Atopic Dermatitis Is Associated with Dermatitis Herpetiformis and Celiac Disease in Children

Atopic dermatitis (AD) is an eczematous inflammatory skin disease that confers an increased risk for several atopic, somatic, and psychiatric comorbidities (Brunner et al., 2017Brunner P.M. Silverberg J.I. Guttman-Yassky E. Paller A.S. Kabashima K. Amagai M. et al.Increasing comorbidities suggest that atopic dermatitis is a systemic disorder.J Invest Dermatol. 2017; 137: 18-25Abstract Full Text Full Text PDF PubMed Scopus (224) Google Scholar, Kauppi et al., 2019Kauppi S. Jokelainen J. Timonen M. Tasanen K. Huilaja L. Adult patients with atopic eczema have a high burden of psychiatric disease: a Finnish Nationwide Registry study.Acta Derm Venereol. 2019; 99: 647-651Crossref PubMed Scopus (15) Google Scholar). Recent epidemiologic data suggest that patients with AD are at risk for several autoimmune diseases (Narla and Silverberg, 2019Narla S. Silverberg J.I. Association between atopic dermatitis and autoimmune disorders in US adults and children: a cross-sectional study.J Am Acad Dermatol. 2019; 80: 382-389Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar). Celiac disease (CD) is an autoimmune disease in which dietary gluten intake induces enteropathy (Collin et al., 2017Collin P. Salmi T.T. Hervonen K. Kaukinen K. Reunala T. Dermatitis herpetiformis: a cutaneous manifestation of coeliac disease.Ann Med. 2017; 49: 23-31Crossref PubMed Scopus (68) Google Scholar). Dermatitis herpetiformis (DH) is an extraintestinal manifestation of CD presenting with intensely pruritic vesicles on specific skin areas (Collin et al., 2017Collin P. Salmi T.T. Hervonen K. Kaukinen K. Reunala T. Dermatitis herpetiformis: a cutaneous manifestation of coeliac disease.Ann Med. 2017; 49: 23-31Crossref PubMed Scopus (68) Google Scholar). Studies analyzing the risk of CD in children and adolescents with AD have found contradictory results (Narla and Silverberg, 2019Narla S. Silverberg J.I. Association between atopic dermatitis and autoimmune disorders in US adults and children: a cross-sectional study.J Am Acad Dermatol. 2019; 80: 382-389Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar, Ress et al., 2014Ress K. Annus T. Putnik U. Luts K. Uibo R. Uibo O. Celiac disease in children with atopic dermatitis.Pediatr Dermatol. 2014; 31: 483-488Crossref PubMed Scopus (35) Google Scholar). Because the highest reported national prevalence of DH was from a study conducted in Finland (Salmi et al., 2011Salmi T.T. Hervonen K. Kautiainen H. Collin P. Reunala T. Prevalence and incidence of dermatitis herpetiformis: a 40-year prospective study from Finland.Br J Dermatol. 2011; 165: 354-359Crossref PubMed Scopus (103) Google Scholar), we designed the present retrospective nationwide hospital register-based study to examine the risks for CD and DH in children and adolescents with AD. Study and control populations (Table 1), databases used, and statistical analyses are presented in the Supplementary Materials and Methods online. We found significant association between AD and DH (OR = 10.42, 95% confidence interval [CI] = 6.56–16.55) as well as between AD and CD (OR = 2.28, 95% CI = 2.07–2.52) in pediatric AD population compared with control individuals. After adjusting for healthcare utilization, the associations remained similar (Table 2). Despite the high ORs found for the associations presented, it is important to take into account the rarity of DH and CD when interpreting the results because the prevalence difference of DH between the AD and the control population is 0.099% (95% CI = 0.073–0.124). In a small Estonian study of 351 pediatric patients with AD, the risk of CD was four times greater than in the general pediatric population (OR = 4.18, 95% CI = 1.12–15.64) (Ress et al., 2014Ress K. Annus T. Putnik U. Luts K. Uibo R. Uibo O. Celiac disease in children with atopic dermatitis.Pediatr Dermatol. 2014; 31: 483-488Crossref PubMed Scopus (35) Google Scholar). In contrast, in a large cohort study of 9,290 adult and 10,196 pediatric patients with AD in the United States, the risk of CD in children and adolescents with AD was not statistically significant (OR = 2.90, 95% CI = 0.88–9.54) (Narla and Silverberg, 2019Narla S. Silverberg J.I. Association between atopic dermatitis and autoimmune disorders in US adults and children: a cross-sectional study.J Am Acad Dermatol. 2019; 80: 382-389Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar). Because DH is relatively rare, a large study population is needed to fully elucidate its association with AD. Moreover, the relatively high prevalence in Finland of both DH and CD (Salmi et al., 2011Salmi T.T. Hervonen K. Kautiainen H. Collin P. Reunala T. Prevalence and incidence of dermatitis herpetiformis: a 40-year prospective study from Finland.Br J Dermatol. 2011; 165: 354-359Crossref PubMed Scopus (103) Google Scholar) may have impacted our findings. The associations between AD and DH and between AD and CD stratified by sex are shown in Supplementary Tables S1 and S2. We found a similar association between AD and both DH and CD in both sexes among the patients with AD. The mean age at AD diagnosis was 6.3 years in patients with AD and 6.8 years versus 5.8 years in girls versus boys, respectively. No statistically significant difference was seen in age at the onset of DH or CD between the AD and the control groups (Supplementary Table S3).Table 1Subject Characteristics of Cases with AD and ControlsGroupADControlTotal96,060228,642Study population, n (%)64,975 (100.0)1Including those AD cases with at least two recorded diagnoses of AD in the register.228,642 (100.0) Girls31,137 (47.92)108,568 (47.5) Boys33,190 (51.08)120,074 (52.5)Abbreviation: AD, atopic dermatitis.1 Including those AD cases with at least two recorded diagnoses of AD in the register. Open table in a new tab Table 2OR for Diagnoses of DH and CD in Patients with AD Adjusted for Healthcare UtilizationDiseaseGroupTotaln (%)OR (95% CI)Adjusted OR (95% CI)DHAD64,97571 (0.1)1A total 1 of 71 in AD group and 3 of 24 in the control group had the diagnosis for DH only once.10.42 (6.56–16.55)9.80 (6.15–15.62)Control228,64224 (0.01)1A total 1 of 71 in AD group and 3 of 24 in the control group had the diagnosis for DH only once.ReferenceReferenceCDAD64,975648 (1.0)2.28 (2.07–2.52)1.92 (1.73–2.13)Control228,6421,005 (0.4)ReferenceReferenceAbbreviations: AD, atopic dermatitis; CD, celiac disease; CI, confidence interval; DH, dermatitis herpetiformis.1 A total 1 of 71 in AD group and 3 of 24 in the control group had the diagnosis for DH only once. Open table in a new tab Abbreviation: AD, atopic dermatitis. Abbreviations: AD, atopic dermatitis; CD, celiac disease; CI, confidence interval; DH, dermatitis herpetiformis. The pathogenesis of DH begins in the gut where ingested gluten induces immune-mediated enteropathy in genetically susceptible individuals, but it is still unclear why only a small proportion of the population develops DH (Reunala et al., 2018Reunala T. Salmi T.T. Hervonen K. Kaukinen K. Collin P. Dermatitis herpetiformis: a common extraintestinal manifestation of coeliac disease.Nutrients. 2018; 10: 602Crossref Scopus (52) Google Scholar). CD is usually latent and asymptomatic in patients with DH (Collin et al., 2017Collin P. Salmi T.T. Hervonen K. Kaukinen K. Reunala T. Dermatitis herpetiformis: a cutaneous manifestation of coeliac disease.Ann Med. 2017; 49: 23-31Crossref PubMed Scopus (68) Google Scholar), which leads to prolonged inflammation in the gut. This may over time lead to the production of antibodies against transglutaminase (TG)3 through epitope spreading and cross-reactivity between tissue TG2 and TG3 (Kárpáti et al., 2018Kárpáti S. Sárdy M. Németh K. Mayer B. Smyth N. Paulsson M. et al.Transglutaminases in autoimmune and inherited skin diseases: the phenomena of epitope spreading and functional compensation.Exp Dermatol. 2018; 27: 807-814Crossref PubMed Scopus (33) Google Scholar). It is suggested that TG3 autoimmunity in DH leads to the formation of circulating IgA–TG3 immunocomplexes that deposit in the dermis (Kárpáti et al., 2018Kárpáti S. Sárdy M. Németh K. Mayer B. Smyth N. Paulsson M. et al.Transglutaminases in autoimmune and inherited skin diseases: the phenomena of epitope spreading and functional compensation.Exp Dermatol. 2018; 27: 807-814Crossref PubMed Scopus (33) Google Scholar). Scratching of itchy AD lesions leads to tissue injury, which results in the release of intracellular autoantigens and thus, activation of immune mechanisms against autoantigens (Tang et al., 2012Tang T.S. Bieber T. Williams H.C. Does "autoreactivity" play a role in atopic dermatitis?.J Allergy Clin Immunol. 2012; 129: 1209-1215.e2Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar). Considering that our study revealed that patients with AD face a significantly higher risk of DH than that of CD, it is possible that skin damage and inflammation in AD predispose for the onset of DH. Our results are in line with previous studies showing that patients with AD carry a heightened risk for autoimmune diseases (Andersen et al., 2017Andersen Y.M. Egeberg A. Gislason G.H. Skov L. Thyssen J.P. Autoimmune diseases in adults with atopic dermatitis.J Am Acad Dermatol. 2017; 76: 274-280.e1Abstract Full Text Full Text PDF PubMed Scopus (69) Google Scholar, Narla and Silverberg, 2019Narla S. Silverberg J.I. Association between atopic dermatitis and autoimmune disorders in US adults and children: a cross-sectional study.J Am Acad Dermatol. 2019; 80: 382-389Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar). A growing body of evidence suggests that autoreactivity plays a significant role in AD pathogenesis (Tang et al., 2012Tang T.S. Bieber T. Williams H.C. Does "autoreactivity" play a role in atopic dermatitis?.J Allergy Clin Immunol. 2012; 129: 1209-1215.e2Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar). It has also been suggested that atopy may be linked to mucosal damage resulting from CD (Ciacci et al., 2004Ciacci C. Cavallaro R. Iovino P. Sabbatini F. Palumbo A. Amoruso D. et al.Allergy prevalence in adult celiac disease.J Allergy Clin Immunol. 2004; 113: 1199-1203Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar). There is a substantial overlap in the genetic backgrounds of AD and autoimmune diseases. A total of 31 specific genomic regions have been associated with AD. Most of these loci have roles in immune responses and have also been implicated in other immune-mediated diseases (Paternoster et al., 2015Paternoster L. Standl M. Waage J. Baurecht H. Hotze M. Strachan D.P. et al.Multi-ancestry genome-wide association study of 21,000 cases and 95,000 controls identifies new risk loci for atopic dermatitis.Nat Genet. 2015; 47: 1449-1456Crossref PubMed Scopus (329) Google Scholar). The HLA-DQ2 and -DQ8 haplotypes have both been associated with susceptibility to DH and CD (Spurkland et al., 1997Spurkland A. Ingvarsson G. Falk E.S. Knutsen I. Sollid L.M. Thorsby E. Dermatitis herpetiformis and celiac disease are both primarily associated with the HLA-DQ (alpha 1∗0501, beta 1∗02) or the HLA-DQ (alpha 1∗03, beta 1∗0302) heterodimers.Tissue Antigens. 1997; 49: 29-34Crossref PubMed Scopus (125) Google Scholar, Kaukinen et al., 2002Kaukinen K. Partanen J. Mäki M. Collin P. HLA-DQ typing in the diagnosis of celiac disease.Am J Gastroenterol. 2002; 97: 695-699Crossref PubMed Google Scholar, Collin et al., 2017Collin P. Salmi T.T. Hervonen K. Kaukinen K. Reunala T. Dermatitis herpetiformis: a cutaneous manifestation of coeliac disease.Ann Med. 2017; 49: 23-31Crossref PubMed Scopus (68) Google Scholar), but no association was detected between AD and these HLA types (Paternoster et al., 2015Paternoster L. Standl M. Waage J. Baurecht H. Hotze M. Strachan D.P. et al.Multi-ancestry genome-wide association study of 21,000 cases and 95,000 controls identifies new risk loci for atopic dermatitis.Nat Genet. 2015; 47: 1449-1456Crossref PubMed Scopus (329) Google Scholar). However, to the best of our knowledge, HLA genetics in patients with concomitant DH and AD has not yet been studied, so it is unknown whether HLA types have any bearing on the association between AD and DH. A major strength of our study is that we had a large study population of patients with AD as well as a random, representative sample of the Finnish population as a control group. The Finnish hospital register was an optimal source of data for studying the link between AD and DH because of the particularly high prevalence of DH found in Finland (Salmi et al., 2011Salmi T.T. Hervonen K. Kautiainen H. Collin P. Reunala T. Prevalence and incidence of dermatitis herpetiformis: a 40-year prospective study from Finland.Br J Dermatol. 2011; 165: 354-359Crossref PubMed Scopus (103) Google Scholar). A weakness of our study is that the data are limited to the hospital setting and do not contain information about lifestyle factors or detailed clinical information. Because this is a registry-based study, misclassification of DH as AD cannot be excluded. In summary, we report a significant association between AD and DH in children and adolescents. In daily practice, it is important for dermatologists to recognize the risk of DH in patients with AD, especially because the symptoms of DH may be obscured by those of AD, although the overall prevalence of DH and CD are low. Further studies are needed to determine the potential mechanisms underlying the associations of AD with DH and AD with CD. Our data are from the Finnish Care Register for Health Care (former name: the Finnish Hospital Discharge Register) maintained by the National Institute for Health and Welfare. According to the Finnish laws and regulations, the data in the social welfare and health care registers and documents are confidential. The National Institute for Health and Welfare can, on a case-by-case basis, grant permission to use the registers and the documents for purposes of scientific research. More information from research authorization applications can be found on https://www.thl.fi/en/web/thlfi-en/statistics/information-for-researchers. Saana Kauppi: http://orcid.org/0000-0003-3609-8874 Jari Jokelainen: http://orcid.org/0000-0003-4629-0560 Markku Timonen: http://orcid.org/0000-0002-9624-6544 Kaisa Tasanen: http://orcid.org/0000-0001-8056-5949 Laura Huilaja: https://orcid.org/0000-0002-7597-4323 SK has received educational grants from Novartis, Leo Pharma, Sanofi Genzyme, AbbVie, and Janssen-Cilag and honoraria from Sanofi Genzyme for speaking. KT has received educational grants from Novartis and Pfizer and honoraria from Novartis, AbbVie, Janssen-Cilag, Sanofi Genzyme, and Eli Lilly for consulting and speaking. LH has received educational grants from Shire, Janssen-Cilag, Novartis, AbbVie, and LeoPharma and honoraria from Sanofi Genzyme, Novartis, AbbVie, and UCB Pharma for consulting and speaking and is an investigator for AbbVie. The remaining authors state no conflict of interest. Conceptualization: SK, MT, KT, LH; Formal Analysis: SK, JJ, LH; Methodology: SK, JJ, KT, LH; Supervision: MT, KT, LH; Writing - Original Draft Preparation: SK; Writing - Review and Editing: SK, MT, KT, LH. This was a retrospective register-based study. All cases of atopic dermatitis (AD) in Finnish hospitals between 1 January 1987 and 31 December 2016 were eligible for inclusion if the patient was aged <18 years at the time of diagnosis. Patient records were obtained from the Finnish Health Register for Health Care (CRHC), which is maintained by the National Institute for Health and Welfare. The CRHC covers all in-patient visits in Finnish hospitals from 1987 and outpatient visits from 1998 onward. Records were selected by a diagnosis of AD (code 6918B in the World Health Organization International Classification of Diseases [ICD]-9 and code L20.0 in ICD-10). In Finland, ICD-9 was used between 1987 and 1995 and ICD-10 has been used since 1996. As controls, we collected data from a random sample of individuals aged >18 years from the Finnish Population Register Centre. The Population Information System contains basic information about Finnish residents, and every individual is recognized by a personal identity code. AD-diagnosed individuals and individuals with no records in the CRHC were excluded from the control group. Those without records in the CRHC were excluded from the control group to avoid the potential bias of this population sample having undiagnosed diseases or diagnoses not recorded in the register. In both groups, we gathered data from the CRHC on the numbers of diagnoses of dermatitis herpetiformis (codes 6940A and 6942A in ICD-9 and L13.0 in ICD-10) and celiac disease (code 5790 in ICD-9 and code K90.0 in ICD-10) recorded before the age of 18 years. The characteristics of the study population are presented as proportions and means. The associations between the groups were evaluated using a logistic regression model and presented as ORs and 95% confidence intervals. All statistical analyses were performed using the SAS software package (version 9.4, SAS Institute, Cary, NC) and two-sided P-values < 0.05 were considered statistically significant. The search of the CRHC database returned data for 151,822 individuals who received a diagnosis of AD between 1987 and 2016 and 96,060 patients who were under the age of 18 years at the time of diagnosis. The final study population of 64,975 AD cases was formed by including only those AD cases with at least two recorded diagnoses of AD in the CRHC to increase validity of the AD diagnosis. Of the 500,000 population register sample, 228,642 individuals who had at least one diagnosis registered in the CRHC but no diagnosis of AD were identified as controls. The characteristics of the AD and control groups are presented in Table 1.Supplementary Table S1OR for the Diagnoses of DH and CD in Patients with AD Stratified By SexDiseaseGroupTotalGirlsBoysn (%)OR (95% CI)n (%)OR (95% CI)DHAD64,97532 (0.1)8.48 (4.45–16.16)39 (0.1)12.74 (6.53–24.88)Control228,64213 (0.0)Reference11 (0.0)ReferenceCDAD64,975394 (0.6)2.21 (1.95–2.51)254 (0.4)2.36 (2.01–2.77)Control228,642617 (0.6)Reference388 (0.3)ReferenceAbbreviations: AD, atopic dermatitis; CD, celiac disease; CI, confidence interval; DH, dermatitis herpetiformis. Open table in a new tab Supplementary Table S2OR for the Diagnoses of DH and CD in Patients with AD Adjusted for Healthcare Utilization Stratified by SexDiseaseGroupTotalGirlsBoysn (%)Adjusted OR (95% CI)n (%)Adjusted OR (95% CI)DHAD64,97532 (0.1)8.63 (4.77–15.64)39 (0.1)11.26 (5.95–21.29)Control228,64213 (0.0)Reference11 (0.0)ReferenceCDAD64,975394 (0.6)1.51 (1.35–1.70)254 (0.4)2.75 (2.42–3.13)Control228,642617 (0.6)Reference388 (0.3)ReferenceAbbreviations: AD, atopic dermatitis; CD, celiac disease; CI, confidence interval; DH, dermatitis herpetiformis. Open table in a new tab Supplementary Table S3Age at Diagnosis of AD, DH, and CD in Cases with AD and ControlsDiseaseGroupAge at diagnosis, Mean (SD)ADAD6.3 (4.5)ControlNADHAD9.6 (3.9)Control7.8 (4.1)CDAD8.2 (4.1)Control7.9 (4.0)Abbreviations: AD, atopic dermatitis; CD, celiac disease; DH, dermatitis herpetiformis; NA, not applicable. Open table in a new tab Abbreviations: AD, atopic dermatitis; CD, celiac disease; CI, confidence interval; DH, dermatitis herpetiformis. Abbreviations: AD, atopic dermatitis; CD, celiac disease; CI, confidence interval; DH, dermatitis herpetiformis. Abbreviations: AD, atopic dermatitis; CD, celiac disease; DH, dermatitis herpetiformis; NA, not applicable.