Familial History of Autoimmune Disorders Among Patients With Pediatric Multiple Sclerosis.

BACKGROUND AND
OBJECTIVE: The objective of this study was to determine whether family members of patients with pediatric multiple sclerosis (MS) have an increased prevalence of autoimmune conditions compared with controls.
METHODS: Data collected during a pediatric MS case-control study of risk factors included information about various autoimmune diseases in family members. The frequency of these disorders was compared between cases and controls.
RESULTS: There was an increased rate of autoimmune diseases among family members of pediatric MS cases compared with controls with first-degree history of MS excluded (OR = 2.27, 95% CI 1.71-3.01, p < 0.001). There was an increased rate of MS among second-degree relatives of pediatric MS cases compared with controls (OR = 3.47, 95% CI 1.36-8.86, p = 0.009). The OR for MS was 2.64 when restricted to maternal relatives and 6.37 when restricted to paternal relatives. DISCUSSION: The increased rates of autoimmune disorders, including thyroid disorders and MS among families of patients with pediatric MS, suggest shared genetic factors among families with children diagnosed with pediatric MS.

Multiple sclerosis (MS) is an autoimmune disorder targeting the CNS that is most commonly diagnosed in women around age 30 years.[1] An estimated 5% of all MS cases have onset under age 18 years.[2] Multiple genetic and environmental risks have been identified for MS.[1] Studying pediatric MS provides a unique opportunity to gain insights into the pathophysiology and causes of MS because of the relatively fewer number of irrelevant exposures preceding disease onset and the possibility that larger exposures result in earlier onset.

Genetic risk factors for pediatric MS have been reported that overlap with adult MS. These include HLA-DRB1*15:01 and multiple class III risk variants.[3] Several risk factors for autoimmune disorders, including genetic variants or vitamin D deficiency, are shared among various conditions, including MS.[4-10] There have been conflicting reports about rates of autoimmune diseases in families of patients with adult MS.[11,12] Data consistently, however, note that MS risk is increased among first-degree relatives of patients with MS, including children of parents with MS.[13] This study sought to understand the incidence of a history of autoimmune diseases in family members of pediatric MS cases compared with pediatric controls.

Methods

Participants of this case-control study included children with MS or clinically isolated syndrome and healthy pediatric subjects recruited at 17 clinics participating as part of a nationwide study of environmental risk factors in pediatric MS between November 1, 2011, and July 1, 2017 (R01NS071463, PI Waubant). Diagnosis and control status was established by the treating neurologist and confirmed by adherence to published criteria for pediatric demyelinating disease as determined by a panel of 3 pediatric MS specialists using the 2010 McDonald Criteria. Centers obtained institutional review board approval as well as written informed consent from parents of pediatric participants and assent as appropriate from children. Criteria mandated that cases were enrolled within 4 years of symptom onset. Healthy controls were recruited from primary care, urgent care, or other pediatric clinics at the participating institutions. Eligibility criteria for controls required (1) absence of any ongoing autoimmune disease (apart from eczema and asthma) and (2) no history of MS in any immediate, biological family member. Demographic data including race and ethnicity, consistent with NIH guidelines, were provided by parents of pediatric participants. Standardized medical history questionnaires were completed by patients and families including family autoimmune disease history.

All analyses were conducted using SAS 9.4 (SAS Institute; Cary, NC). Patient characteristics were summarized using medians and quartiles or counts and percentages as appropriate. Chi-squared tests were used to compare reported family history of autoimmune disease between cases and controls, for both overall history of any disease and for specific diseases (e.g., MS).

Logistic regression models were used to test for differences between cases and controls in reporting a family history of autoimmune diseases, when adjusting for sex, race, age, ethnicity, and mother's education level. ORs and 95% CIs were calculated based on the logistic regression models.

The diseases that were found to be significant in the overall tests were investigated within 5 relative groups: first-degree relatives, second-degree relatives, parents, maternal relatives, and paternal relatives. The same statistical methods applied to the overall analyses were implemented in the relative group analyses. When expected cell counts were less than 5, Fisher exact tests were performed in place of χ2 tests. A significance level of 0.05 was used for all analyses.

Data Availability

Deidentified data sets from this study can be made available to qualified investigators with appropriate ethics/IRB approval. Data are stored for up to 5 years postpublication.

Results

Participants' Characteristics

This study included 1,204 participants with 709 controls and 495 cases (37 of which were categorized as clinically isolated syndrome). The breakdown of cases and controls based on age, sex, race, and ethnicity is summarized in Table 1. Groups were well matched except for ethnicity and mother's education.

Table 1

Subject Characteristics

Family History of Autoimmune Disorders

Because controls could not have a history of MS in immediate biological family members (study exclusion criteria), analyses involving familial MS excluded first-degree relatives. There was a significant difference in prevalence of a family history of autoimmune disease in patients with MS compared with controls (68.1% vs 49.5%, p < 0.001) (Table 2). The conditions with significant differences in prevalence between cases and controls were childhood-onset and adult-onset diabetes, thyroid disorders, rheumatoid arthritis, MS (in non–first-degree relatives), and systemic lupus erythematosus. Significant differences between cases and controls were also found for the “other” category. Rates of eczema, psoriasis, and bowel disorders did not differ significantly. With adjusted analyses (adjusted for age, sex, race, ethnicity, and biological mother's highest level of education), the OR of any family history of autoimmune disease was 2.27 among cases compared with controls (95% CI, 1.71–3.01, p < 0.001) (Table 3). With adjusted analyses, the OR of any family history of MS (in non–first-degree relatives) was 4.16 among cases compared with controls (95% CI, 2.57–6.75, p < 0.001) (Table 3). Furthermore, with the adjusted analyses, the other category of autoimmune diseases lost statistical significance while inflammatory bowel diseases were more prevalent in families of patients with pediatric MS (p = 0.024).

Table 2

Unadjusted Analyses of Reported Family History of Autoimmune Disease

Table 3

Adjusted Analyses of Reported Family History of Autoimmune Disease

Separate analyses were performed to understand the association between family history of autoimmune diseases and pediatric MS based on which relatives were reported as having autoimmune diseases. The analysis was performed for several categories of relatives: all first-degree relatives, just parents, second-degree relatives, maternal relatives, and paternal relatives. History of MS in pediatric MS patient's mothers was excluded from the maternal relative analysis for MS, and history of MS in patient's fathers was excluded from the paternal relative analysis for MS, as controls could not have a history of MS in either parent. Adjusted analyses revealed an OR of 3.47 of MS among second-degree relatives of patients with pediatric MS. The OR was 2.64 when restricted to maternal relatives and 6.37 when restricted to paternal relatives (Table 4). Adjusted analyses revealed that there was a statistically significant difference between cases and controls in prevalence of adult-onset diabetes, childhood-onset diabetes, and MS in both maternal and paternal secondary relatives. Rheumatoid arthritis and inflammatory bowel disease had a higher prevalence in paternal second-degree relatives, but not maternal. In contrast, thyroid disorders had a higher prevalence in maternal second-degree relatives, but not paternal.

Table 4

Analyses of Reported History of Autoimmune Disease Within Various Relative Groups

Discussion

We sought to identify the prevalence of familial autoimmune disorders in pediatric MS compared with controls as most studies have focused on adult MS.[12] As children who develop MS 20–30 years earlier than average age at onset likely have larger genetic burden and/or environmental exposures than adults with the disease, the prevalence of autoimmune disorders in their family may be higher than reported for adults with MS. Studies of a familial history of autoimmune disease in patients with adult MS have yielded conflicting results.[14,15]

We have identified a number of autoimmune conditions with increased prevalence among the family members of patients with pediatric MS. Of note, 13.7% of patients with pediatric MS reported a familial history of MS compared with only 4.1% of controls (for whom parents could not have the disease as specified by study exclusion criteria). In another study, 25.5% of patients with adult MS reported a family history of MS compared with only 4.1% of controls.[14] There were significant differences between cases and controls relative to the frequency of autoimmune diseases (e.g., diabetes and rheumatoid arthritis), suggesting some shared risk factors for these conditions. Indeed, prior work has suggested shared genetic variants across several autoimmune disorders.[15]

Notably, not all autoimmune conditions followed this pattern. There was no increase in the odds of psoriasis and eczema among family members of pediatric MS cases compared with controls.

Presumably, the link between various autoimmune diseases might be genetic and/or environmental exposures that trigger an aberrant immune response. Although various autoimmune diseases differ in their target organs and antigens, they share a common loss of self-tolerance. When analyzing the rate of familial autoimmune diseases, especially among non–first-degree relatives, an increased risk of autoimmune conditions would likely be caused by shared genetic risk factors, rather than environmental exposures. Understanding families with increased rates of autoimmune conditions could provide an opportunity for dissecting out the universal genetic risks for autoimmunity separate from the genetic risk factors associated with single conditions. This study found that there was a higher rate of MS in second-degree relatives among the patients with pediatric MS compared with controls, and this difference was most pronounced when the history of MS was paternal. Risk of various autoimmune diseases based on parent of origin has been described in numerous genetics studies, with some conditions having maternal associations (e.g., Crohn and juvenile idiopathic arthritis) and others having paternal associations (e.g., psoriatic arthritis).[16,17]

Strengths of the study include the prospective collection of data in multiple centers in the United States representing various geographical areas, careful case ascertainment, and the number of pediatric patients recruited. In addition, controls were enrolled at the same institutions than cases. Weaknesses include the survey-based format of data collection (i.e., the lack of medical confirmation of autoimmune disorders reported by families) and the potential for a biased sampling of controls since recruitment occurred at academic pediatric hospitals and clinics. There are risks for both under- and over-reporting of family history and the possibility of recall bias among cases. Although recall bias could definitely lead to an overestimation of autoimmune diseases in the case cohort, the fact that several conditions were found to have no statistical difference between cases and control reduces the bias concern.

The higher rate of autoimmune diseases among family members of patients with pediatric MS compared with controls, especially the higher rate of MS in second-degree relatives among the patients with pediatric MS, supports the possibility of shared genetic variants between several autoimmune diseases and MS. The higher rate of specific disorders among paternal relatives raises the question of possible imprinting. The rate of autoimmune disease in this control cohort was higher than the general population, particularly in terms of the percent of participants reporting adult onset diabetes within their family (i.e., 29%). This represents a potential bias to the conclusions that could skew the significance of the findings, but not likely change the overall conclusion. Future studies could be designed to compare the genetic profiles of various pediatric autoimmune disorders in an attempt to determine disease specific genes separate from the genes that dictate risk of losing self-tolerance. Combining cohort and family history diagnostic data with expanded genotyping could help establish relative risk of autoimmunity in families based on genetic predispositions. This would require expanded genotyping programs beyond traditional trio designed genetic studies.