Taiwan.
Methods. A total of 1974 patients with pSS were eligible for inclusion in the study group. We randomly selected 9870 enrollees matched with the study subjects, using the Taiwan National Health Insurance Research Dataset for 2006 and 2007, inclusive. Conditional logistic regression analyses conditioned on sex, age, monthly income, and level of urbanization of the patient’s community were used to calculate the odds ratios (OR) of various comorbid conditions.
Results. Pearson chi-square tests revealed that patients with pSS had significantly higher prevalence of hyperlipidemia, cardiac arrhythmias, headaches, migraines, fibromyalgia (FM), asthma, pulmonary circulation disorders, hypothyroidism, liver disease, peptic ulcers, hepatitis B, deficiency anemias, depression, and psychoses. Conditional regression analyses showed that, compared to patients without the condition, patients with pSS were more likely to have hyperlipidemia (OR 1.42), cardiac arrhythmias (OR 1.32), headaches (OR 1.47), migraines (OR 1.86), FM (OR 1.71), asthma (OR 1.54), pulmonary circulation disorders (OR 1.42), hypothyroidism (OR 2.37), liver disease (OR 1.89), peptic ulcers (OR 1.88), hepatitis B (OR 2.34), deficiency anemias (OR 1.33), depression (OR 2.57), and psychoses (OR 2.15).
Conclusion. The prevalence of several comorbidities was increased among the patients with pSS. Our study provides epidemiological data for comorbidities among pSS patients in an ethnic Chinese population.
Sjögren’s syndrome (SS) is a relatively common autoimmune disease that mainly affects the exocrine glands, such as the lacrimal and salivary glands1. Patients with primary SS (pSS) usually have characteristic chronic symptoms of ocular and oral dryness. Although the pathomechanisms of SS are still unclear, environmental factors are considered to trigger abnormal immunological reactions in people with genetic susceptibility1. Its clinical features, such as the appearance of pSS autoantibody profiles, share some similarity with systemic lupus erythematosus (SLE)1,2. Secondary SS can also be accompanied by other autoimmune diseases, such as rheumatoid arthritis, SLE, and primary systemic sclerosis1.
As in other autoimmune diseases, the immunological derangement caused by pSS can be systemic. Although the data are scanty, one recent study shows about 40%–50% of patients with pSS develop extraglandular disease3. Although multiple extraglandular conditions have been recognized in previous studies, many of these diseases have been reported only sporadically1. In addition, the prevalence and risk of these medical comorbidities among patients with pSS cannot be accurately ascertained, due to limitations such as potential selection bias and relatively small sample sizes for such studies. To our knowledge, the data regarding the prevalence of medical comorbidities among Asian or Chinese patients with pSS are still lacking.
The adverse effects on general health, functional status, and quality of life of comorbidity in chronic illness were demonstrated in previous studies4. The data regarding the comorbidity associated with pSS are vital to improve the care of patients with pSS. Our goal was to explore the prevalence and risk of a broad spectrum of comorbid medical conditions among patients with pSS, using a nationwide case-controlled approach for a Han Chinese population. We further compared the comorbidity profiles of pSS in Han Chinese and Western populations.
MATERIALS AND METHODS
Database
This study used the Taiwan National Health Insurance Research Dataset (NHIRD). Taiwan began the National Health Insurance (NHI) program in 1995, and the program currently includes over 21 million enrollees, representing over 98% of the island’s population. The NHIRD contains all ambulatory care and inpatient claims, a registry of contracted medical facilities, and a registry of all beneficiaries of the program. The NHIRD therefore provides a unique opportunity to explore the risk of various comorbidities among patients with pSS.
Since the NHIRD consists of de-identified secondary data released to the public for research purposes, this study was exempt from full review by the Institutional Review Board.
Study sample
The study design consisted of a study group and a comparison group. The study group comprised all patients who sought ambulatory care during a 2-year period from 2006 to 2007, receiving a principal diagnosis of pSS (ICD-9-CM code 710.2; n = 4187). The diagnoses were considered reliable because the features of dryness of the eyes and mouth and blood antibody testing (such as antinuclear antibodies) are required for initial diagnosis of pSS in Taiwan. To further assure validity of diagnosis, we selected only patients who had at least 2 consensus pSS diagnoses for the study group (n = 2238). In addition, we excluded patients who had comorbid SLE, rheumatoid arthritis, scleroderma, sarcoidosis, antiphospholipid syndrome, hepatitis C, lymphoma, radiation therapy mucositis, or ankylosing spondylitis, in order to limit our study sample to primary SS. In total, 1974 patients with SS were eligible for inclusion in the study group.
The patients in the comparison group were drawn from a subset database released by the Taiwan National Health Research Institute (TNHRI) in 2007. This database includes 1,000,000 subjects selected randomly and systematically from the NHIRD, representing about 5% of all enrollees in the NHI program. The TNHRI identified no significant differences in the distribution of age and sex between this representative dataset and the entire NHIRD. We randomly selected our comparison group from this subset database, after excluding subjects who had previously been diagnosed with pSS, identifying 9870 enrollees (5 for every patient with pSS) matched with the study subjects for sex, age (≤ 44, 45–64, and ≥ 65 yrs), monthly income (< NT$15,000, $15,000–30,000, $30,001–50,000, and ≥ $50,001) and level of urbanization of the community in which the patient resided (urbanization levels in Taiwan are divided into 5 strata, level 1 referring to the “most urbanized” and level 5 the “least urbanized” communities).
Comorbidity
In this study, the medical comorbidities for analysis were selected based on the Elixhauser Comorbidity Index, which was created in 1997 and has been widely used for risk adjustment in administrative datasets. The Elixhauser method of comorbidity measurement includes 30 comorbidity measures, but we selected only 19 comorbidities from the index, since some conditions such as obesity, weight loss, coagulopathy, and alcohol abuse have very low prevalence in Taiwan. These 19 comorbidities included hypertension, congestive heart failure, cardiac arrhythmias, peripheral vascular disorder, paralysis, other neurological disorders, pulmonary circulation disorders (ICD-9-CM codes 415, 416.0~416.9, 417.9), hypothyroidism, liver disease (ICD-9-CM codes 070.32, 070.33, 070.54, 456.0, 456.1, 456.20, 456.21, 571.0, 571.2, 571.3, 571.40~571.49, 571.5, 571.6, 571.8, 571.9, 572.3, 572.8), diabetes without complications, peptic ulcer, diabetes with complications, renal failure, fluid and electrolyte disorders, deficiency anemias, depression, psychoses, metastatic cancer excluding lymphoma, and solid tumors. We also included 19 additional comorbid conditions (stroke, ischemic heart disease, hyperlipidemia, migraines, multiple sclerosis, polyneuropathy, fibromyalgia (FM), myasthenia gravis, hepatitis B, biliary cirrhosis, chronic interstitial cystitis/chronic cystitis, tuberculosis (TB), headaches, dementia, epilepsy, urinary tract infection, lymphoma, asthma, and chronic obstructive pulmonary disease (COPD), conditions that are prevalent in patients with pSS and in the general adult population in Taiwan. In our study, comorbid conditions were counted only if the condition occurred either in the inpatient setting or in 2 or more ambulatory care claims coded in the years 2006 and 2007.
Statistical analysis
All statistical analyses were performed with the SAS statistical package (SAS for Windows, Version 8.2). Pearson chi-square tests were performed to explore the prevalence of comorbidities among patients with and without pSS. Conditional logistic regression analyses conditioned on sex, age, monthly income, and level of urbanization of the community in which the patient resided were used to calculate odds ratios (OR) and 95% confidence intervals (CI) for each of the 31 comorbid conditions among patients with and without SS. A 2-sided significance test requiring p ≤ 0.05 was used.
RESULTS
Overall, the mean age of the sampled patients was 49.7 years (standard deviation 21.0). Table 1 gives demographic characteristics of the sampled patients. The overwhelming majority of patients were female (76.2%) and about two-thirds (66.8%) were over 44 years of age.
Sociodemographic characteristics of patients with Sjögren’s syndrome (SS) compared with controls (n = 11,844). Data are number (%).
Table 2 presents the prevalence of comorbidities by patient group. After matching for sex, age, monthly income, and urbanization level of the patient’s community, Pearson chi-square tests showed that patients with pSS had significantly higher rates of hyperlipidemia (p < 0.001), cardiac arrhythmias (p = 0.031), other neurological disorders (p = 0.014), headaches (p < 0.001), migraines (p < 0.001), FM (p < 0.001), asthma (p < 0.001), pulmonary circulation disorders, (p < 0.001), hypothyroidism (p < 0.001), liver disease (p < 0.001), peptic ulcers (p < 0.001), hepatitis B (p < 0.001), deficiency anemias (p = 0.043), depression (p < 0.001), and psychoses (p < 0.001). No significant difference was observed between patients with and those without pSS in the prevalence of hypertension, ischemic heart disease, congestive heart failure, peripheral vascular disorder, paralysis, epilepsy, dementia, COPD, diabetes without complications, diabetes with complications, renal failure, fluid and electrolyte disorders, TB, urinary tract infections, metastatic cancer, and solid tumors.
Prevalence of medical comorbidities in patients with Sjögren’s syndrome (SS) compared with controls (n = 11,844). Data are number (%).
Table 3 gives OR for each of 31 comorbidities, comparing patients with and without pSS. Conditional regression analyses conditioned on sex, age, monthly income, and level of urbanization showed that, compared to patients without pSS, patients with pSS were more likely to have hyperlipidemia (OR 1.42, 95% CI 1.23–1.64), cardiac arrhythmias (OR 1.32, 95% CI 1.03–1.71), headaches (OR 1.47, 95% CI 1.30–1.67), migraines (OR 1.86, 95% CI 1.30–1.67), fibromyalgia (OR 1.71, 95% CI 1.48–1.97), asthma (OR 1.54, 95% CI 1.22–1.93), pulmonary circulation disorders (OR 1.42, 95% CI 1.21–1.68), hypothyroidism (OR 2.37, 95% CI 1.92–2.93), liver disease (OR 1.89, 95% CI 1.60–2.22), peptic ulcers (OR 1.88, 95% CI 1.62–2.19), hepatitis B (OR 2.34, 95% CI 1.82–3.02), deficiency anemias (OR 1.33, 95% CI 1.01–1.77), depression (OR 2.57, 95% CI 2.03–3.25), and psychoses (OR 2.15, 95% CI 1.65–2.80).
Prevalence of medical comorbidities in patients with Sjögren’s syndrome (SS) compared with controls (n = 11,844). Data are odds ratios (95% CI). OR were calculated by regression analyses conditioned on patient’s age, sex, urbanization level, and monthly income.
DISCUSSION
The influence and consequences of comorbid medical conditions among patients with pSS remain unclear. Although studies show that overall mortality is not increased among patients with pSS, compared with the general population3,5,6, significantly decreased quality of life has been noted among the patients with pSS7,8. In addition, patients with pSS experience a high prevalence of fatigue, pain, and depression, which could be associated with coexisting comorbidities7,9. We found higher prevalence of several medical conditions, including cardiovascular, neurological, pulmonary, endocrine, gastrointestinal, infections, hematological, and psychological diseases among patients with pSS. As about 98% of inhabitants of Taiwan are Han Chinese, our data should be highly representative of the Han Chinese population. In general, we found the profiles of comorbidity among pSS patients of Han Chinese ethnicity were similar to previous reports on Western populations. However, some of our results were inconsistent with previous reports.
Cardiovascular involvement, such as heart conduction problems and autonomic dysfunctions, has been identified as associated with pSS10–13. Our data showed higher prevalence of cardiac arrhythmias, which is compatible with previous observations. Derangement of lipid metabolism has been reported in association with many autoimmune diseases14. Limited, but notable data regarding dyslipidemia in patients with pSS have been published15–17. Lodde, et al showed a significant difference in lipid profiles associated with serological measurement of inflammation in a small case-control study15. Our data support that patients with pSS have a higher prevalence of hyperlipidemia. Chronic systemic inflammation is also considered a risk factor for developing atherosclerosis18. However, contrary to what we expected, the prevalence of atherosclerotic consequences such as ischemic heart disease, peripheral vascular disorders, and strokes was not significantly increased among patients with pSS. We proposed that the complex relationships among inflammation, vascularity, coagulation, and the autonomic system could be associated with atherosclerosis in patients with pSS. Further studies are needed to confirm our data.
Although hepatitis C virus (HCV) infection is recognized as being associated with pSS19,20, the role of HBV infection in pSS is still controversial. One case report described pSS occurring after HBV vaccination21. In contrast, Ram, et al proposed that HBV infection may play a protective role against some autoimmune disorders22. Interestingly, we found a higher prevalence of hepatitis B among patients with pSS. HBV infection is highly endemic in Taiwan23. Further study is needed to determine the role of HBV infection in pSS, particularly in Chinese populations.
Our data are compatible with studies that demonstrated higher prevalence of headaches and migraines among patients with pSS24–26. Vascular involvement is hypothesized to contribute to headaches and migraine symptoms in patients with pSS25. Varying degrees of neurological involvement have been demonstrated in pSS. One survey reported that up to 71% patients with pSS have central or peripheral neurological system involvement27. Further, neuropsychiatric surveys have demonstrated a significant prevalence of cognitive impairment among patients with pSS24. However, we found a trend that did not reach statistical significance suggesting increased risk of peripheral neuropathy among patients with pSS. We did not find increased prevalence of dementia among the patients with pSS. Clinical diagnosis of neurological involvement in patients with pSS could be difficult, since the clinical course is highly variable, and subclinical presentations are not unusual. Therefore we propose that some patients with pSS have mild neuropathy or cognitive impairment that fails to raise clinical attention.
Our data confirm several comorbidity patterns observed by studies with small sample sizes or in Western populations. For example, Amin, et al demonstrated increased hypersensitivity of bronchial airways in pSS similar to atopic asthma, despite some differences in inflammation patterns28. Our study supports this association in a Chinese population. Although there is still some controversy about it, many articles have suggested diffuse myalgia and widespread pain are associated with pSS1,29,30. Our data support a higher prevalence of FM among patients with pSS in a Chinese population. Higher prevalence of hypothyroidism and hepatic dysfunction among patients with pSS compared to the general population noted in our study is also consistent with previous findings, despite the lower prevalence in our data31. It is worth noting that we also found higher prevalence of peptic ulcers among the patients with pSS. Heliobacter pylori are proposed to be associated with rheumatic diseases, including pSS with extragastric presentation. Our findings may provide epidemiological evidence for this association.
Depression and psychoses are recognized to be associated with many kinds of autoimmune diseases. Recent imaging and neuropsychiatric evaluations have indicated such psychiatric dysfunction in patients with pSS may be associated with psychological stress from a disabling chronic disorder, but may also have an organic neuronal basis32. Our findings showed that of all comorbidities we surveyed, depression posed the highest risk to patients with pSS. Nevertheless, the prevalence of depression and psychoses among patients with pSS in our data is still much lower than in previous studies in Western populations. Underdiagnosis and undertreatment of psychiatric conditions is quite common among patients with other autoimmune diseases. Our data may reflect a similar phenomenon also exists in Taiwan. Recognizing the need for psychological or psychiatric care should be emphasized for these patients.
Our results are in some ways incompatible with results from studies in Western populations. For example, previous reports have suggested high risk for developing lymphoma among patients with pSS3,33. A nationwide population-based study in Sweden noted the risk of non-Hodgkin’s lymphoma is about 6.4 times higher in patients with pSS33. Nevertheless, we found only a positive trend toward developing lymphoma among patients with pSS in Taiwan that did not reach statistical significance. Such contradictions could come from different selection criteria, the duration of followup, and ethnic differences. Further study is needed to confirm our findings. Previous studies found biliary cirrhosis to be associated with pSS, but there were only rare cases in our survey. Some authors have suggested interstitial cystitis and diabetes mellitus are associated with pSS17,34; however, we did not find an association with either condition in our data.
Our study has several limitations. First, the validity of pSS diagnoses is difficult to determine using a nationwide health database. Miscoding of patients with pSS with mild symptoms could occur under different diagnostic criteria. In accord with diagnostic criteria set up by international consensus, our analysis revealed that over 95% of our sampled patients had previously received at least 2 of the following objective tests: autoantibodies (Ro/La antibodies), histopathology (biopsy of oral mucosa), or objective salivary gland investigation (sialoscintigraphy) during the period from 2004 to 2007. Although detailed results cannot be determined from the dataset used in our study, the diagnoses of pSS were made by physicians after performing these tests on the sampled patients, which strongly suggests these patients had positive results on these tests. Therefore, we believe the validity of diagnoses of our sampled patents was acceptable for analysis of epidemiological profiles. Second, it is sometimes difficult to differentiate secondary SS from pSS, because it is not unusual for manifestations accompanying autoimmune diseases to be delayed for a long period. Third, cause-effect relationships could not be determined in a cross-sectional study. A longitudinal study would be necessary to clarify the interaction between pSS and specific comorbid conditions. Fourth, variables including physical activity, smoking and alcohol consumption, dietary habits, and family history have a confounding effect on development of certain comorbidities, but could not be determined in our database. Finally, selection bias could have occurred in our study, since patients with mild disease may not seek medical attention, with the result that not all medical conditions come to doctors’ attention and are recorded in the database. In such circumstances, the prevalence of medical conditions is likely to be underestimated.
The challenge remains to determine the underlying pathomechanisms of medical conditions comorbid with pSS. Shared genetic, environmental, and secondary behavior contributing to disease, as well as medication, could all account for the development of specific comorbid medical conditions.
To our knowledge, this is the first study to investigate prevalence data for a broad spectrum of medical conditions comorbidly associated with pSS using a large, population-based dataset of Han Chinese. Our data indicate increased prevalence of multiple comorbidities among patients with pSS. In general, the pattern of comorbidities appears similar in Chinese and Western populations with pSS. Clinicians should be aware of specific comorbid medical conditions common among patients with pSS.
Footnotes
-
This study is based in part on data from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health, Taiwan, and managed by the National Health Research Institutes. The interpretations and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health, or the National Health Research Institutes.
- Accepted for publication January 7, 2010.
