521515 research-article2014
MSJ0010.1177/1352458514521515Multiple Sclerosis JournalMagyari et al.
MULTIPLE SCLEROSIS MSJ JOURNAL
Research Paper
Gender and autoimmune comorbidity in multiple sclerosis
Multiple Sclerosis Journal 2014, Vol. 20(9) 1244–1251 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1352458514521515 msj.sagepub.com
Melinda Magyari1,2,3, Nils Koch-Henriksen2,4, Claudia C Pfleger5 and Per Soelberg Sørensen1,3
Abstract Background: The female preponderance in incidence of multiple sclerosis (MS) calls for investigations into sex differences in comorbidity with other autoimmune diseases (ADs). Objectives: To determine whether male and female patients with MS have a higher frequency of autoimmune comorbidity than controls, and to describe the type and frequency of ADs that are associated with MS. Methods: Our database was established by linkage of the Danish MS Registry to The Danish National Patient Register and consisted of 1403 patients of both sexes with clinical onset of MS between 2000 and 2004, and 25 matched controls for every case. Results: None of the ADs occurred more frequently in female cases than in controls. Male cases were more likely to have Type I diabetes mellitus (odds ratio (OR) = 3.34; 95% CI 1.40 – 7.02; p < 0.008), Crohn’s disease (OR = 5.03; 95% CI 1.18 – 16.10; p = 0.03) and systemic lupus erythematosus (OR = 12.55; 95% CI 1.62 – 69.95; p = 0.02) than male controls. Conclusions: Autoimmune disorders are rare, but some of them tend to occur together with MS at a higher rate than in controls. Although women are generally more prone to ADs than men, significantly increased occurrence of other ADs were only found in male MS patients. Keywords Autoimmune disease, comorbidity, Crohn’s disease, Denmark, diabetes, epidemiology, gender, lupus, multiple sclerosis, registry Date received: 30 July 2013; accepted: 15 December 2013
Introduction Multiple sclerosis (MS) is considered to be an autoimmune disease1 (AD) that, in common with the majority of ADs, occurs more frequently in women.2 The strong female preponderance in incidence of MS calls for investigations into sex differences in comorbidity with other ADs in MS patients, to see whether and to what degree this gender difference might deviate from the population’s gender difference for ADs. ADs are treated by different medical specialties and have conventionally been studied as separate entities,3 but the amount of literature suggesting the co-existence of MS with other autoimmune disorders is growing, although some inconsistent findings in various studies have been pointed out.4,5 The aetiology of ADs is not fully understood, but cooccurrence of various ADs suggests common features in their aetiology and pathogenesis.
Several studies using different methods and designs investigated the co-existence of ADs in MS patients6,7 and the co-occurrence of autoimmune disorders within families,8 but did not establish any common predisposing 1Danish
Multiple Sclerosis Center, Department of Neurology, Neuroscience Centre, Rigshospitalet, Copenhagen, Denmark. 2Danish Multiple Sclerosis Registry, Neuroscience Centre, Rigshospitalet, Copenhagen, Denmark. 3University of Copenhagen, Denmark. 4Clinical Institute, Department of Clinical Epidemiology, University of Aarhus, Denmark. 5Department of Neurology, Aalborg University Hospital, Denmark. Corresponding author: Melinda Magyari, Danish Multiple Sclerosis Research Center, Department of Neurology, Neuroscience Centre, Rigshospitalet, section 2082, Blegdamsvej 9, Copenhagen, 2100, Denmark. Email: [email protected]
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1245
Magyari et al. factors to autoimmunity.9 As ADs are shown to be more common in family members of patients with MS than in the general population,10,11 a common genetic and environmental background for susceptibility to autoimmunity is presumed. Genome-wide association studies (GWAS) show an overlap in the genetic risk foci for MS and some other core ADs.12 Recently, ImmunoChip analyses disclosed overlapping of approximately 22% of MS signals with at least one other AD signal, most frequently with inflammatory bowel disease and less frequently with rheumatoid arthritis (RA) and autoimmune thyroid disease.13 However, in the Multiple Autoimmune Disease Genetics Consortium (MADGC) family study, some genetic loci for susceptibility to autoimmunity were not common to MS and other ADs.14 Co-morbidity has different aspects, and the occurrence of some ADs can be either more or less frequent in patients with MS, compared to the general population. MS may either predispose or protect against later onset of some other diseases or, conversely, some diseases may predispose or protect against later onset of MS. Common traits may affect the risk of MS and other ADs independently. If so, the order of onset of the diseases is not important. The high and increasing female:male sex ratio in MS15 underlines the need to search for gender differences in aetiology or pathogenesis. Autoimmune comorbidity in MS could be one of the factors that differentiates male from female MS patients, taking into consideration that most ADs are more common in women than in men.2 The main objective of this study is to determine whether male and female patients with MS have a higher co-morbidity with other ADs than controls, in particular in the time before onset of MS.
Patients and methods A database was established by linking the Danish MS Registry to The Danish National Patient Registry. We included all patients with definite MS according to the 2001 criteria of McDonald,16 with clinical onset from 2000 through 2004, within the 15 – 55 years of age interval. Each case studied was matched with 25 control persons by sex, year of birth and residential municipality by the reference year. We chose 25 controls per MS patient because a high number of controls carried no additional cost. We drew control persons at random from the Danish Civil Registration System. Those persons among the controls who had been identified as having MS in the Danish MS Registry were removed. Like the MS patients, controls had to be alive at the time of onset of the corresponding patients, and both cases and controls born in Denmark. The mean age at clinical onset of MS was 35.3 years. A detailed description of patients and methods was
previously published.17 Our MS onset cohort was chosen for the years 2000 – 2004, to have a sufficiently long retrospective observation period after the establishment of other registers. The follow-up for cases and controls was the same on average: 7 years. Using data from the Danish MS Registry database, we found that the 7-year cumulative survival probability was 97.9%, compared with 98.4% in the matched population, so the bias caused by differential survival among MS cases and controls was negligible. The Danish National Patient Register (NPR) is a nationwide registry that was established in 1977 and contains information about hospital admissions since 1977. Outpatient-hospital contacts were also included since 1995. From 1977 to 1993, the diagnoses were made in accordance with The International Classification of Diseases, version 8 (ICD 8); and after 1994, according to ICD version 10 (ICD 10).We chose the 18 most common ADs in both cohorts, to determine their association with MS (Table 1). We used data from the NPR from 1977 up to 2010, in this study. The ICD8 classification did not distinguish between Type 1 and Type 2 diabetes before 1987, but after that, and in the subsequent ICD10, a distinction was made. Since some milder forms of Type 1 diabetes mellitus (DM1) may be treated by general practitioners, we completed the data for these diseases with information from the Danish National Prescription Registry, which is an individual-level register containing data on all prescription drugs sold in Danish community pharmacies since 1994. The identification of the type of medication used was based on the second level of the Anatomical Therapeutic Chemical classification (ATC). For insulin and insulin analogs, we used the ATC code A10. A person who started insulin treatment below the age of 41, was treated with insulin preparations for more than 2 years and is registered in the NPR with the diagnosis DM1 was considered to have the diagnosis of DM1. For other diseases, we accepted one occurrence of the diagnoses as a main diagnosis, but for RA we required two occurrences.18 We used conditional logistic regression to calculate odds ratios (ORs) with a 95% confidence interval (CI) for a study subject developing an additional AD in two periods: The period prior to clinical onset of MS, as far back as permitted by our data sources and the age of the subjects; and the total observational period, up to and including 2010. With counts of less than 10 persons, exact confidence limits were calculated. We performed separate analyses for each AD and each sex. To determine the diagnostic sequences, we assessed the year of first recorded diagnosis of ADs, relative to the year of clinical onset of MS (which became the reference year for our controls).
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1246
Multiple Sclerosis Journal 20(9)
Table 1. Classification of autoimmune diseases. Disease
ICD8
ICD10
Diabetes mellitus Type I Rheumatoid arthritis Iridocyclitis Crohn’s disease Ulcerative colitis Glomerulonephritis Systemic lupus erythematosus Sjögren’s syndrome Graves’ disease Autoimmune thyroiditis Polyarteritis nodosa Dermatomyositis Systemic sclerosis Myasthenia gravis Juvenile rheumatoid arthritis Psoriasis vulgaris Ankylosing spondylitis Pemphigoid
249, 249.09 712.19, 712.39, 712.59 364.00 – 09 563.00 – 09 563.19, 569.04 582.00 – 09 734.19 734.90 242.00 245.03 446.09 716.09, 716.19 734.00 – 02, 734.08, 734.09 733.09 712.09 696.09, 696.10, 696.19 726.99, 727.09, 712.49 694.00 – 09
E10.0 – 9 M05.0 – 9,-M06.0 – 9 H 20.0, H20.1, H20.8, H20.9 K50.0, K50.1, K50.8, K50.9 K51.0 – K51.3, K51.8, K51.9 N02.01 – 09 M32.0, M32.1 M35.0 E05.0 E06.3 M30.0 M33.0 – 9 M34.0 – 9 G70.0 M08.0, M08.2 – 4, M08.8, M08.9 L40.0 – 9 M45.0 – 9, M08.1 L10.0 – 9, L12.0 – 9
ICD: International Classification of Diseases (with version number following).
We computed P-values for associations using the Fisher Exact Test, as many of the observations involved only a few index persons. We defined statistical significance at p < 0.05 and the correction for false discovery rate was done by the Benjamini-Hochberg method.19 We performed all statistical analyses and data management with the statistical package SPSS, version 19.
Results The study dataset consisted of 1403 MS cases, 939 women and 464 men, with a clinical onset from 2000 up to and including 2004, as well as 35,045 matched controls (23,455 women and 11,600 men). The female-to-male ratio was 2.02:1. The mean age of clinical onset of MS was 35.3 years. Of the MS patients, 16.3% (241) were primary progressive MS and 83.7% (1242) were relapsing–remitting MS from the onset. As a whole, the prevalence of AD in our cohort was low, and there was no general difference between MS cases and controls as to the frequency of having any ADs: In men, 3.66% of MS cases and 3.11% of controls had another autoimmune diagnosis (p = 0.59); and in women, 10.12% of MS cases and 9.42% of controls had an AD diagnosis (p = 0.47). Eight male (1.72%) and 14 female (1.49%) MS cases experienced comorbidity with more than one AD. Among the controls, 51 men (0.44%) and 322 women (1.37%) experienced more than one AD within the same time frame. The difference was statistically significant for men (p = 0.002), but not for women
(p = 0.92). The presence of more than one AD (other than MS) was, in MS cases as well as in controls, typically associated with DM1. The frequency of coexistence of other ADs for men and women with MS in the whole observational period is presented in Figure 1. For a better overview, the diseases not represented by MS-cases nor controls were excluded from the bar diagram. Male MS cases had a significantly higher probability of being diagnosed with DM1 before the clinical onset of MS (OR = 3.34; 95% CI 1.40 – 7.02; p = 0.008); however, there was no difference in women (p = 0.81). Male MS cases also had a significantly higher probability of Crohn’s disease than male controls, but only in the period preceding clinical onset of MS (OR = 5.03; 95% CI 1.18 – 16.10; p = 0.03), and this association was not seen in female MS cases (p = 0.82). For ulcerative colitis, there was a trend with the OR for the occurrence in male MS cases during the whole period of 2.22 (95% CI 0.93 – 4.59; p = 0.057); but not in female MS cases (p = 0.64). Male MS cases also were at a higher risk of developing systemic lupus erythematosus (SLE) (n = 2) prior to MS onset, compared to controls (OR = 12.55; 95% CI 1.62 – 69.95; p = 0.02). With only two cases, the importance of this observation was limited. There were no occurrences of pemphigus in male MS cases and it occurred only in one female MS case (p = 0.11). The mean age at diagnosis of the autoimmune comorbidities was 38.79 for the MS cases and 39.69 for the controls (p = 0.34).
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1247
Magyari et al.
Figure 1. Frequency of coexistence of autoimmune diseases with MS in the whole observational period. MS: multiple sclerosis; SLE: systemic lupus erythematosus.
When assuming a p-value limit of 0.0083 after correcting the standard value of 0.05 for false discovery rate, the increased OR for DM1 in male MS patients with a p-value limit of 0.008 remained statistically significant. The distribution of ADs with ORs for the period before onset of MS symptoms are presented in Figure 2 for men and in Figure 3 for women. For rare ADs with none of the MS cases affected by the disease, we show the upper 95% confidence limit for the zero ORs in the figures. None of the ADs were found to have a protective effect against developing MS, but as ADs are rare, some ADs were represented neither in the case nor in the control groups.
Discussion In our case-control study, we found no increased nor decreased risk of autoimmune comorbidity in female MS patients. Male patients, however, had an increased risk of DM1 and Crohn’s disease, but only in the period preceding onset of MS; however, only the significance for DM1 survived correction for multiple testing.
Although ADs are generally more common in women than in men, they did not occur more frequently in female MS patients than in female controls. In our study, the associations between MS and other ADs could only be found in men. MS cases with DM1 were more prone to have yet another AD. A higher co-occurrence of DM1 in MS patients and in their first-degree relatives, but without significant gender differences, is reported by another Danish study.20 MS has previously been shown to coexist with inflammatory bowel diseases.21,22 Other studies report co-occurrence of MS with other ADs, such as SLE,23 scleroderma,24 myasthenia gravis,25 psoriasis11 and RA.8,22 A Danish follow-up study of a large MS cohort shows a significantly lower risk of RA, as compared with a matched background (general non-MS control) population with a RR of 0.5,25 indicating there is a protective effect of MS against RA. This study was particularly designed to obtain a long follow-up and did not comprise the time before onset of MS, which was the main part of our study. This is in accordance with the genetic finding of a less pronounced genetic overlap with RA.13 Because of our much shorter follow-up, combined with a higher age at onset in RA, we failed to detect this effect.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1248
Multiple Sclerosis Journal 20(9)
Figure 2. Associations between MS and other investigated autoimmune diseases before clinical onset of MS, in men. The number of MS cases with the respective autoimmune disease are indicated. MS: multiple sclerosis; OR: odds ratio; SLE: systemic lupus erythematosus.
Figure 3. Associations between MS and other investigated autoimmune diseases before clinical onset of MS in women. The number of MS cases with the respective autoimmune disease are indicated. MS: multiple sclerosis; OR: odds ratio; SLE: systemic lupus erythematosus.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1249
Magyari et al. The prevalence of autoimmune thyroid disease is one of the most investigated co-morbidities in MS, and there is a large variation in the study designs and populations used. A significantly higher prevalence of autoimmune thyroiditis in female MS patients is reported in some studies,26 while others find a higher prevalence of autoimmune thyroiditis in male patients,27 or a significant co-occurrence without gender differences.28 Our study was too small to detect a higher risk of pemphigus in MS patients, as is reported in several studies.7,11 The strengths of the present registry-based study are that we used a population-based onset-defined MS cohort, an appropriately matched large control group derived unselected from the background (general) population and independent registry data, not influenced by the responsiveness or recall of the subjects. Furthermore, the population was ethnically homogeneous, only including individuals born in Denmark, to avoid problems with ethnic matching. We restricted inclusion of MS patients to the age interval 15 – 55 years, because there were very few patients with clinical onset of MS outside this interval. With a first demyelinating event below the age of 15, a diagnosis may be confused with acute disseminated encephalomyelitis (ADEM). We also decided to exclude these patients, because they would contribute with a very short backward time for the retrospective analysis. Patients with an onset age > 55 could not be followed far enough back in time, because the NPR was only established in 1977. The weakness of our study was the size of our MS cohort, which was limited to onset in the years 2000 – 2004, because we wanted to have a more recent MS onset cohort that allowed us to look sufficiently back in time; however, at the expense of follow-up time after onset of MS. This choice may also have caused underestimation of ADs with a higher age of onset than MS, e.g. RA, and reduce the statistical power to detect rarer AD combinations. Another weakness was that until 1995, hospital out-patients were not included in the NPR, and the patients in private practice were not included at all, which may explain the low frequency of ADs often treated in general practice alone, like autoimmune thyroid disease and psoriasis in both cases and controls, but this filter applied to MS patients as well as to controls. This limitation may apply to the other comorbidities studied and suggests the possibility that our findings were biased towards more severe diseases. Algorithms using multiple ICD codes derived from several sources were validated in a study from Manitoba,29 but health services are different in Denmark from many other countries. Practicing specialists only cover a small part of the activity, and in most cases more serious diseases are diagnosed and treated in highly-specialized hospital clinics, meaning that the probability of a disease being recorded in central hospital registries was higher than in countries
where the private sector has a higher share of the health care system. Another issue is the number of times that a patient must be registered in the NPR with a diagnosis for it to be considered valid. The problem is manifest for RA, where the validity of a first-time diagnosis in the NPR was rather low;18 however, it improved if the diagnosis was repeated. For this reason, we required at least two hospital contacts with the diagnosis RA as the main diagnosis. Index persons were classified with a diagnosis of AD if they had been admitted into a hospital or had been treated in an outpatient clinic with their main or secondary discharge diagnosis of an AD. Accepting one-time diagnoses could affect validity. The validity of the diverse diagnoses recorded in NPR depends on the disease.30–32 A validity study of RA in NPR points to a higher validity of the registered diagnosis for inpatient contacts and for more than one hospital registration with RA.18 Otherwise, with inflammatory bowel disease and Crohn’s disease, the validity as well as the completeness of the NPR is > 90% without a requirement for repeated diagnoses.33 The increase in specificity by requiring several contacts for a diagnosis to be confirmed would be at the expense of sensitivity, and a balance point may differ between the specific diseases. Another even more significant and bias-causing problem with requiring multiple contacts would be that patients with MS tend to have more hospital contacts than controls, and hence, their comorbidities could be noted repeatedly in the registries. All patients and controls were Danish born and alive at the time of disease onset of the MS patient, so immortal time bias up to the time of onset of MS in the retrospective part of the study was non-existent; however, in the 5 – 10 year follow-up window after MS onset, there is a possibility that some MS patients would have died early and provided less observational time than the controls, as the 7-year cumulative survival after onset of MS was 97.9% and in the matched background population, it was 98.4%. Thus, the bias caused by not taking into account differences in survival between MS patients and controls was negligible. Our study also confirmed that ADs were relatively infrequent, both in the general population and in MS patients. The coexistence of MS with some autoimmune disorders and the lack of association with others could lead to a better understanding of the complex pathogenesis and identification of common genetic risk factors for these diseases. The impact of ADs on MS patients is of importance for the individual patient. The estimated combined prevalence of having hospital-recorded ADs in Denmark was 5.3 %.34 Knowledge about comorbidities in MS patients is important, as it may affects clinical features, treatment decisions and responses, health outcomes and inclusion in clinical
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1250
Multiple Sclerosis Journal 20(9)
trials. An improved understanding of the autoimmune comorbidity in MS may give novel insights into shared pathophysiology, genetic predisposition and etiological risk factors for these diseases. Conflict of interest Melinda Magyari has served on scientific advisory board for Biogen Idec; TEVA, has received honoraria for lecturing from Biogen Idec, Merck Serono, Sanofi-Aventis, Teva; has received support for congress participation from Biogen Idec, Merck Serono, Novartis, Genzyme. Nils Koch-Henriksen has received honoraria for lecturing and participation in advisory councils, travel expenses for attending congresses and meetings, and financial support for monitoring the Danish MS Treatment Register from Bayer-Schering, MerckSerono, BiogenIdec, TEVA, Sanofi-Avensis and Novartis. The Danish Multiple Sclerosis Registry is funded by the Danish Multiple Sclerosis Society. Claudia Christina Pfleger has served on scientific advisory board for Novartis, Amirall, Biogen Idec; has received support for congress participation from Biogen Idec, Teva and Novartis. Per Soelberg Sorensen has served on scientific advisory boards Biogen Idec, Merck Serono, Novartis, Genmab, TEVA, Elan, GSK; has been on steering committees or independent data monitoring boards in clinical trials sponsored by Merck Serono, Genmab, TEVA, GSK, Bayer Schering, and he has received funding of travel for these activities; has received speaker honoraria from Biogen Idec, Merck Serono, TEVA, Bayer Schering, Sanofiaventis, Genzyme, and Novartis.
Funding This work was supported by the Danish MS Society.
References 1. Ramagopalan SV, Knight JC and Ebers GC. Multiple sclerosis and the major histocompatibility complex. Curr Opin Neurol 2009; 22: 219–225. 2. Gleicher N and Barad DH. Gender as risk factor for autoimmune diseases. J Autoimmun 2007; 28: 1–6. 3. Cooper GS, Bynum ML and Somers EC. Recent insights in the epidemiology of autoimmune diseases: Improved prevalence estimates and understanding of clustering of diseases. J Autoimmun 2009; 33: 197–207. 4. Marrie RA. Autoimmune disease and multiple sclerosis: Methods, methods, methods. Lancet Neurol 2007; 6: 575–576. 5. Marrie RA and Horwitz RI. Emerging effects of comorbidities on multiple sclerosis. Lancet Neurol 2010; 9: 820–828. 6. Edwards LJ and Constantinescu CS. A prospective study of conditions associated with multiple sclerosis in a cohort of 658 consecutive outpatients attending a multiple sclerosis clinic. Mult Scler 2004; 10: 575–581. 7. Langer-Gould A, Albers KB, Van Den Eeden SK, et al. Autoimmune diseases prior to the diagnosis of multiple sclerosis: A population-based case-control study. Mult Scler 2010; 16: 855–861.
8. Henderson RD, Bain CJ and Pender MP. The occurrence of autoimmune diseases in patients with multiple sclerosis and their families. J Clin Neurosci 2000; 7: 434–437. 9. Anaya JM, Kim-Howard X, Prahalad S, et al. Evaluation of genetic association between an ITGAM non-synonymous SNP (rs1143679) and multiple autoimmune diseases. Autoimmun Rev 2012; 11: 276–280. 10. Barcellos LF, Kamdar BB, Ramsay PP, et al. Clustering of autoimmune diseases in families with a high risk for multiple sclerosis: A descriptive study. Lancet Neurol 2006; 5: 924–931. 11. Nielsen NM, Frisch M, Rostgaard K, et al. Autoimmune diseases in patients with multiple sclerosis and their first-degree relatives: A nationwide cohort study in Denmark. Mult Scler 2008; 14: 823–829. 12. Lettre G and Rioux JD. Autoimmune diseases: Insights from genome-wide association studies. Hum Mol Genet 2008;17: R116–121. 13. Beecham AH, Patsopoulos NA, Xifara DK, et al. Analysis of immune-related loci identifies 48 new susceptibility variants for multiple sclerosis. Nat Genet 2013; 45: 1353–1360. 14. Criswell LA, Pfeiffer KA, Lum RF, et al. Analysis of families in the multiple autoimmune disease genetics consortium (MADGC) collection: The PTPN22 620W allele associates with multiple autoimmune phenotypes. Am J Hum Genet 2005; 76: 561–571. 15. Koch-Henriksen N and Sorensen PS. The changing demographic pattern of multiple sclerosis epidemiology. Lancet Neurol 2010; 9: 520–532. 16. McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: Guidelines from the International Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol 2001; 50: 121–127. 17. Magyari M, Koch-Henriksen N, Pfleger CC, et al. Reproduction and the risk of multiple sclerosis. Mult Scler 2013; 18 Mult Scler 2013; 19: 1604–1609. 18. Pedersen M, Klarlund M, Jacobsen S, et al. Validity of rheumatoid arthritis diagnoses in the Danish National Patient Registry. Eur J Epidemiol 2004; 19: 1097–1103. 19. Benjamini H and Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Statist Soc B 1995; 57: 289–300. 20. Nielsen NM, Westergaard T, Frisch M, et al. Type 1 diabetes and multiple sclerosis: A Danish population-based cohort study. Arch Neurol 2006; 63: 1001–1004. 21. Gupta G, Gelfand JM and Lewis JD. Increased risk for demyelinating diseases in patients with inflammatory bowel disease. Gastroenterology 2005; 129: 819–826. 22. Midgard R, Gronning M, Riise T, et al. Multiple sclerosis and chronic inflammatory diseases. A case-control study. Acta Neurol Scand 1996; 93: 322–328. 23. Hietaharju A, Peltola J, Seppa J, et al. The coexistence of systemic lupus erythematosus and multiple sclerosis in a mother and daughter. Scand J Rheumatol 2001; 30: 120–122. 24. Trostle DC, Helfrich D and Medsger TA Jr. Systemic sclerosis (scleroderma) and multiple sclerosis. Arthrit Rheum 1986; 29: 124–127. 25. Shakir RA, Hussien JM and Trontelj JV. Myasthenia gravis and multiple sclerosis. J Neuroimmunol 1983; 4: 161–165.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1251
Magyari et al. 26. Munteis E, Cano JF, Flores JA, et al. Prevalence of autoimmune thyroid disorders in a Spanish multiple sclerosis cohort. Eur J Neurol 2007; 14: 1048–1052. 27. Niederwieser G, Buchinger W, Bonelli RM, et al. Prevalence of autoimmune thyroiditis and non-immune thyroid disease in multiple sclerosis. J Neurol 2003; 250: 672–675. 28. Sloka JS, Phillips PW, Stefanelli M, et al. Co-occurrence of autoimmune thyroid disease in a multiple sclerosis cohort. J Autoimmune Dis 2005; 2: 9:2–9. 29. Marrie RA, Yu BN, Leung S, et al. Rising prevalence of vascular comorbidities in multiple sclerosis: Validation of administrative definitions for diabetes, hypertension and hyperlipidemia. Mult Scler 2012; 18: 1310–1319. 30. Vest-Hansen B, Riis AH and Christiansen CF. Registration of acute medical hospital admissions in the Danish National
31.
32.
33.
34.
Patient Registry: A validation study. Clin Epidemiol 2013; 5: 129–133. Madsen M, Balling H and Eriksen LS. [The validity of the diagnosis of acute myocardial infarction in 2 registries: The Heart Registry compared to the National Patient Registry]. Ugeskr Laeger 1990; 152: 308–314. Nickelsen TN. [Data validity and coverage in the Danish National Health Registry. A literature review]. Ugeskr Laeger 2001; 164: 33–37. Fonager K, Sorensen HT, Rasmussen SN, et al. Assessment of the diagnoses of Crohn’s disease and ulcerative colitis in a Danish hospital information system. Scand J Gastroenterol 1996; 31: 154–159. Eaton WW, Rose NR, Kalaydjian A, et al. Epidemiology of autoimmune diseases in Denmark. J Autoimmun 2007; 29: 1–9.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
MSJ0010.1177/1352458514521515Multiple Sclerosis JournalMagyari et al.
MULTIPLE SCLEROSIS MSJ JOURNAL
Research Paper
Gender and autoimmune comorbidity in multiple sclerosis
Multiple Sclerosis Journal 2014, Vol. 20(9) 1244–1251 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1352458514521515 msj.sagepub.com
Melinda Magyari1,2,3, Nils Koch-Henriksen2,4, Claudia C Pfleger5 and Per Soelberg Sørensen1,3
Abstract Background: The female preponderance in incidence of multiple sclerosis (MS) calls for investigations into sex differences in comorbidity with other autoimmune diseases (ADs). Objectives: To determine whether male and female patients with MS have a higher frequency of autoimmune comorbidity than controls, and to describe the type and frequency of ADs that are associated with MS. Methods: Our database was established by linkage of the Danish MS Registry to The Danish National Patient Register and consisted of 1403 patients of both sexes with clinical onset of MS between 2000 and 2004, and 25 matched controls for every case. Results: None of the ADs occurred more frequently in female cases than in controls. Male cases were more likely to have Type I diabetes mellitus (odds ratio (OR) = 3.34; 95% CI 1.40 – 7.02; p < 0.008), Crohn’s disease (OR = 5.03; 95% CI 1.18 – 16.10; p = 0.03) and systemic lupus erythematosus (OR = 12.55; 95% CI 1.62 – 69.95; p = 0.02) than male controls. Conclusions: Autoimmune disorders are rare, but some of them tend to occur together with MS at a higher rate than in controls. Although women are generally more prone to ADs than men, significantly increased occurrence of other ADs were only found in male MS patients. Keywords Autoimmune disease, comorbidity, Crohn’s disease, Denmark, diabetes, epidemiology, gender, lupus, multiple sclerosis, registry Date received: 30 July 2013; accepted: 15 December 2013
Introduction Multiple sclerosis (MS) is considered to be an autoimmune disease1 (AD) that, in common with the majority of ADs, occurs more frequently in women.2 The strong female preponderance in incidence of MS calls for investigations into sex differences in comorbidity with other ADs in MS patients, to see whether and to what degree this gender difference might deviate from the population’s gender difference for ADs. ADs are treated by different medical specialties and have conventionally been studied as separate entities,3 but the amount of literature suggesting the co-existence of MS with other autoimmune disorders is growing, although some inconsistent findings in various studies have been pointed out.4,5 The aetiology of ADs is not fully understood, but cooccurrence of various ADs suggests common features in their aetiology and pathogenesis.
Several studies using different methods and designs investigated the co-existence of ADs in MS patients6,7 and the co-occurrence of autoimmune disorders within families,8 but did not establish any common predisposing 1Danish
Multiple Sclerosis Center, Department of Neurology, Neuroscience Centre, Rigshospitalet, Copenhagen, Denmark. 2Danish Multiple Sclerosis Registry, Neuroscience Centre, Rigshospitalet, Copenhagen, Denmark. 3University of Copenhagen, Denmark. 4Clinical Institute, Department of Clinical Epidemiology, University of Aarhus, Denmark. 5Department of Neurology, Aalborg University Hospital, Denmark. Corresponding author: Melinda Magyari, Danish Multiple Sclerosis Research Center, Department of Neurology, Neuroscience Centre, Rigshospitalet, section 2082, Blegdamsvej 9, Copenhagen, 2100, Denmark. Email: [email protected]
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1245
Magyari et al. factors to autoimmunity.9 As ADs are shown to be more common in family members of patients with MS than in the general population,10,11 a common genetic and environmental background for susceptibility to autoimmunity is presumed. Genome-wide association studies (GWAS) show an overlap in the genetic risk foci for MS and some other core ADs.12 Recently, ImmunoChip analyses disclosed overlapping of approximately 22% of MS signals with at least one other AD signal, most frequently with inflammatory bowel disease and less frequently with rheumatoid arthritis (RA) and autoimmune thyroid disease.13 However, in the Multiple Autoimmune Disease Genetics Consortium (MADGC) family study, some genetic loci for susceptibility to autoimmunity were not common to MS and other ADs.14 Co-morbidity has different aspects, and the occurrence of some ADs can be either more or less frequent in patients with MS, compared to the general population. MS may either predispose or protect against later onset of some other diseases or, conversely, some diseases may predispose or protect against later onset of MS. Common traits may affect the risk of MS and other ADs independently. If so, the order of onset of the diseases is not important. The high and increasing female:male sex ratio in MS15 underlines the need to search for gender differences in aetiology or pathogenesis. Autoimmune comorbidity in MS could be one of the factors that differentiates male from female MS patients, taking into consideration that most ADs are more common in women than in men.2 The main objective of this study is to determine whether male and female patients with MS have a higher co-morbidity with other ADs than controls, in particular in the time before onset of MS.
Patients and methods A database was established by linking the Danish MS Registry to The Danish National Patient Registry. We included all patients with definite MS according to the 2001 criteria of McDonald,16 with clinical onset from 2000 through 2004, within the 15 – 55 years of age interval. Each case studied was matched with 25 control persons by sex, year of birth and residential municipality by the reference year. We chose 25 controls per MS patient because a high number of controls carried no additional cost. We drew control persons at random from the Danish Civil Registration System. Those persons among the controls who had been identified as having MS in the Danish MS Registry were removed. Like the MS patients, controls had to be alive at the time of onset of the corresponding patients, and both cases and controls born in Denmark. The mean age at clinical onset of MS was 35.3 years. A detailed description of patients and methods was
previously published.17 Our MS onset cohort was chosen for the years 2000 – 2004, to have a sufficiently long retrospective observation period after the establishment of other registers. The follow-up for cases and controls was the same on average: 7 years. Using data from the Danish MS Registry database, we found that the 7-year cumulative survival probability was 97.9%, compared with 98.4% in the matched population, so the bias caused by differential survival among MS cases and controls was negligible. The Danish National Patient Register (NPR) is a nationwide registry that was established in 1977 and contains information about hospital admissions since 1977. Outpatient-hospital contacts were also included since 1995. From 1977 to 1993, the diagnoses were made in accordance with The International Classification of Diseases, version 8 (ICD 8); and after 1994, according to ICD version 10 (ICD 10).We chose the 18 most common ADs in both cohorts, to determine their association with MS (Table 1). We used data from the NPR from 1977 up to 2010, in this study. The ICD8 classification did not distinguish between Type 1 and Type 2 diabetes before 1987, but after that, and in the subsequent ICD10, a distinction was made. Since some milder forms of Type 1 diabetes mellitus (DM1) may be treated by general practitioners, we completed the data for these diseases with information from the Danish National Prescription Registry, which is an individual-level register containing data on all prescription drugs sold in Danish community pharmacies since 1994. The identification of the type of medication used was based on the second level of the Anatomical Therapeutic Chemical classification (ATC). For insulin and insulin analogs, we used the ATC code A10. A person who started insulin treatment below the age of 41, was treated with insulin preparations for more than 2 years and is registered in the NPR with the diagnosis DM1 was considered to have the diagnosis of DM1. For other diseases, we accepted one occurrence of the diagnoses as a main diagnosis, but for RA we required two occurrences.18 We used conditional logistic regression to calculate odds ratios (ORs) with a 95% confidence interval (CI) for a study subject developing an additional AD in two periods: The period prior to clinical onset of MS, as far back as permitted by our data sources and the age of the subjects; and the total observational period, up to and including 2010. With counts of less than 10 persons, exact confidence limits were calculated. We performed separate analyses for each AD and each sex. To determine the diagnostic sequences, we assessed the year of first recorded diagnosis of ADs, relative to the year of clinical onset of MS (which became the reference year for our controls).
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1246
Multiple Sclerosis Journal 20(9)
Table 1. Classification of autoimmune diseases. Disease
ICD8
ICD10
Diabetes mellitus Type I Rheumatoid arthritis Iridocyclitis Crohn’s disease Ulcerative colitis Glomerulonephritis Systemic lupus erythematosus Sjögren’s syndrome Graves’ disease Autoimmune thyroiditis Polyarteritis nodosa Dermatomyositis Systemic sclerosis Myasthenia gravis Juvenile rheumatoid arthritis Psoriasis vulgaris Ankylosing spondylitis Pemphigoid
249, 249.09 712.19, 712.39, 712.59 364.00 – 09 563.00 – 09 563.19, 569.04 582.00 – 09 734.19 734.90 242.00 245.03 446.09 716.09, 716.19 734.00 – 02, 734.08, 734.09 733.09 712.09 696.09, 696.10, 696.19 726.99, 727.09, 712.49 694.00 – 09
E10.0 – 9 M05.0 – 9,-M06.0 – 9 H 20.0, H20.1, H20.8, H20.9 K50.0, K50.1, K50.8, K50.9 K51.0 – K51.3, K51.8, K51.9 N02.01 – 09 M32.0, M32.1 M35.0 E05.0 E06.3 M30.0 M33.0 – 9 M34.0 – 9 G70.0 M08.0, M08.2 – 4, M08.8, M08.9 L40.0 – 9 M45.0 – 9, M08.1 L10.0 – 9, L12.0 – 9
ICD: International Classification of Diseases (with version number following).
We computed P-values for associations using the Fisher Exact Test, as many of the observations involved only a few index persons. We defined statistical significance at p < 0.05 and the correction for false discovery rate was done by the Benjamini-Hochberg method.19 We performed all statistical analyses and data management with the statistical package SPSS, version 19.
Results The study dataset consisted of 1403 MS cases, 939 women and 464 men, with a clinical onset from 2000 up to and including 2004, as well as 35,045 matched controls (23,455 women and 11,600 men). The female-to-male ratio was 2.02:1. The mean age of clinical onset of MS was 35.3 years. Of the MS patients, 16.3% (241) were primary progressive MS and 83.7% (1242) were relapsing–remitting MS from the onset. As a whole, the prevalence of AD in our cohort was low, and there was no general difference between MS cases and controls as to the frequency of having any ADs: In men, 3.66% of MS cases and 3.11% of controls had another autoimmune diagnosis (p = 0.59); and in women, 10.12% of MS cases and 9.42% of controls had an AD diagnosis (p = 0.47). Eight male (1.72%) and 14 female (1.49%) MS cases experienced comorbidity with more than one AD. Among the controls, 51 men (0.44%) and 322 women (1.37%) experienced more than one AD within the same time frame. The difference was statistically significant for men (p = 0.002), but not for women
(p = 0.92). The presence of more than one AD (other than MS) was, in MS cases as well as in controls, typically associated with DM1. The frequency of coexistence of other ADs for men and women with MS in the whole observational period is presented in Figure 1. For a better overview, the diseases not represented by MS-cases nor controls were excluded from the bar diagram. Male MS cases had a significantly higher probability of being diagnosed with DM1 before the clinical onset of MS (OR = 3.34; 95% CI 1.40 – 7.02; p = 0.008); however, there was no difference in women (p = 0.81). Male MS cases also had a significantly higher probability of Crohn’s disease than male controls, but only in the period preceding clinical onset of MS (OR = 5.03; 95% CI 1.18 – 16.10; p = 0.03), and this association was not seen in female MS cases (p = 0.82). For ulcerative colitis, there was a trend with the OR for the occurrence in male MS cases during the whole period of 2.22 (95% CI 0.93 – 4.59; p = 0.057); but not in female MS cases (p = 0.64). Male MS cases also were at a higher risk of developing systemic lupus erythematosus (SLE) (n = 2) prior to MS onset, compared to controls (OR = 12.55; 95% CI 1.62 – 69.95; p = 0.02). With only two cases, the importance of this observation was limited. There were no occurrences of pemphigus in male MS cases and it occurred only in one female MS case (p = 0.11). The mean age at diagnosis of the autoimmune comorbidities was 38.79 for the MS cases and 39.69 for the controls (p = 0.34).
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1247
Magyari et al.
Figure 1. Frequency of coexistence of autoimmune diseases with MS in the whole observational period. MS: multiple sclerosis; SLE: systemic lupus erythematosus.
When assuming a p-value limit of 0.0083 after correcting the standard value of 0.05 for false discovery rate, the increased OR for DM1 in male MS patients with a p-value limit of 0.008 remained statistically significant. The distribution of ADs with ORs for the period before onset of MS symptoms are presented in Figure 2 for men and in Figure 3 for women. For rare ADs with none of the MS cases affected by the disease, we show the upper 95% confidence limit for the zero ORs in the figures. None of the ADs were found to have a protective effect against developing MS, but as ADs are rare, some ADs were represented neither in the case nor in the control groups.
Discussion In our case-control study, we found no increased nor decreased risk of autoimmune comorbidity in female MS patients. Male patients, however, had an increased risk of DM1 and Crohn’s disease, but only in the period preceding onset of MS; however, only the significance for DM1 survived correction for multiple testing.
Although ADs are generally more common in women than in men, they did not occur more frequently in female MS patients than in female controls. In our study, the associations between MS and other ADs could only be found in men. MS cases with DM1 were more prone to have yet another AD. A higher co-occurrence of DM1 in MS patients and in their first-degree relatives, but without significant gender differences, is reported by another Danish study.20 MS has previously been shown to coexist with inflammatory bowel diseases.21,22 Other studies report co-occurrence of MS with other ADs, such as SLE,23 scleroderma,24 myasthenia gravis,25 psoriasis11 and RA.8,22 A Danish follow-up study of a large MS cohort shows a significantly lower risk of RA, as compared with a matched background (general non-MS control) population with a RR of 0.5,25 indicating there is a protective effect of MS against RA. This study was particularly designed to obtain a long follow-up and did not comprise the time before onset of MS, which was the main part of our study. This is in accordance with the genetic finding of a less pronounced genetic overlap with RA.13 Because of our much shorter follow-up, combined with a higher age at onset in RA, we failed to detect this effect.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1248
Multiple Sclerosis Journal 20(9)
Figure 2. Associations between MS and other investigated autoimmune diseases before clinical onset of MS, in men. The number of MS cases with the respective autoimmune disease are indicated. MS: multiple sclerosis; OR: odds ratio; SLE: systemic lupus erythematosus.
Figure 3. Associations between MS and other investigated autoimmune diseases before clinical onset of MS in women. The number of MS cases with the respective autoimmune disease are indicated. MS: multiple sclerosis; OR: odds ratio; SLE: systemic lupus erythematosus.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1249
Magyari et al. The prevalence of autoimmune thyroid disease is one of the most investigated co-morbidities in MS, and there is a large variation in the study designs and populations used. A significantly higher prevalence of autoimmune thyroiditis in female MS patients is reported in some studies,26 while others find a higher prevalence of autoimmune thyroiditis in male patients,27 or a significant co-occurrence without gender differences.28 Our study was too small to detect a higher risk of pemphigus in MS patients, as is reported in several studies.7,11 The strengths of the present registry-based study are that we used a population-based onset-defined MS cohort, an appropriately matched large control group derived unselected from the background (general) population and independent registry data, not influenced by the responsiveness or recall of the subjects. Furthermore, the population was ethnically homogeneous, only including individuals born in Denmark, to avoid problems with ethnic matching. We restricted inclusion of MS patients to the age interval 15 – 55 years, because there were very few patients with clinical onset of MS outside this interval. With a first demyelinating event below the age of 15, a diagnosis may be confused with acute disseminated encephalomyelitis (ADEM). We also decided to exclude these patients, because they would contribute with a very short backward time for the retrospective analysis. Patients with an onset age > 55 could not be followed far enough back in time, because the NPR was only established in 1977. The weakness of our study was the size of our MS cohort, which was limited to onset in the years 2000 – 2004, because we wanted to have a more recent MS onset cohort that allowed us to look sufficiently back in time; however, at the expense of follow-up time after onset of MS. This choice may also have caused underestimation of ADs with a higher age of onset than MS, e.g. RA, and reduce the statistical power to detect rarer AD combinations. Another weakness was that until 1995, hospital out-patients were not included in the NPR, and the patients in private practice were not included at all, which may explain the low frequency of ADs often treated in general practice alone, like autoimmune thyroid disease and psoriasis in both cases and controls, but this filter applied to MS patients as well as to controls. This limitation may apply to the other comorbidities studied and suggests the possibility that our findings were biased towards more severe diseases. Algorithms using multiple ICD codes derived from several sources were validated in a study from Manitoba,29 but health services are different in Denmark from many other countries. Practicing specialists only cover a small part of the activity, and in most cases more serious diseases are diagnosed and treated in highly-specialized hospital clinics, meaning that the probability of a disease being recorded in central hospital registries was higher than in countries
where the private sector has a higher share of the health care system. Another issue is the number of times that a patient must be registered in the NPR with a diagnosis for it to be considered valid. The problem is manifest for RA, where the validity of a first-time diagnosis in the NPR was rather low;18 however, it improved if the diagnosis was repeated. For this reason, we required at least two hospital contacts with the diagnosis RA as the main diagnosis. Index persons were classified with a diagnosis of AD if they had been admitted into a hospital or had been treated in an outpatient clinic with their main or secondary discharge diagnosis of an AD. Accepting one-time diagnoses could affect validity. The validity of the diverse diagnoses recorded in NPR depends on the disease.30–32 A validity study of RA in NPR points to a higher validity of the registered diagnosis for inpatient contacts and for more than one hospital registration with RA.18 Otherwise, with inflammatory bowel disease and Crohn’s disease, the validity as well as the completeness of the NPR is > 90% without a requirement for repeated diagnoses.33 The increase in specificity by requiring several contacts for a diagnosis to be confirmed would be at the expense of sensitivity, and a balance point may differ between the specific diseases. Another even more significant and bias-causing problem with requiring multiple contacts would be that patients with MS tend to have more hospital contacts than controls, and hence, their comorbidities could be noted repeatedly in the registries. All patients and controls were Danish born and alive at the time of disease onset of the MS patient, so immortal time bias up to the time of onset of MS in the retrospective part of the study was non-existent; however, in the 5 – 10 year follow-up window after MS onset, there is a possibility that some MS patients would have died early and provided less observational time than the controls, as the 7-year cumulative survival after onset of MS was 97.9% and in the matched background population, it was 98.4%. Thus, the bias caused by not taking into account differences in survival between MS patients and controls was negligible. Our study also confirmed that ADs were relatively infrequent, both in the general population and in MS patients. The coexistence of MS with some autoimmune disorders and the lack of association with others could lead to a better understanding of the complex pathogenesis and identification of common genetic risk factors for these diseases. The impact of ADs on MS patients is of importance for the individual patient. The estimated combined prevalence of having hospital-recorded ADs in Denmark was 5.3 %.34 Knowledge about comorbidities in MS patients is important, as it may affects clinical features, treatment decisions and responses, health outcomes and inclusion in clinical
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1250
Multiple Sclerosis Journal 20(9)
trials. An improved understanding of the autoimmune comorbidity in MS may give novel insights into shared pathophysiology, genetic predisposition and etiological risk factors for these diseases. Conflict of interest Melinda Magyari has served on scientific advisory board for Biogen Idec; TEVA, has received honoraria for lecturing from Biogen Idec, Merck Serono, Sanofi-Aventis, Teva; has received support for congress participation from Biogen Idec, Merck Serono, Novartis, Genzyme. Nils Koch-Henriksen has received honoraria for lecturing and participation in advisory councils, travel expenses for attending congresses and meetings, and financial support for monitoring the Danish MS Treatment Register from Bayer-Schering, MerckSerono, BiogenIdec, TEVA, Sanofi-Avensis and Novartis. The Danish Multiple Sclerosis Registry is funded by the Danish Multiple Sclerosis Society. Claudia Christina Pfleger has served on scientific advisory board for Novartis, Amirall, Biogen Idec; has received support for congress participation from Biogen Idec, Teva and Novartis. Per Soelberg Sorensen has served on scientific advisory boards Biogen Idec, Merck Serono, Novartis, Genmab, TEVA, Elan, GSK; has been on steering committees or independent data monitoring boards in clinical trials sponsored by Merck Serono, Genmab, TEVA, GSK, Bayer Schering, and he has received funding of travel for these activities; has received speaker honoraria from Biogen Idec, Merck Serono, TEVA, Bayer Schering, Sanofiaventis, Genzyme, and Novartis.
Funding This work was supported by the Danish MS Society.
References 1. Ramagopalan SV, Knight JC and Ebers GC. Multiple sclerosis and the major histocompatibility complex. Curr Opin Neurol 2009; 22: 219–225. 2. Gleicher N and Barad DH. Gender as risk factor for autoimmune diseases. J Autoimmun 2007; 28: 1–6. 3. Cooper GS, Bynum ML and Somers EC. Recent insights in the epidemiology of autoimmune diseases: Improved prevalence estimates and understanding of clustering of diseases. J Autoimmun 2009; 33: 197–207. 4. Marrie RA. Autoimmune disease and multiple sclerosis: Methods, methods, methods. Lancet Neurol 2007; 6: 575–576. 5. Marrie RA and Horwitz RI. Emerging effects of comorbidities on multiple sclerosis. Lancet Neurol 2010; 9: 820–828. 6. Edwards LJ and Constantinescu CS. A prospective study of conditions associated with multiple sclerosis in a cohort of 658 consecutive outpatients attending a multiple sclerosis clinic. Mult Scler 2004; 10: 575–581. 7. Langer-Gould A, Albers KB, Van Den Eeden SK, et al. Autoimmune diseases prior to the diagnosis of multiple sclerosis: A population-based case-control study. Mult Scler 2010; 16: 855–861.
8. Henderson RD, Bain CJ and Pender MP. The occurrence of autoimmune diseases in patients with multiple sclerosis and their families. J Clin Neurosci 2000; 7: 434–437. 9. Anaya JM, Kim-Howard X, Prahalad S, et al. Evaluation of genetic association between an ITGAM non-synonymous SNP (rs1143679) and multiple autoimmune diseases. Autoimmun Rev 2012; 11: 276–280. 10. Barcellos LF, Kamdar BB, Ramsay PP, et al. Clustering of autoimmune diseases in families with a high risk for multiple sclerosis: A descriptive study. Lancet Neurol 2006; 5: 924–931. 11. Nielsen NM, Frisch M, Rostgaard K, et al. Autoimmune diseases in patients with multiple sclerosis and their first-degree relatives: A nationwide cohort study in Denmark. Mult Scler 2008; 14: 823–829. 12. Lettre G and Rioux JD. Autoimmune diseases: Insights from genome-wide association studies. Hum Mol Genet 2008;17: R116–121. 13. Beecham AH, Patsopoulos NA, Xifara DK, et al. Analysis of immune-related loci identifies 48 new susceptibility variants for multiple sclerosis. Nat Genet 2013; 45: 1353–1360. 14. Criswell LA, Pfeiffer KA, Lum RF, et al. Analysis of families in the multiple autoimmune disease genetics consortium (MADGC) collection: The PTPN22 620W allele associates with multiple autoimmune phenotypes. Am J Hum Genet 2005; 76: 561–571. 15. Koch-Henriksen N and Sorensen PS. The changing demographic pattern of multiple sclerosis epidemiology. Lancet Neurol 2010; 9: 520–532. 16. McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: Guidelines from the International Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol 2001; 50: 121–127. 17. Magyari M, Koch-Henriksen N, Pfleger CC, et al. Reproduction and the risk of multiple sclerosis. Mult Scler 2013; 18 Mult Scler 2013; 19: 1604–1609. 18. Pedersen M, Klarlund M, Jacobsen S, et al. Validity of rheumatoid arthritis diagnoses in the Danish National Patient Registry. Eur J Epidemiol 2004; 19: 1097–1103. 19. Benjamini H and Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Statist Soc B 1995; 57: 289–300. 20. Nielsen NM, Westergaard T, Frisch M, et al. Type 1 diabetes and multiple sclerosis: A Danish population-based cohort study. Arch Neurol 2006; 63: 1001–1004. 21. Gupta G, Gelfand JM and Lewis JD. Increased risk for demyelinating diseases in patients with inflammatory bowel disease. Gastroenterology 2005; 129: 819–826. 22. Midgard R, Gronning M, Riise T, et al. Multiple sclerosis and chronic inflammatory diseases. A case-control study. Acta Neurol Scand 1996; 93: 322–328. 23. Hietaharju A, Peltola J, Seppa J, et al. The coexistence of systemic lupus erythematosus and multiple sclerosis in a mother and daughter. Scand J Rheumatol 2001; 30: 120–122. 24. Trostle DC, Helfrich D and Medsger TA Jr. Systemic sclerosis (scleroderma) and multiple sclerosis. Arthrit Rheum 1986; 29: 124–127. 25. Shakir RA, Hussien JM and Trontelj JV. Myasthenia gravis and multiple sclerosis. J Neuroimmunol 1983; 4: 161–165.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
1251
Magyari et al. 26. Munteis E, Cano JF, Flores JA, et al. Prevalence of autoimmune thyroid disorders in a Spanish multiple sclerosis cohort. Eur J Neurol 2007; 14: 1048–1052. 27. Niederwieser G, Buchinger W, Bonelli RM, et al. Prevalence of autoimmune thyroiditis and non-immune thyroid disease in multiple sclerosis. J Neurol 2003; 250: 672–675. 28. Sloka JS, Phillips PW, Stefanelli M, et al. Co-occurrence of autoimmune thyroid disease in a multiple sclerosis cohort. J Autoimmune Dis 2005; 2: 9:2–9. 29. Marrie RA, Yu BN, Leung S, et al. Rising prevalence of vascular comorbidities in multiple sclerosis: Validation of administrative definitions for diabetes, hypertension and hyperlipidemia. Mult Scler 2012; 18: 1310–1319. 30. Vest-Hansen B, Riis AH and Christiansen CF. Registration of acute medical hospital admissions in the Danish National
31.
32.
33.
34.
Patient Registry: A validation study. Clin Epidemiol 2013; 5: 129–133. Madsen M, Balling H and Eriksen LS. [The validity of the diagnosis of acute myocardial infarction in 2 registries: The Heart Registry compared to the National Patient Registry]. Ugeskr Laeger 1990; 152: 308–314. Nickelsen TN. [Data validity and coverage in the Danish National Health Registry. A literature review]. Ugeskr Laeger 2001; 164: 33–37. Fonager K, Sorensen HT, Rasmussen SN, et al. Assessment of the diagnoses of Crohn’s disease and ulcerative colitis in a Danish hospital information system. Scand J Gastroenterol 1996; 31: 154–159. Eaton WW, Rose NR, Kalaydjian A, et al. Epidemiology of autoimmune diseases in Denmark. J Autoimmun 2007; 29: 1–9.
Downloaded from msj.sagepub.com at NORTH DAKOTA STATE UNIV LIB on May 25, 2015
