The prevalence of celiac disease (CD) in children with type 1 diabetes (T1D) is 5.1%, and it is often asymptomatic (1). Thus, children with T1D are routinely screened for CD with assessment of IgA antibodies against tissue transglutaminase (TGA-IgA). However, elevated TGA-IgA at T1D onset can normalize spontaneously without a gluten-free diet (2,3). The cause of this remains unknown, but genetic variations may influence CD outcomes (2). The European Society of Paediatric Gastroenterology Hepatology and Nutrition recommends a nonbiopsy approach for diagnosis of CD in children with TGA-IgA values >10 times the upper limit of normal (>10×ULN) who have a positive result on an endomysial IgA test in a second blood sample (4). An endomysial IgA test is expensive, with limited availability in some regions of the world, which raised the question of CD diagnosis based on TGA-IgA alone. Therefore, we performed a retrospective analysis to explore the diagnostic outcome of children with newly diagnosed T1D with elevated TGA-IgA in an Australian center where duodenal biopsy is pursued for diagnosing CD.
Data were collected from electronic medical records of children <17 years of age diagnosed with new-onset T1D at John Hunter Children’s Hospital between 2012 and 2022. Children with prior diagnosis of CD or low IgA test result were excluded (ethics approval: Hunter New England Local Health District Research Ethics Committee, 16/07/20/4.05). Information was recorded on TGA-IgA levels, at diagnosis of T1D and at ∼6- to 12-month intervals, and serum ferritin, iron, and vitamin B12 levels; HbA1c; sex; comorbidities; diabetic ketoacidosis; family history of autoimmune conditions; and anthropometry (height, weight, and BMI z scores). Stage 3 T1D was diagnosed according to elevated random glucose >11.1 mmol/L, HbA1c >6.5%, and positivity for one or more islet antibodies (5). Due to varying assays, TGA-IgA elevation was defined as any level >1×ULN. Children whose serial TGA-IgA titers were not declining were referred for duodenal biopsy. All maintained a gluten-containing diet (GCD), with guidance from a specialist dietitian. CD was disproven by negative biopsy or normalizing (<1×ULN) of TGA-IgA while a GCD was maintained. After biopsies were performed, children were divided into two groups (CD-positive vs. CD-negative). Statistical analysis was undertaken with unpaired Student t tests for normally distributed variables and χ2 test for categorical variables using Stata, version 18 (StataCorp, College Station, TX). P values <0.05 were considered statistically significant.
A total of 358 children with new-onset T1D with no prior history of CD were included in the study; 48 (13.4%) had elevated TGA-IgA on diagnosis, and 24 (6.7%) had TGA-IgA >10×ULN. Table 1 details demographics and clinical features by CD status, and laboratory parameters including TGA-IgA levels. There was no difference between the groups in age at diagnosis of T1D, sex, autoimmune comorbidities, or family history of autoimmune disease in first-degree relatives. Of the anthropological measurements at diagnosis, weight z score was significantly lower for CD-positive patients (P = 0.04).
Demographic and clinical characteristics and TGA-IgA levels of children with newly diagnosed T1D and elevated TGA-IgA with CD proven or disproven
. | CD-positive, n = 22 . | CD-negative, n = 20 . | P . |
---|---|---|---|
Age, years | 9.5 (3.4) | 9.5 (3.7) | 0.990 |
Female, n (%) | 16 (72) | 12 (60) | 0.380 |
Male, n (%) | 6 (27) | 8 (40) | |
Comorbidities, n (%)* | 2 (9) | 2 (10) | 0.920 |
Family history, n (%)** | 14 (63) | 12 (50) | 0.810 |
DKA, n (%) | 8 (37) | 9 (45) | 0.570 |
HbA1c, mmol/mol | 93 (13) | 110 (9) | 0.120 |
Weight z score | 0.02 (0.99) | 0.65 (0.92) | 0.040 |
Height z score | 0.06 (0.75) | 0.46 (1.07) | 0.180 |
BMI z score | −0.04 (1.24) | 0.47 (1.17) | 0.190 |
TGA-IgA 1.0–4.9×ULN, n (%) | 4 (18.0) | 12 (60.0) | 0.005 |
TGA-IgA 5.0–9.9×ULN, n (%) | 1 (4.5) | 1 (5.0) | 0.900 |
TGA-IgA >10.0×ULN, n (%) | 17 (77.3) | 7 (35.0) | 0.006 |
Serum iron, μmol/L | 13.6 (6.2) | 11.7 (7.1) | 0.350 |
Low ferritin for age, n (%) | 3 (13.6) | 2 (10) | 0.600 |
Low transferrin sat for age, n (%) | 1 (4.5) | 4 (20) | 0.100 |
Serum B12, pmol/L | 491 (158) | 520 (261) | 0.660 |
. | CD-positive, n = 22 . | CD-negative, n = 20 . | P . |
---|---|---|---|
Age, years | 9.5 (3.4) | 9.5 (3.7) | 0.990 |
Female, n (%) | 16 (72) | 12 (60) | 0.380 |
Male, n (%) | 6 (27) | 8 (40) | |
Comorbidities, n (%)* | 2 (9) | 2 (10) | 0.920 |
Family history, n (%)** | 14 (63) | 12 (50) | 0.810 |
DKA, n (%) | 8 (37) | 9 (45) | 0.570 |
HbA1c, mmol/mol | 93 (13) | 110 (9) | 0.120 |
Weight z score | 0.02 (0.99) | 0.65 (0.92) | 0.040 |
Height z score | 0.06 (0.75) | 0.46 (1.07) | 0.180 |
BMI z score | −0.04 (1.24) | 0.47 (1.17) | 0.190 |
TGA-IgA 1.0–4.9×ULN, n (%) | 4 (18.0) | 12 (60.0) | 0.005 |
TGA-IgA 5.0–9.9×ULN, n (%) | 1 (4.5) | 1 (5.0) | 0.900 |
TGA-IgA >10.0×ULN, n (%) | 17 (77.3) | 7 (35.0) | 0.006 |
Serum iron, μmol/L | 13.6 (6.2) | 11.7 (7.1) | 0.350 |
Low ferritin for age, n (%) | 3 (13.6) | 2 (10) | 0.600 |
Low transferrin sat for age, n (%) | 1 (4.5) | 4 (20) | 0.100 |
Serum B12, pmol/L | 491 (158) | 520 (261) | 0.660 |
Data are means (SD) unless otherwise indicated. DKA, diabetic ketoacidosis; sat, saturation.
*Thyroid disease, Addison disease, Down syndrome, juvenile arthritis, systemic lupus erythematosus.
**Thyroid disease, CD, T1D, systemic lupus erythematosus.
CD was excluded for 20 of 48 (42%) patients with elevated TGA-IgA at diagnosis according to either normalization (n = 14) or negative biopsy (n = 6) while on a GCD. Among the 34 with a repeat elevated TGA-IgA test result, CD was confirmed for 22 (64%) by duodenal biopsy. Of the 24 children who had levels >10×ULN at diagnosis of T1D, 17 (70.8%) had biopsy-proven CD, 7 (29.1%) had CD disproven, and 1 received a serological diagnosis of likely CD due to severe symptoms and family declining duodenal biopsy.
Six children (12.5%) had repeat elevated serology but did not undergo biopsy, TGA-IgA subsequently normalized in one (2%) child while on a GCD, and TGA-IgA has not normalized in five (10%) children, who are currently awaiting endoscopy.
In grouping the children according to TGA-IgA levels (Table 1), those with milder elevations (1.1–4.9×ULN) were less likely to subsequently have proven CD (P = 0.005) and those with initial titers >10×ULN were more likely to have CD (P = 0.006). Among the children with levels >10×ULN, 7 of 24 (29.1%) had CD disproven. Among the group with TGA-IgA 1.0–4.9×ULN, 4 of 16 (25%) were CD-positive. In comparisons of TGA-IgA titers and baseline clinical parameters with use of multivariate logistic regression, only TGA-IgA >10×ULN remained significantly associated with a confirmed CD diagnosis.
Our study has several limitations. The small, single-center population may limit the generalizability of the findings. Specifically, this analysis is not designed to estimate an accurate prevalence of CD in T1D. While some children may have started a gluten-free diet without our recommendation, this is unlikely, since all families received specialized dietetic support to maintain a GCD prior to endoscopy.
Elevated TGA-IgA levels are common at diagnosis of T1D and do not consistently predict CD. Our data support repeating serology and pursuing biopsy in those with persistent elevation of TGA-IgA to avoid overdiagnosis of CD in children with newly diagnosed T1D.
Article Information
Acknowledgments. The authors thank the children/adolescents and families who participated in this study and the Paediatric Endocrinology team at the John Hunter Children’s Hospital for clinical care and maintenance of the endocrine clinical database.
Funding. D.E. received an International Society for Pediatric and Adolescent Diabetes (ISPAD) travel grant.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. D.E. was involved in the acquisition, analysis, and interpretation of data and wrote the first draft of the manuscript, and all authors edited, reviewed, and approved the final version of the manuscript. C.E.S. was involved in development of the study concept and design, supervision of data acquisition, and critical review and editing of the manuscript. K.C. assisted with data acquisition and review of the manuscript. B.R.K. was involved in development of the study concept, contributed to study design and data interpretation, and critically reviewed the manuscript. S.N. was involved in development of the study concept and design, performed statistical analysis, and critically review and editing of the manuscript. S.N. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. Parts of this study were presented in abstract form at the 49th Annual ISPAD Conference, Rotterdam, the Netherlands, 18–21 October 2023.
Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Elizabeth Selvin and Matthew J. Crowley.