Association of Gluten Intake During the First 5 Years of Life With Incidence of Celiac Disease Autoimmunity and Celiac Disease Among Children at Increased Risk

Carin Andrén Aronsson; Hye-Seung Lee; Elin M Hård af Segerstad; Ulla Uusitalo; Jimin Yang; Sibylle Koletzko; Edwin Liu; Kalle Kurppa; Polly J Bingley; Jorma Toppari; Anette G Ziegler; Jin-Xiong She; William A Hagopian; Marian Rewers; Beena Akolkar; Jeffrey P Krischer; Suvi M Virtanen; Jill M Norris; Daniel Agardh

doi:10.1001/jama.2019.10329

. 2019 Aug 13;322(6):514–523. doi: 10.1001/jama.2019.10329

Association of Gluten Intake During the First 5 Years of Life With Incidence of Celiac Disease Autoimmunity and Celiac Disease Among Children at Increased Risk

Carin Andrén Aronsson ¹, Hye-Seung Lee ², Elin M Hård af Segerstad ¹, Ulla Uusitalo ², Jimin Yang ², Sibylle Koletzko ^3,⁴, Edwin Liu ⁵, Kalle Kurppa ⁶, Polly J Bingley ⁷, Jorma Toppari ^8,⁹, Anette G Ziegler ¹⁰, Jin-Xiong She ¹¹, William A Hagopian ¹², Marian Rewers ¹³, Beena Akolkar ¹⁴, Jeffrey P Krischer ², Suvi M Virtanen ^15,^16,¹⁷, Jill M Norris ¹⁸, Daniel Agardh ^1,^✉, for the TEDDY Study Group

¹Department of Clinical Sciences, Lund University, Malmö, Sweden

²Health Informatics Institute, Department of Pediatrics, Morsani College of Medicine, University of South Florida, Tampa

³Dr von Hauner Children’s Hospital, Ludwig Maximilians University, Munich, Germany

⁴University of Warmia and Mazuri, Olsztyn, Poland

⁵Digestive Health Institute, University of Colorado Denver, Children’s Hospital Colorado, Denver

⁶Tampere Centre for Child Health Research, University of Tampere, Tampere University Hospital, Tampere, Finland

⁷School of Clinical Sciences, University of Bristol, Bristol, England

⁸Research Centre for Integrative Physiology and Pharmacology, Institute of Biomedicine, University of Turku, Turku, Finland

⁹Department of Pediatrics, Turku University Hospital, Turku, Finland

¹⁰Institute of Diabetes Research, Helmholtz Zentrum München, Klinikum rechts der Isar, Technische Universität München, and Forschergruppe Diabetes eV, Neuherberg, Germany

¹¹Center for Biotechnology and Genomic Medicine, Augusta University, Augusta, Georgia

¹²Pacific Northwest Diabetes Research Institute, Seattle, Washington

¹³Barbara Davis Center for Childhood Diabetes, University of Colorado School of Medicine, Aurora

¹⁴National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, Maryland

¹⁵National Institute for Health and Welfare, Department of Public Health Solutions, Helsinki, Finland

¹⁶Faculty of Social Sciences/Health Sciences, University of Tampere, Tampere, Finland

¹⁷Research Center for Child Health, Tampere University, University Hospital, Science Center of Pirkanmaa Hospital District, Tampere, Finland

¹⁸Colorado School of Public Health, Department of Epidemiology, University of Colorado, Aurora

^✉

Corresponding Author: Daniel Agardh, MD, PhD, Clinical Research Centre, Department of Clinical Sciences, Diabetes and Celiac Disease Unit, Lund University, Jan Waldenströms Gata 35, 20502 Malmö, Sweden (daniel.agardh@med.lu.se).

Group Information: The TEDDY Study Group members appear at the end of the article.

Author Contributions: Dr Lee had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Lee, Liu, Kurppa, Ziegler, She, Hagopian, Rewers, Akolkar, Virtanen, Agardh.

Acquisition, analysis, or interpretation of data: Andrén Aronsson, Lee, Hård af Segerstad, Uusitalo, Yang, Koletzko, Kurppa, Bingley, Toppari, Ziegler, She, Hagopian, Rewers, Krischer, Virtanen, Norris, Agardh.

Drafting of the manuscript: Andrén Aronsson, Lee, Hård af Segerstad, Bingley, Agardh.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Andrén Aronsson, Lee, Yang, Krischer, Virtanen.

Obtained funding: Toppari, Ziegler, She, Hagopian, Rewers.

Administrative, technical, or material support: Yang, Koletzko, Kurppa, Toppari, Hagopian, Rewers, Krischer, Norris.

Supervision: Yang, Kurppa, Toppari, Ziegler, She, Hagopian, Rewers, Akolkar, Virtanen, Norris, Agardh.

Conflict of Interest Disclosures: Dr Koletzko reported being a member of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition guideline group for celiac disease and a member of the PreventCD consortium. No other disclosures were reported.

Funding/Support: This study was funded by grants U01 DK63829, U01 DK63861, U01 DK63821, U01 DK63865, U01 DK63863, U01 DK63836, U01 DK63790, UC4 DK63829, UC4 DK63861, UC4 DK63821, UC4 DK63865, UC4 DK63863, UC4 DK63836, UC4 DK95300, UC4 DK100238, UC4 DK106955, UC4 DK112243, and UC4 DK117483 and contract HHSN267200700014C from the National Institute of Diabetes and Digestive and Kidney Diseases, the National Institute of Allergy and Infectious Diseases, the National Institute of Child Health and Human Development, the National Institute of Environmental Health Sciences, the Centers for Disease Control and Prevention, and JDRF. This work was supported in part by the clinical and translational science awards given to the University of Florida (UL1 TR000064) and the University of Colorado (UL1 TR001082) from the National Institutes of Health and the National Center for Advancing Translational Sciences.

Role of the Funder/Sponsor: The sponsors were represented in The Environmental Determinants of Diabetes in the Young (TEDDY) steering committee. The sponsors had a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and review and approval to submit the manuscript for publication. The sponsors did not have the right to veto submission to any particular journal, but did participate in the writing group’s discussion when selecting an appropriate journal for submission. The corresponding author had the final say in submitting the manuscript for publication.

Group Information: The TEDDY Study Group clinical sites and members are located in Finland: Jorma Toppari, MD, PhD, primary investigator (University of Turku, Turku University Hospital, Hospital District of Southwest Finland), Olli G. Simell, MD, PhD (University of Turku), Annika Adamsson, PhD (Turku University Hospital, Hospital District of Southwest Finland), Suvi Ahonen (University of Tampere, Tampere University Hospital, and National Institute for Health and Welfare, Finland), Mari Åkerlund (University of Tampere, Tampere University Hospital, and National Institute for Health and Welfare, Finland), Anne Hekkala, MD (University of Oulu, Oulu University Hospital), Henna Holappa (University of Oulu, Oulu University Hospital), Heikki Hyöty, MD, PhD (University of Tampere, Tampere University Hospital), Anni Ikonen (University of Oulu, Oulu University Hospital), Jorma Ilonen, MD, PhD (University of Turku and University of Kuopio), Sinikka Jäminki (University of Tampere, Tampere University Hospital), Sanna Jokipuu (Turku University Hospital, Hospital District of Southwest Finland), Leena Karlsson (Turku University Hospital, Hospital District of Southwest Finland), Miia Kähönen (University of Oulu, Oulu University Hospital), Mikael Knip, MD, PhD (University of Tampere, Tampere University Hospital), Minna-Liisa Koivikko (University of Oulu, Oulu University Hospital), Mirva Koreasalo (University of Tampere, Tampere University Hospital, and National Institute for Health and Welfare, Finland), Kalle Kurppa, MD, PhD (University of Tampere, Tampere University Hospital), Jarita Kytölä (University of Tampere, Tampere University Hospital), Tiina Latva-aho (University of Oulu, Oulu University Hospital), Katri Lindfors, PhD (University of Tampere), Maria Lönnrot, MD, PhD (University of Tampere, Tampere University Hospital), Elina Mäntymäki (Turku University Hospital, Hospital District of Southwest Finland), Markus Mattila (University of Tampere), Katja Multasuo (University of Oulu, Oulu University Hospital), Teija Mykkänen (University of Oulu, Oulu University Hospital), Tiina Niininen (Tampere University Hospital, University of Tampere), Sari Niinistö (Tampere University Hospital and National Institute for Health and Welfare, Finland), Mia Nyblom (University of Tampere, Tampere University Hospital), Sami Oikarinen, PhD (University of Tampere, Tampere University Hospital), Paula Ollikainen (University of Oulu, Oulu University Hospital), Sirpa Pohjola (University of Oulu, Oulu University Hospital), Petra Rajala (Turku University Hospital, Hospital District of Southwest Finland), Jenna Rautanen (Tampere University Hospital and National Institute for Health and Welfare, Finland), Anne Riikonen (University of Tampere, Tampere University Hospital, and National Institute for Health and Welfare, Finland), Minna Romo (Turku University Hospital, Hospital District of Southwest Finland), Suvi Ruohonen (Turku University Hospital, Hospital District of Southwest Finland), Satu Simell, MD, PhD (University of Turku), Maija Sjöberg (University of Turku, Turku University Hospital, Hospital District of Southwest Finland), Aino Stenius (University of Oulu, Oulu University Hospital), Päivi Tossavainen, MD (University of Oulu, Oulu University Hospital), Mari Vähä-Mäkilä (Turku University Hospital, Hospital District of Southwest Finland), Sini Vainionpää (Turku University Hospital, Hospital District of Southwest Finland), Eeva Varjonen (University of Turku, Turku University Hospital, Hospital District of Southwest Finland), Riitta Veijola, MD, PhD (University of Oulu, Oulu University Hospital), Irene Viinikangas (University of Oulu, Oulu University Hospital), and Suvi M. Virtanen, MD, PhD (University of Tampere, Tampere University Hospital, National Institute for Health and Welfare, Finland); Germany: Anette G. Ziegler, MD, primary investigator (all investigators affiliated with Forschergruppe Diabetes eV, Institute of Diabetes Research, Helmholtz Zentrum München, Klinikum rechts der Isar, Technische Universität München, and Forschergruppe Diabetes eV, Neuherberg, unless otherwise noted), Ezio Bonifacio, PhD (Center for Regenerative Therapies, TU Dresden), Miryam D'Angelo, Anita Gavrisan, Cigdem Gezginci, Anja Heublein, Verena Hoffmann, PhD, Sandra Hummel, PhD, Andrea Keimer (University of Bonn, Department of Nutritional Epidemiology), Annette Knopff, Charlotte Koch, Sibylle Koletzko, MD (Dr von Hauner Children’s Hospital, Department of Gastroenterology, Ludwig Maximillians University Munich), Claudia Ramminger, Roswith Roth, PhD, Marlon Scholz, Joanna Stock, Katharina Warncke, MD, Lorena Wendel, and Christiane Winkler, PhD; Sweden: Åke Lernmark, PhD, primary investigator (all investigators affiliated with Lund University), Daniel Agardh, MD, PhD, Carin Andrén Aronsson, PhD, Maria Ask, Jenny Bremer, Corrado Cilio, PhD, MD, Emelie Ericson-Hallström, Annika Fors, Lina Fransson, Thomas Gard, Rasmus Bennet, Monika Hansen, Hanna Jisser, Fredrik Johansen, Berglind Jonsdottir, MD, PhD, Silvija Jovic, Helena Elding Larsson, MD, PhD, Marielle Lindström, Markus Lundgren, MD, PhD, Maria Månsson-Martinez, Maria Markan, Jessica Melin, Zeliha Mestan, Caroline Nilsson, Karin Ottosson, Kobra Rahmati, Anita Ramelius, Falastin Salami, Anette Sjöberg, Birgitta Sjöberg, Carina Törn, PhD, Anne Wallin, Åsa Wimar, and Sofie Åberg; United States: Colorado: Marian Rewers, MD, PhD, primary investigator (all investigators affiliated with the University of Colorado, Anschutz Medical Campus, Barbara Davis Center for Childhood Diabetes), Kimberly Bautista, Judith Baxter, Daniel Felipe-Morales, Kimberly Driscoll, PhD, Brigitte I. Frohnert, MD, Marisa Gallant, MD, Patricia Gesualdo, Michelle Hoffman, Rachel Karban, Edwin Liu, MD, Jill Norris, PhD, Andrea Steck, MD, and Kathleen Waugh; Georgia and Florida: Jin-Xiong She, PhD, primary investigator (Center for Biotechnology and Genomic Medicine, Augusta University), Desmond Schatz, MD (University of Florida), Diane Hopkins (Center for Biotechnology and Genomic Medicine, Augusta University), Leigh Steed (Center for Biotechnology and Genomic Medicine, Augusta University), Jennifer Bryant (Center for Biotechnology and Genomic Medicine, Augusta University), Katherine Silvis (Center for Biotechnology and Genomic Medicine, Augusta University), Michael Haller, MD (University of Florida), Melissa Gardiner (Center for Biotechnology and Genomic Medicine, Augusta University), Richard McIndoe, PhD (Center for Biotechnology and Genomic Medicine, Augusta University), Ashok Sharma (Center for Biotechnology and Genomic Medicine, Augusta University), Stephen W. Anderson, MD (Pediatric Endocrine Associates, Atlanta), Laura Jacobsen, MD (University of Florida), John Marks, DHSc (University of Florida), and P. D. Towe (University of Florida); Pennsylvania Satellite Center: Dorothy Becker, MD (all investigators affiliated with Children’s Hospital of Pittsburgh of UPMC), Margaret Franciscus, MaryEllen Dalmagro-Elias Smith, Ashi Daftary, MD, Mary Beth Klein, and Chrystal Yates; Washington: William A. Hagopian, MD, PhD, primary investigator (all investigators affiliated with Pacific Northwest Research Institute), Michael Killian, Claire Cowen Crouch, Jennifer Skidmore, Ashley Akramoff, Masumeh Chavoshi, Kayleen Dunson, Rachel Hervey, Rachel Lyons, Arlene Meyer, Denise Mulenga, Jared Radtke, Matei Romancik, Davey Schmitt, Julie Schwabe, and Sarah Zink; Data Coordinating Center: Jeffrey P. Krischer, PhD, primary investigator (all investigators affiliated with the University of South Florida unless otherwise noted), Sarah Austin-Gonzalez, Maryouri Avendano, Sandra Baethke, Rasheedah Brown, Brant Burkhardt, PhD, Martha Butterworth, Joanna Clasen, David Cuthbertson, Christopher Eberhard, Steven Fiske, Jennifer Garmeson, Veena Gowda, Kathleen Heyman, Belinda Hsiao, Christina Karges, Francisco Perez Laras, Hye-Seung Lee, PhD, Qian Li, Shu Liu, Xiang Liu, PhD, Kristian Lynch, PhD, Colleen Maguire, Jamie Malloy, Cristina McCarthy, Aubrie Merrell, Steven Meulemans, Hemang Parikh, PhD, Ryan Quigley, Cassandra Remedios, Chris Shaffer, Laura Smith, PhD, Susan Smith, Noah Sulman, PhD, Roy Tamura, PhD, Dena Tewey, Michael Toth, Ulla Uusitalo, PhD, Kendra Vehik, PhD, Ponni Vijayakandipan, Keith Wood, and Jimin Yang, PhD, RD (former staff members: Michael Abbondondolo, Lori Ballard, David Hadley, PhD, and Wendy McLeod); Autoantibody Reference Laboratories: Liping Yu, MD (Barbara Davis Center for Childhood Diabetes, University of Colorado Denver), Dongmei Miao, MD (Barbara Davis Center for Childhood Diabetes, University of Colorado Denver), Polly Bingley, MD, FRCP (Bristol Medical School, University of Bristol UK), Alistair Williams (Bristol Medical School, University of Bristol UK), Kyla Chandler (Bristol Medical School, University of Bristol UK), Olivia Ball (Bristol Medical School, University of Bristol UK), Ilana Kelland (Bristol Medical School, University of Bristol UK), Sian Grace (Bristol Medical School, University of Bristol UK), and Ben Gillard (Bristol Medical School, University of Bristol UK); HLA Reference Laboratory: William Hagopian, MD, PhD (all investigators affiliated with Pacific Northwest Research Institute unless otherwise noted), Masumeh Chavoshi, Jared Radtke, and Julie Schwabe (former staff members: Henry Erlich, PhD, Steven J. Mack, PhD, and Anna Lisa Fear); Repository: Sandra Ke and Niveen Mulholland, PhD (both with the National Institutes of Diabetes and Digestive and Kidney Diseases biosample repository at Fisher BioServices); Project Scientist: Beena Akolkar, PhD (National Institutes of Diabetes and Digestive and Kidney Diseases); and Other Contributors: Kasia Bourcier, PhD (National Institutes of Allergy and Infectious Diseases), Thomas Briese, PhD (Columbia University), Suzanne Bennett Johnson, PhD (Florida State University), and Eric Triplett, PhD (University of Florida).

^✉

Corresponding author.

PMCID: PMC6692672 PMID: 31408136

Key Points

Question

Is the amount of gluten intake during the first 5 years of life associated with the risk of celiac disease autoimmunity and celiac disease in at-risk children?

Findings

In this multinational prospective birth cohort consisting of 6605 genetically predisposed children, higher gluten intake was associated with a statistically significant increase in celiac disease autoimmunity (absolute risk difference, 6.1%) and celiac disease (absolute risk difference, 7.2%) for every gram increase of gluten intake per day.

Meaning

Increased intake of gluten during the first 5 years of life was an independent risk factor for celiac disease autoimmunity and celiac disease in genetically predisposed children.

Abstract

Importance

High gluten intake during childhood may confer risk of celiac disease.

Objectives

To investigate if the amount of gluten intake is associated with celiac disease autoimmunity and celiac disease in genetically at-risk children.

Design, Setting, and Participants

The participants in The Environmental Determinants of Diabetes in the Young (TEDDY), a prospective observational birth cohort study designed to identify environmental triggers of type 1 diabetes and celiac disease, were followed up at 6 clinical centers in Finland, Germany, Sweden, and the United States. Between 2004 and 2010, 8676 newborns carrying HLA antigen genotypes associated with type 1 diabetes and celiac disease were enrolled. Screening for celiac disease with tissue transglutaminase autoantibodies was performed annually in 6757 children from the age of 2 years. Data on gluten intake were available in 6605 children (98%) by September 30, 2017.

Exposures

Gluten intake was estimated from 3-day food records collected at ages 6, 9, and 12 months and biannually thereafter until the age of 5 years.

Main Outcomes and Measures

The primary outcome was celiac disease autoimmunity, defined as positive tissue transglutaminase autoantibodies found in 2 consecutive serum samples. The secondary outcome was celiac disease confirmed by intestinal biopsy or persistently high tissue transglutaminase autoantibody levels.

Results

Of the 6605 children (49% females; median follow-up: 9.0 years [interquartile range, 8.0-10.0 years]), 1216 (18%) developed celiac disease autoimmunity and 447 (7%) developed celiac disease. The incidence for both outcomes peaked at the age of 2 to 3 years. Daily gluten intake was associated with higher risk of celiac disease autoimmunity for every 1-g/d increase in gluten consumption (hazard ratio [HR], 1.30 [95% CI, 1.22-1.38]; absolute risk by the age of 3 years if the reference amount of gluten was consumed, 28.1%; absolute risk if gluten intake was 1-g/d higher than the reference amount, 34.2%; absolute risk difference, 6.1% [95% CI, 4.5%-7.7%]). Daily gluten intake was associated with higher risk of celiac disease for every 1-g/d increase in gluten consumption (HR, 1.50 [95% CI, 1.35-1.66]; absolute risk by age of 3 years if the reference amount of gluten was consumed, 20.7%; absolute risk if gluten intake was 1-g/d higher than the reference amount, 27.9%; absolute risk difference, 7.2% [95% CI, 6.1%-8.3%]).

Conclusions and Relevance

Higher gluten intake during the first 5 years of life was associated with increased risk of celiac disease autoimmunity and celiac disease among genetically predisposed children.

This cohort study investigates the association between gluten intake in the first 5 years of life and celiac disease autoimmunity and biopsy-confirmed celiac disease in genetically at-risk children over a median 9-year follow-up.

Introduction

Gluten is a food antigen found in wheat, rye, and barley. It has a high content of proteins rich in gliadin peptides, which are resistant to complete digestion by gastrointestinal enzymes, and may cause an inflammatory response leading to celiac disease in genetically predisposed individuals.¹ Celiac disease is an autoimmune enteropathy affecting approximately 1% of the Western population and is attributable to both genetic and environmental factors.² Although gluten consumption and certain HLA antigen genotypes are key factors for celiac disease development, not all individuals with a predisposing genetic background develop lifelong intolerance to gluten,³ and the risk is likely to be modified by the timing or quantities of gluten consumed as well as other potential pathophysiological factors.^4,5

Celiac disease commonly presents during early childhood,⁶ highlighting the importance of studying early life events to identify triggers of the disease.⁷ It was initially reported that early or late introduction of gluten to infants increased the risk of celiac disease.^8,9 The timing of infant gluten exposure has not been consistently associated with celiac disease risk,^10,11 and this has led to changing recommendations for infant feeding.¹² It remains unclear whether the amount of gluten consumed triggers celiac disease.^11,13,14,15

Gluten intake during the first 5 years of life was assessed in genetically at-risk children followed up in The Environmental Determinants of Diabetes in the Young (TEDDY), a multinational prospective birth cohort study. The aim was to investigate whether the amount of gluten in the diet was associated with development of celiac disease autoimmunity and celiac disease to allow better understanding of the pathogenesis and inform feeding recommendations to minimize disease burden.

Methods

Study Population

This prospective cohort study was designed to follow up children from birth up to 15 years of age at 6 clinical research centers in Finland, Germany, Sweden, and the United States (one center in Colorado, one center for Florida and Georgia, and one center in Washington state).¹⁶ The enrollment period was from September 2004 through February 2010 and the final date of follow-up was September 30, 2017.

The primary goal was to identify genetic and environmental factors associated with increased risk of type 1 diabetes, celiac disease, or both. Newborn infants were screened for HLA antigen genotypes associated with type 1 diabetes and celiac disease.¹⁷ Distribution of the HLA antigen genotypes in the study appear in Table 1.

Table 1. Descriptive Characteristics of the Children by Study Outcome.

	No. (%) of Children^a
	Always Negative for tTG Autoantibodies (n = 5194)	Celiac Disease Autoimmunity (n = 1216)	Celiac Disease (n = 447)
Sex
Male	2741 (52.8)	523 (43.0)	166 (37.1)
Female	2453 (47.2)	693 (57.0)	281 (62.9)
Finland
Total	1218 (23.5)	251 (20.6)	78 (17.4)
HLA DR3-DQ2/DR3-DQ2^b	124 (10.2)	79 (31.5)	36 (46.2)
HLA DR3-DQ2/DR4-DQ8^c	376 (30.9)	120 (47.8)	30 (38.5)
Other HLA antigen genotypes^d	718 (58.9)	52 (20.7)	12 (15.4)
Germany
Total	314 (6.1)	57 (4.7)	16 (3.6)
HLA DR3-DQ2/DR3-DQ2^b	50 (15.9)	22 (38.6)	9 (56.2)
HLA DR3-DQ2/DR4-DQ8^c	131 (41.7)	19 (33.3)	4 (25.0)
Other HLA antigen genotypes^d	133 (42.4)	16 (28.1)	3 (18.8)
Sweden
Total	1554 (29.9)	464 (38.2)	222 (49.7)
HLA DR3-DQ2/DR3-DQ2^b	225 (14.5)	202 (43.5)	108 (48.6)
HLA DR3-DQ2/DR4-DQ8^c	690 (44.4)	152 (32.8)	66 (29.7)
Other HLA antigen genotypes^d	639 (41.1)	110 (23.7)	48 (21.6)
United States
Total	2108 (40.5)	444 (36.5)	131 (29.3)
HLA DR3-DQ2/DR3-DQ2^b	391 (18.5)	194 (43.7)	69 (52.7)
HLA DR3-DQ2/DR4-DQ8^c	849 (40.3)	183 (41.2)	50 (38.2)
Other HLA antigen genotypes^d	868 (41.2)	67 (15.1)	12 (9.2)
First-degree relative with celiac disease	129 (2.5)	126 (10.4)	77 (17.2)
Breastfeeding duration, median (IQR), mo	7.8 (3.5-12.0)	8.3 (5.0-12.0)	8.1 (5.0-12.0)
Age at gluten introduction, mean (SD), mo	6.2 (1.9)	6.1 (1.8)	5.9 (1.9)

Open in a new tab

Abbreviations: IQR, interquartile range; tTG, tissue transglutaminase.

^{^a}

Unless otherwise indicated.

^{^b}

Detailed description of HLA antigen genotypes: DR3-DQA1*05:01-DQB1*02:01/DR3-DQA1*05:01-DQB1*02:01.

^{^c}

Detailed description of HLA antigen genotypes: DR4-DQA1*03:0X-DQB1*03:02/DR3-DQA1*05:01-DQB1*02:01.

^{^d}

Detailed description of HLA antigen genotypes: DR4-DQA1*03:0X-DQB1*03:02/DR4-DQA1*03:0X-DQB1*03:02 or DR3-DQA1*05:01-DQB103:02/DR8-DQA1*04:01-DQB1*04:02, DR4-DQA1*03-DQB1*03:02/DR3-DQA1*05:01-DQB1*02:01, DR4-DQA1*03-DQB1*03:02/DR4-DQA1*03-DQB1*03:02, DR4-DQA1*03-DQB1*03:02/DR8-DQA1*04:01-DQB1*04:02, DR3-DQA1*05:01-DQB1*02:01/DR3-DQA1*05:01-DQB1*02:01, DR4-DQA1*03-DQB1*03:02/DR4-DQA1*03-DQB1*02, DR4-DQA1*03-DQB1*03:02/DR1-DQA1*01:01-DQB1*05:01, DR4-DQA1*03-DQB1*03:02/DR13-DQA1*01:02-DQB1*06:04, DR4-DQA1*03-DQB1*03:02/DR9-DQA1*03-QB1*03:03, or DR3-DQA1*05:01-DQB1*02:01/DR9-DQA1*03-DQB1*03:03.

For all study participants, separate written informed consent was obtained from a parent or primary caretaker for genetic screening and participation in the prospective follow-up beginning at birth. Local institutional or regional ethics review boards in all participating countries approved the study. Details of the study design, eligibility, and methods have been published.^16,18,19,20

Dietary Assessment

Gluten intake was estimated from 3-day food records collected at ages 6, 9, and 12 months and biannually (ie, at 18, 24, 30, and 36 months) thereafter until the age of 5 years. Parents were asked to keep a food record documenting all foods and drinks consumed by the child over 3-day periods (2 weekdays and 1 weekend day) before the scheduled clinic visit. Normal food habits were encouraged during the time of food record collection. Portion sizes were estimated using household measurements, food models, pictures, drawings, and shapes of foods as references. A specific booklet was developed and used in all countries to facilitate estimation of food portion sizes. The dietary assessment method used in the study has been described elsewhere.^15,21

Dietary intake was analyzed using the food composition databases from each participating country. For the analyses at the food-group level, a harmonized food-grouping system was developed with comparable food groups and quantification of food intakes among the databases used in the individual countries.²² Composite foods and recipes were broken down to ingredients. Mean intake (grams/day) was calculated from total intake of gluten-containing flours (wheat, rye, and barley) reported during the 3-day recording period. Vegetable protein content (using country-specific values) was obtained from the daily intake of gluten-containing flours and converted to the amount of gluten using a conversion factor of 0.8 (the gluten content in wheat protein).²³ The converted amount was analyzed as absolute gluten intake (grams/day).

Measurement of Tissue Transglutaminase Autoantibodies

Testing for serum tissue transglutaminase (tTG) autoantibodies started at the 24-month clinic visit and was continued yearly thereafter. Radiobinding assays were used to measure tTG autoantibody levels at 2 laboratories as described.¹⁹ Briefly, serum samples from US centers were screened for IgA-tTG autoantibodies at the Barbara Davis Center for Childhood Diabetes, University of Colorado (Denver laboratory).²⁴

Serum samples from the European centers were tested at the University of Bristol (Bristol laboratory) using an assay that detected both IgA and IgG autoantibodies against tTG.²⁵ To harmonize the results, all samples with a tTG autoantibody index greater than 0.01 at the Denver laboratory were sent for quantification at the Bristol laboratory, which was the reference laboratory for the study.¹⁹

The results were expressed in arbitrary units derived from a standard curve consisting of dilutions of serum samples taken from a patient with celiac disease. If a sample tested positive from the Bristol laboratory (≥1.3 U),²⁵ the child’s earlier serum samples were retrospectively analyzed at the Bristol laboratory to determine the age at which the tTG autoantibodies first became detectable. Persistence of tTG autoantibodies was confirmed if positive results were found for 2 consecutive serum samples collected at least 3 months apart.²⁶

Primary and Secondary Outcomes

The primary outcome was celiac disease autoimmunity (defined as positive tTG autoantibodies found at the Bristol laboratory for 2 consecutive serum samples). Children meeting the criteria for persistence of tTG autoantibodies were referred to a gastroenterologist at the clinical discretion of their usual physician. The decision whether to perform a biopsy was not determined by the study protocol.

The secondary outcome was celiac disease (defined as an intestinal biopsy showing a Marsh score of ≥2 or, if biopsy was not performed, when the average of 2 samples was ≥100 U).²⁶

Statistical Analyses

The time to an event was defined as the age of the first positive tTG autoantibody sample for children who later fulfilled the criteria for both celiac disease autoimmunity and celiac disease. The right-censored time for celiac disease autoimmunity was the age at the last negative tTG autoantibody sample and for celiac disease was the age at the last clinic visit when celiac disease had not been diagnosed. To control for differences in age or body size, we analyzed energy and age-adjusted gluten intake using the residual method,²⁷ as well as gluten intake per 10 kg of body weight at a given age, in addition to absolute daily intake of gluten.

To address concerns regarding missing data and variability in dietary data, joint modeling was selected as the prespecified analysis and was chosen to assess the association between gluten intake over time and the risk of celiac disease autoimmunity and celiac disease.^28,29 Joint modeling assesses the association by fitting an individual trajectory for gluten intake over time. Based on the patterns detected, a linear trajectory for gluten intake was assumed for the longitudinal model, and the incidence peak during the beginning of the study was considered for the baseline hazard estimation, assuming a piecewise constant. Seven intervals without weighting were applied per the best model fit based on the difference in the Akaike information criterion.³⁰ The longitudinal model was adjusted for energy intake (kilocalories/day) at the same time, and the time-to-event model was adjusted for HLA antigen genotype, sex, country of residence, and family history of celiac disease (mother, father, or sibling). The SAS macro JMFit was used for the analyses.³¹ From the log-hazard model fitted by joint modeling, absolute risk by the age of 3 years was estimated as the cumulative hazard in relation to the mean daily gluten intake at 2 years of age. The hazard ratios (HRs) and absolute risk differences were assessed at a 1-g/d or 1-g/d/10 kg (of body weight) increase of gluten intake.

In addition, 2 Cox regression analyses were performed as sensitivity analyses that included the most recent gluten intake prior to the event and energy intake as time-dependent covariates. One analysis included all children, and the other analysis included children with gluten intake available within 1 year prior to each risk set to control for various lag times between gluten exposure and the event. Because an early peak incidence of seroconversion was found, we examined the effects of age-specific gluten intake as a post hoc analysis. The association with subsequent incidence of celiac disease autoimmunity and celiac disease was assessed using Cox regression, focusing on absolute intake reported at the age of each study visit. For children whose gluten intake at the specific age was the most recent data prior to the event, the standard Cox regression model assessed the effects of gluten intake as a time constant covariate. For children who had additional gluten intake data available after the specific age, the most recent gluten intake prior to the event needed to be controlled for to assess the effects of the intake. To assess the effects of age-specific gluten intake in addition to the effect of current intake, the model considered the most recent intake prior to the event as a time-dependent covariate and the intake at the specific age as a time-constant covariate.

The proportional hazards assumption was examined using the martingale residual analysis with the supremum test. The functional form in the martingale residual plot, as well as a change-point analysis based on the log-rank test,³² suggested a dichotomization for absolute gluten intake at 2 years of age. Two-sided P values are reported. Statistical significance was determined when the P value was <.05. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc).

Results

Between September 2004 and February 2010, 424 788 newborn infants were screened for HLA antigens. Of those screened, there were 21 589 infants (5%) identified as being eligible for study inclusion and 8676 (40%) were enrolled before the age of 4 months. The most common reasons for not enrolling were related to protocol characteristics (eg, blood draw, demanding protocol) or family circumstances (eg, changing contact information).³³

There were 6757 children screened for tTG autoantibodies and 6605 (97.8%) submitted at least one 3-day food record during the first 5 years of life or prior to detection of tTG autoantibody positivity (Figure 1). Descriptive characteristics of the study population appear in Table 1.

Of 6605 children included in the study, 3233 (49%) were girls. Gluten consumption linearly increased with age, but there were some national differences (Figure 2). Mean daily gluten intake per visit and country is presented in eTable 1 in the Supplement. Of the 52 952 visits for which parallel tTG autoantibody results were available, data on gluten intake were missing or of inadequate quality at 4465 visits (8%). In total, 204 (3%) participants completed only 1 food record. Among children with celiac disease autoimmunity, 20 (1.6%) completed only 1 food record more than 3 months prior to their seroconversion.

Figure 2. — Error bars indicate 95% CIs.

Primary and Secondary Outcome Analyses

As of September 30, 2017, among the 6605 children included in the analysis, 1411 (21%) tested positive for tTG autoantibodies on at least 1 occasion. During a median follow up of 9.0 years (range, 1.0-13.0 years; interquartile range, 8.0-10.0 years), 1216 children (18%) developed celiac disease autoimmunity (seroconverted to positive tTG autoantibodies at a median age of 3.3 years [range, 0.9-11.5 years]). There were 447 children (7%) who developed celiac disease (seroconverted at a median age of 3.0 years [range, 0.9-11.2 years]). The incidence of seroconversion for both outcomes peaked at 2 to 3 years of age (eFigure 1 in the Supplement).

Children homozygous for HLA DR3-DQ2 were at the highest risk of celiac disease autoimmunity and celiac disease. Swedish residence, female sex, and family history of celiac disease were also associated with increased risk for both outcomes (eTable 2 in the Supplement).

Higher intake of gluten during the first 5 years of life was associated with increased risk and absolute risk by the age of 3 years in relation to mean daily gluten intake at the age of 2 years for both celiac disease autoimmunity and celiac disease (Table 2 and Table 3). Daily (absolute) gluten intake was associated with higher risk of celiac disease autoimmunity for every 1-g/d increase in gluten consumption (HR, 1.30 [95% CI, 1.22-1.38], P < .001; absolute risk by the age of 3 years if the reference amount of gluten was consumed, 28.1%; absolute risk if the gluten amount consumed was 1-g/d higher than the reference amount, 34.2%; absolute risk difference, 6.1% [95% CI, 4.5%-7.7%]). Daily (absolute) gluten intake was associated with higher risk of celiac disease for every 1-g/d increase in gluten consumption (HR, 1.50 [95% CI, 1.35-1.66], P < .001; absolute risk by the age of 3 years if the reference amount of gluten was consumed, 20.7%; absolute risk if the gluten amount consumed was 1-g/d higher than the reference amount, 27.9%; absolute risk difference, 7.2% [95% CI, 6.1%-8.3%]).

Table 2. Daily Gluten Intake and Risk for Developing Celiac Disease Autoimmunity and Celiac Disease.

Type of Analysis^a	No. of Children With Celiac Disease Autoimmunity	Measurements of Gluten	Celiac Disease Autoimmunity		Celiac Disease
Type of Analysis^a	No. of Children With Celiac Disease Autoimmunity	Measurements of Gluten	HR (95% CI)	P Value	HR (95% CI)	P Value
Joint modeling	1216	Absolute intake, g/d	1.30 (1.22-1.38)	<.001	1.50 (1.35-1.66)	<.001
		Residual intake, g/d^b	1.40 (1.30-1.52)	<.001	1.43 (1.23-1.68)	<.001
		Intake/10 kg of body weight	1.87 (1.66-2.11)	<.001	2.18 (1.75-2.71)	<.001
Cox regression	1216	Absolute intake, g/d	1.14 (1.11-1.17)	<.001	1.14 (1.09-1.20)	<.001
		Residual intake, g/d^b	1.12 (1.09-1.15)	<.001	1.07 (1.02-1.13)	.01
		Intake/10 kg of body weight	1.19 (1.14-1.23)	<.001	1.14 (1.07-1.22)	<.001
Cox regression including only those with gluten consumption available within 1 y prior to time of event	905	Absolute intake, g/d	1.12 (1.08-1.16)	<.001	1.07 (1.02-1.13)	.01
		Residual intake, g/d^b	1.09 (1.05-1.13)	<.001	1.04 (0.99-1.10)	.14
		Intake/10 kg of body weight	1.15 (1.10-1.20)	<.001	1.12 (1.05-1.21)	.002

Open in a new tab

Abbreviation: HR, hazard ratio.

^{^a}

Adjusted for HLA antigen genotype, country, sex, first-degree relative with celiac disease, and energy intake.

^{^b}

Adjusted for age and energy intake using the residual method.²⁷

Table 3. Absolute Risk for Developing Celiac Disease Autoimmunity and Celiac Disease^a.

Measurements of Gluten	Reference Gluten Intake, Mean (SD)^b	Absolute Risk for Developing Celiac Disease Autoimmunity by Age of 3 y		Absolute Risk Difference, % (95% CI)^d	Absolute Risk for Developing Celiac Disease by Age of 3 y		Absolute Risk Difference, % (95% CI)^d
Measurements of Gluten	Reference Gluten Intake, Mean (SD)^b	If Reference Amount of Gluten Was Consumed, %^c	If Amount of Gluten Consumed Was 1-U Higher Than Reference Amount, %	Absolute Risk Difference, % (95% CI)^d	If Reference Amount of Gluten Was Consumed, %^c	If Amount of Gluten Consumed Was 1-U Higher Than Reference Amount, %	Absolute Risk Difference, % (95% CI)^d
Absolute intake, g/d	3.71 (1.75)	28.1	34.2	6.1 (4.5-7.7)	20.7	27.9	7.2 (6.1-8.3)
Residual intake, g/d^e	0.48 (1.67)	18.7	24.6	5.9 (4.4-7.4)	7.8	10.7	2.9 (1.9-3.8)
Intake/10 kg of body weight	2.91 (1.39)	51.9	70.2	18.3 (16.7-19.9)	35.0	55.0	20.0 (19.0-21.0)

Open in a new tab

^{^a}

Analysis was conditioned on HLA antigen genotype, country, sex, first-degree relative with celiac disease, and energy intake. Cumulative hazard from the log-hazard model fit by joint modeling in Table 2.

^{^b}

Average amount reported at the 2-year visit.

^{^c}

Assessed in relation to the average daily gluten intake at 2 years of age.

^{^d}

Implies the risk increase if the gluten intake was 1-g/d higher than the average gluten intake at 2 years of age.

^{^e}

Adjusted for age and energy intake using the residual method.²⁷

Age- and energy-adjusted (residual) gluten intake was associated with higher risk of celiac disease autoimmunity for every per 1-g/d increase in gluten consumption (HR, 1.40 [95% CI, 1.30-1.52], P < .001; absolute risk by the age of 3 years if the reference amount of gluten was consumed, 18.7%; absolute risk if the gluten amount consumed was 1-g/d higher than the reference amount, 24.6%; absolute risk difference, 5.9% [95% CI, 4.4%-7.4%]). Age- and energy-adjusted (residual) gluten intake was associated with higher risk of celiac disease for every per 1-g/d increase in gluten consumption (HR, 1.43 [95% CI, 1.23-1.68], P < .001; absolute risk by the age of 3 years if the reference amount of gluten was consumed, 7.8%; absolute risk if the gluten amount consumed was 1-g/d higher than the reference amount, 10.7%; absolute risk difference, 2.9% [95% CI, 1.9%-3.8%]).

In addition, gluten intake per 10 kg of body weight was associated with a higher risk of celiac disease autoimmunity for every 1-g/d/10 kg increase in gluten consumption (HR, 1.87 [95% CI, 1.66-2.11], P < .001; absolute risk by the age of 3 years if the reference amount of gluten was consumed, 51.9%; absolute risk if the gluten amount consumed was 1-g/d/10 kg higher than the reference amount, 70.2%; absolute risk difference, 18.3% [95% CI, 16.7%-19.9%]). Gluten intake per 10 kg of body weight was associated with a higher risk of celiac disease for every 1-g/d/10 kg increase in gluten consumption (HR, 2.18 [95% CI, 1.75 -2.71], P < .001; absolute risk by the age of 3 years if the reference amount of gluten was consumed, 35.0%; absolute risk if the gluten amount consumed was 1-g/d/10 kg higher than the reference amount, 55.0%; absolute risk difference, 20.0% [95% CI, 19.0%-21.0%]).

Sensitivity Analyses

The sensitivity analyses using Cox regression models generally supported the statistical significance found from the joint modeling analysis (Table 2). In the country-specific analyses, higher gluten intake was associated with an increased risk of celiac disease autoimmunity in all countries and at all sites (eTable 3 in the Supplement). Absolute gluten intake and age- and energy-adjusted intake were associated with increased risk for celiac disease in the United States (specifically, in Colorado and Washington State) and in Sweden. The analyses could not be performed in Germany because there were only 16 cases of celiac disease.

Post Hoc Analysis

Gluten intake reported at the 2-year visit was available for 833 children with celiac disease autoimmunity. Gluten intake reported at the 3-year visit was available for 526 children with celiac disease autoimmunity. The post hoc analysis showed that gluten intake at 2 years of age had an independent association with the risk of celiac disease autoimmunity and celiac disease, in addition to the current intake during the first 5 years of life (eTable 4 in the Supplement).

The supremum test showed no indication of violating the proportional hazards assumption, but there was a deviation with gluten intake greater than 2 g/d in the martingale residual plot (eFigure 2 in the Supplement). In addition, the change point analysis showed a significant risk difference between gluten intake greater than 2 g/d and gluten intake of 2 g/d or less. Based on these analyses, we dichotomized the gluten intake reported at 2 years as greater than 2 g/d and 2 g/d or less and examined the adjusted HRs with the outcomes (eTable 5 in the Supplement).

Gluten consumption greater than 2 g/d at 2 years of age was associated with a higher risk of celiac disease autoimmunity (HR, 1.49 [95% CI, 1.16-1.91], P = .002) and celiac disease (HR, 1.75 [95% CI, 1.10-2.81], P = .02) compared with those who consumed 2 g/d or less. When analyzing absolute gluten intake reported at the 2-year visit and risk for developing celiac disease autoimmunity and celiac disease, a linear increase in HRs were seen with higher intake of gluten (eTable 6 in the Supplement).

Discussion

Higher gluten intake during the first 5 years of life was associated with statistically significantly increased risks of celiac disease autoimmunity and celiac disease among genetically predisposed children. The incidence of both outcomes peaked at 2 to 3 years of age. If gluten intake was 1-g/d higher than the mean at 2 years of age (corresponding to a half slice of white bread), the absolute risk differences for the respective outcome were 6% and 7% higher by 3 years of age. The 6% to 7% increase in risk for a small 1-g/d increase in gluten intake appears clinically important.

In addition, the finding of a cut point at which gluten intake was associated with an increased risk is relevant for gluten feeding recommendations in at-risk children; however, this conclusion was based on a post hoc analysis and should be confirmed. The association of gluten intake with these risks was significantly increased if the child consumed more than 2 g/d of gluten at around 2 years of age, which corresponds to approximately 1 slice (35 g) of white bread or 1 portion of cooked pasta (150 g). The HR increased with subsequent higher gluten intake at the 2-year visit, further supporting the results that higher intakes of gluten were associated with higher risks of celiac disease autoimmunity and celiac disease.

The findings from a previous case-control study of gluten intake among Swedish children born during the mid-1980s showed that children subsequently diagnosed with celiac disease had been introduced to larger amounts of gluten-containing foods compared with children who did not develop celiac disease.¹³ The hypothesis that gluten given in small amounts at 5 to 6 months of age would protect at-risk children from developing celiac disease was addressed in a randomized placebo-controlled intervention trial; however, that study produced null results.¹¹ In the same study population, mean daily gluten intake (from 10 months of age when unrestricted gluten consumption was allowed) was not associated with celiac disease up to 3 years of age, except in children carrying the HLA antigen genotype HLA-DQ2.2/-DQ7.¹⁴

In contrast to the randomized placebo-controlled intervention trial, gluten consumption during the first 2 years of life was found to be associated with increased risk of celiac disease in a subset of Swedish children from the present cohort, and furthermore, children in the upper tertile of gluten intake were at a 2-fold increased risk of celiac disease vs children with lower gluten intake. This nested case-control study including 146 children with biopsy-confirmed celiac disease and 436 matched controls indicated that the amount of gluten consumed could be a risk factor for celiac disease.¹⁵

For the current study, food record data from all the participating countries were harmonized, which made it feasible to perform a longitudinal analysis of the full birth cohort. In addition, the current analysis included gluten intake up to 5 years of age and another 301 children diagnosed with celiac disease, which made it possible to do country-specific analyses using time-to-event analyses. In the country-specific analyses, a higher gluten intake was associated with an increased risk of celiac disease autoimmunity in all countries, whereas absolute gluten intake and age- and energy-adjusted intake were only associated with increased risk for celiac disease in the United States and Sweden.

Despite similar dietary assessment methods and calculation of gluten intake, discrepancies in results among the studies are likely attributable to study design and population size. In the randomized clinical trial, the time of gluten introduction was not accounted for and gluten amounts were fixed,¹¹ which differed from the present observational study consisting of a larger population that reflected the natural variations of gluten intake in real life. Other contributing factors may be differences in exposures to various triggering environmental factors, such as gastrointestinal infections or rotavirus vaccination status,^4,5 which could partly explain why Swedish children are more prone to develop celiac disease compared with children from other countries.

Because gluten intake was measured as the mean from 3-day food records, there were missing data and day-to-day variation. The joint modeling method simultaneously models longitudinal data and time-to-event data. By fitting the longitudinal pattern through modeling, missing gluten intake was imputed by the fit, and variability also was considered. However, Cox regression includes observed data only and assumes covariates without any variability. Therefore, joint modeling was considered as the primary analytic method, but various sensitivity analyses were conducted using Cox regression. Joint modeling tends to produce greater effect sizes.²⁸

A major strength of this study is its prospective design, enrolling a large cohort of children with the same genetic risk from 4 countries with different infant feeding habits and following the same study protocol. Another strength is the dietary assessment method that allowed repeated measurements to capture changes in dietary habits in growing infants and young children over time prior to disease onset. The prospective design also reduced the effect of changes in dietary habits because parents were unaware of their child’s autoantibody status when the food records were collected. The analyses also were adjusted for known confounders for celiac disease (HLA antigen, country, sex, and having a family member with celiac disease).²⁶ Moreover, potential confounders such as socioeconomic status, maternal smoking during pregnancy, maternal education, and maternal age had previously been analyzed and were not associated with risk of celiac disease,³⁴ and are therefore considered less likely to confound the results.

Limitations

This study has several limitations. First, information on analyzed gluten content in foods in national food composition databases was lacking. Due to variability in protein content in different types, cultivars, and crops of wheat, the accuracy of gluten intake would improve if gluten content were available in food composition databases. The same conversion factor for estimation of gluten content in wheat, rye, and barley was chosen because this method has been used in several studies.^10,14,15,35 Other studies have used cereal-specific conversion factors for the estimation of gluten content.³⁶

Second, calculations of gluten content are approximate because they were based on self-reported dietary data. Different dietary assessment methods together with differences in methods of estimating gluten content are challenging when comparing results from previous studies. A randomized clinical trial of different amounts of gluten during early childhood in genetically at-risk individuals would be warranted to confirm our findings.

Conclusions

Higher gluten intake during the first 5 years of life was associated with increased risk of celiac disease autoimmunity and celiac disease among genetically predisposed children.

Supplement.

eTable 1. Daily gluten intake (gram per day) at clinic visits by country

eTable 2. Non-dietary risk factors for celiac disease autoimmunity and celiac disease

eTable 3. Daily gluten intake and risk for developing celiac disease autoimmunity and celiac disease by study site and country

eTable 4. Daily absolute gluten intake reported at the 2 and 3 year study visits and risk for developing celiac disease autoimmunity and celiac disease

eTable 5. Daily absolute gluten intake reported at the 2-year study visit and risk for developing celiac disease autoimmunity and celiac disease

eTable 6. Daily absolute gluten intake reported at the 2-year study visit and risk for developing celiac disease autoimmunity and celiac disease

eFigure 1. Incidence of seroconversion to celiac disease autoimmunity and celiac disease over time

eFigure 2. Martingale residual plots and local linear regression (LOESS line; blue line) for absolute gluten intake (g/day) at the 2 year study visit from the Cox regression model

Click here for additional data file.^{(472.4KB, pdf)}

References

1.Biesiekierski JR. What is gluten? J Gastroenterol Hepatol. 2017;32(suppl 1):78-81. doi: 10.1111/jgh.13703 [DOI] [PubMed] [Google Scholar]
2.Lebwohl B, Sanders DS, Green PHR. Coeliac disease. Lancet. 2018;391(10115):70-81. doi: 10.1016/S0140-6736(17)31796-8 [DOI] [PubMed] [Google Scholar]
3.Tjon JM, van Bergen J, Koning F. Celiac disease: how complicated can it get? Immunogenetics. 2010;62(10):641-651. doi: 10.1007/s00251-010-0465-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Bouziat R, Hinterleitner R, Brown JJ, et al. Reovirus infection triggers inflammatory responses to dietary antigens and development of celiac disease. Science. 2017;356(6333):44-50. doi: 10.1126/science.aah5298 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Kemppainen KM, Lynch KF, Liu E, et al. ; TEDDY Study Group . Factors that increase risk of celiac disease autoimmunity after a gastrointestinal infection in early life. Clin Gastroenterol Hepatol. 2017;15(5):694-702.e5, e5. doi: 10.1016/j.cgh.2016.10.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Hagopian W, Lee HS, Liu E, et al. ; TEDDY Study Group . Co-occurrence of type 1 diabetes and celiac disease autoimmunity. Pediatrics. 2017;140(5):e20171305. doi: 10.1542/peds.2017-1305 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Agardh D, Lee HS, Kurppa K, et al. ; TEDDY Study Group . Clinical features of celiac disease: a prospective birth cohort. Pediatrics. 2015;135(4):627-634. doi: 10.1542/peds.2014-3675 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Norris JM, Barriga K, Hoffenberg EJ, et al. Risk of celiac disease autoimmunity and timing of gluten introduction in the diet of infants at increased risk of disease. JAMA. 2005;293(19):2343-2351. doi: 10.1001/jama.293.19.2343 [DOI] [PubMed] [Google Scholar]
9.Størdal K, White RA, Eggesbø M. Early feeding and risk of celiac disease in a prospective birth cohort. Pediatrics. 2013;132(5):e1202-e1209. doi: 10.1542/peds.2013-1752 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Lionetti E, Castellaneta S, Francavilla R, et al. ; SIGENP (Italian Society of Pediatric Gastroenterology, Hepatology, and Nutrition) Working Group on Weaning and CD Risk . Introduction of gluten, HLA status, and the risk of celiac disease in children. N Engl J Med. 2014;371(14):1295-1303. doi: 10.1056/NEJMoa1400697 [DOI] [PubMed] [Google Scholar]
11.Vriezinga SL, Auricchio R, Bravi E, et al. Randomized feeding intervention in infants at high risk for celiac disease. N Engl J Med. 2014;371(14):1304-1315. doi: 10.1056/NEJMoa1404172 [DOI] [PubMed] [Google Scholar]
12.Szajewska H, Shamir R, Mearin L, et al. Gluten introduction and the risk of coeliac disease: a position paper by the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr. 2016;62(3):507-513. doi: 10.1097/MPG.0000000000001105 [DOI] [PubMed] [Google Scholar]
13.Ivarsson A, Hernell O, Stenlund H, Persson LA. Breast-feeding protects against celiac disease. Am J Clin Nutr. 2002;75(5):914-921. doi: 10.1093/ajcn/75.5.914 [DOI] [PubMed] [Google Scholar]
14.Crespo-Escobar P, Mearin ML, Hervás D, et al. The role of gluten consumption at an early age in celiac disease development: a further analysis of the prospective PreventCD cohort study. Am J Clin Nutr. 2017;105(4):890-896. doi: 10.3945/ajcn.116.144352 [DOI] [PubMed] [Google Scholar]
15.Andrén Aronsson C, Lee HS, Koletzko S, et al. ; TEDDY Study Group . Effects of gluten intake on risk of celiac disease: a case-control study on a Swedish birth cohort. Clin Gastroenterol Hepatol. 2016;14(3):403-409.e3, e3. doi: 10.1016/j.cgh.2015.09.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.TEDDY Study Group The Environmental Determinants of Diabetes in the Young (TEDDY) study: study design. Pediatr Diabetes. 2007;8(5):286-298. doi: 10.1111/j.1399-5448.2007.00269.x [DOI] [PubMed] [Google Scholar]
17.Hagopian WA, Erlich H, Lernmark A, et al. ; TEDDY Study Group . The Environmental Determinants of Diabetes in the Young (TEDDY): genetic criteria and international diabetes risk screening of 421 000 infants. Pediatr Diabetes. 2011;12(8):733-743. doi: 10.1111/j.1399-5448.2011.00774.x [DOI] [PMC free article] [PubMed] [Google Scholar]
18.TEDDY Study Group The Environmental Determinants of Diabetes in the Young (TEDDY) study. Ann N Y Acad Sci. 2008;1150:1-13. doi: 10.1196/annals.1447.062 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Vehik K, Fiske SW, Logan CA, et al. ; TEDDY Study Group . Methods, quality control and specimen management in an international multicentre investigation of type 1 diabetes: TEDDY. Diabetes Metab Res Rev. 2013;29(7):557-567. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Rewers M, Hyöty H, Lernmark Å, et al. ; TEDDY Study Group . The Environmental Determinants of Diabetes in the Young (TEDDY) study: 2018 update. Curr Diab Rep. 2018;18(12):136. doi: 10.1007/s11892-018-1113-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Yang J, Lynch KF, Uusitalo UM, et al. ; TEDDY Study Group . Factors associated with longitudinal food record compliance in a paediatric cohort study. Public Health Nutr. 2016;19(5):804-813. doi: 10.1017/S1368980015001883 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Joslowski G, Yang J, Aronsson CA, et al. ; TEDDY Study Group . Development of a harmonized food grouping system for between-country comparisons in the TEDDY study. J Food Compost Anal. 2017;63:79-88. doi: 10.1016/j.jfca.2017.07.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.van Overbeek FM, Uil-Dieterman IG, Mol IW, Köhler-Brands L, Heymans HS, Mulder CJ. The daily gluten intake in relatives of patients with coeliac disease compared with that of the general Dutch population. Eur J Gastroenterol Hepatol. 1997;9(11):1097-1099. doi: 10.1097/00042737-199711000-00013 [DOI] [PubMed] [Google Scholar]
24.Bao F, Yu L, Babu S, et al. One third of HLA DQ2 homozygous patients with type 1 diabetes express celiac disease-associated transglutaminase autoantibodies. J Autoimmun. 1999;13(1):143-148. doi: 10.1006/jaut.1999.0303 [DOI] [PubMed] [Google Scholar]
25.Bingley PJ, Williams AJ, Norcross AJ, et al. ; Avon Longitudinal Study of Parents and Children Study Team . Undiagnosed coeliac disease at age seven: population based prospective birth cohort study. BMJ. 2004;328(7435):322-323. doi: 10.1136/bmj.328.7435.322 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Liu E, Lee HS, Aronsson CA, et al. ; TEDDY Study Group . Risk of pediatric celiac disease according to HLA haplotype and country. N Engl J Med. 2014;371(1):42-49. doi: 10.1056/NEJMoa1313977 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr. 1997;65(4)(suppl):1220S-1228S. doi: 10.1093/ajcn/65.4.1220S [DOI] [PubMed] [Google Scholar]
28.Asar Ö, Ritchie J, Kalra PA, Diggle PJ. Joint modelling of repeated measurement and time-to-event data: an introductory tutorial. Int J Epidemiol. 2015;44(1):334-344. doi: 10.1093/ije/dyu262 [DOI] [PubMed] [Google Scholar]
29.Tsiatis A, Davidian M. Joint modeling of longitudinal and time-to-event data: an overview. Stat Sin. 2004;14(3):809-834. [Google Scholar]
30.Zhang D, Chen MH, Ibrahim JG, Boye ME, Wang P, Shen W. Assessing model fit in joint models of longitudinal and survival data with applications to cancer clinical trials. Stat Med. 2014;33(27):4715-4733. doi: 10.1002/sim.6269 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Zhang D, Chen MH, Ibrahim JG, Boye ME, Shen W. JMFit: a SAS macro for joint models of longitudinal and survival data. J Stat Softw. 2016;71(3). doi: 10.18637/jss.v071.i03 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Contal C, O’Quigley J. An application of changepoint methods in studying the effect of age on survival in breast cancer. Comput Stat Data Anal. 1999;30:253-270. doi: 10.1016/S0167-9473(98)00096-6 [DOI] [Google Scholar]
33.Lernmark B, Johnson SB, Vehik K, et al. Enrollment experiences in a pediatric longitudinal observational study: The Environmental Determinants of Diabetes in the Young (TEDDY) study. Contemp Clin Trials. 2011;32(4):517-523. doi: 10.1016/j.cct.2011.03.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Aronsson CA, Lee HS, Liu E, et al. ; TEDDY Study Group . Age at gluten introduction and risk of celiac disease. Pediatrics. 2015;135(2):239-245. doi: 10.1542/peds.2014-1787 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Hopman EG, Pruijn R, Tabben EH, le Cessie S, Mearin ML. Food questionnaire for the assessment of gluten intake by children 1 to 4 years old. J Pediatr Gastroenterol Nutr. 2012;54(6):791-796. doi: 10.1097/MPG.0b013e31825144fe [DOI] [PubMed] [Google Scholar]
36.Hoppe C, Trolle E, Gondolf UH, Husby S. Gluten intake in 6-36-month-old Danish infants and children based on a national survey. J Nutr Sci. 2013;2:e7. doi: 10.1017/jns.2013.1 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials