Type 1 diabetes (T1D) emerge when the auto-immune destruction exceeds the beta cell's regenerative capacity. The patients' beta-cell capacity increases shortly after onset when glucotoxicity decreases after the start of insulin therapy. Children have fewer beta cells and therefore shorter remission; but the expansion potential is larger the younger the child is. The problem with the majority of intervention studies is the many and serious side effects, or a quite marginal effect on the residual beta-cell function. However, in animals that had received gluten-free diet, the T1D incidence fell from 61% to only 6%. Gluten-free diet increases the number of regulatory T cells in Peyer's patches, affect the composition of intestinal microflora and modify the balance between pro and anti-inflammatory cytokines in T cells. Therefore, the aim of our study is to prolong the remission phase by introducing a gluten-free diet intervention to children at T1D onset.

Introduction Type 1 diabetes (T1D) is characterised by a progressive loss and destruction of beta cells. The optimal timing of intervention is during remission when the beta cells regenerative power is still present, because we are treating only already diseased children, and there are enough beta cells to improve outcome if the cells survive.

Thus, the overall aim of this study is to prolong the remission phase by introducing glutenfree diet intervention to children with T1D and in addition, investigate the role of physical activity/fitness.

Hypothesis:

• Gluten-free diet leads to a shift in balance between the destruction and the regenerative capacity as well as increases insulin sensitivity.
• The diet have an impact on the gut microbiome composition and projected function, metabolic, immune and inflammatory biomarkers.

This is the first study with a non-pharmacological intervention in children combining both factors with a potential effect on the immune system as well as on insulin sensitivity and the first to include dietary factors, measures of physical fitness and measures of inflammation.

Perspectives: If glutenfree prolong remission other research areas could be interesting such as change in microRNA following diet change, the influence of diet on fat and carbohydrate metabolism in children, diet habits in children with diabetes, who makes the healthy choices or more research of the influence of glucose transporters (GLUT4) in children and insulin sensitivity. Finally, patients will benefit from a prolonged remission with a decreased burden of the disease by fewer severe hypoglycemic events, fewer long-term complications and it will lead to a substantial reduction in the costs for the society.

Background:

T1D is a chronic disease with high morbidity and mortality but with preserved beta-cells there are less hypoglycemic events and less progression in late complications (1), therefore the preservation and expansion of beta-cells is a goal. Intervention studies in type 1 patients have primarily focused on immune suppressing or immune modulation (2). Most of those studies have several and serious side effects or no effect on preserving or expanding the number of beta cells.

Why is gluten a key factor? Animal models have proven that hydrolyzed diet and gluten-free diet reduces the incidence of diabetes substantially (3-5). Gluten consists of glutenin and gliadin. Incomplete degradation of gluten irritates the intestinal mucosa, resulting in an unspecific subclinical inflammation. Compared with conventional foods, a gluten-free diet increases the number of regulatory T cells in Peyer's patches in the gut (6). Gluten is also known to affect the composition of the intestinal microflora (7). A gluten-rich diet has been shown to modify the balance between pro and anti-inflammatory cytokines in T cells (8) Rats have gut dysfunction and increased gut-permeability, the same is found in pre-T1D humans (9), and the probable mechanism is that gliadin increases the gut hormone zonulin, which again increases the gut permeability (10). Enterovirus also increases the gut permeability (11). The consequences of this can be an increased uptake of bacterial toxins such as lipopolysaccharides, and incompletely degraded gliadin passing through the gut barrier. Gliadin has been found in maternal breast milk in healthy mothers. The gliadin must be transported by the blood, which means that organs with a relatively high blood flow, such as the islets of Langerhans, could also be exposed to gliadin and may actually accumulate gliadin. Gluten affect the micro flora in the gut (12). It has been demonstrated that human microbiota can be transplanted into germ-free mice and transfer host phenotypes ('personalized gnotobiotics'). Human studies have indicated that the introduction of wheat in the diet before the age of four months increases the risk of beta-cell auto-immunity by up to four times (13;14).Two human intervention studies of individuals with a high risk of developing diabetes who were treated with a gluten-free diet over six or 12 months showed no impact on diabetes auto-immunity (15) or diabetes incidence (16), but they did show increased insulin sensitivity during the period with gluten-free diet. The intervention may have been too short to affect autoimmunity.

Rationale for focusing on the composition of the diet and physical fitness in addition to gluten During remission phase the insulin can be reduced from a need for up to 2 units/kg to less than 0.2 units/kg within days/weeks after onset of insulin treatment proving the influence of glucotoxicity on beta cells.

The dietary requirements of fewer carbohydrates and low glycaemic index have been relaxed with the introduction of faster-acting types of insulin and pump treatment. But several studies have shown that intensive treatment can contribute to preserving more beta cells (17,18), perhaps because the beta cells are quiescent to a higher degree, rather than active, as active beta cells are more sensitive to toxic stimuli (19). In addition, the composition of the diet is important for the glycaemic control (20) and potentially affects the immune system indirectly by affecting the intestinal microflora (21). The quantity and quality of carbohydrates can also be important for the insulin requirement.

Physical activity and sedentary lifestyle is known to be associated with insulin resistance (IR). The association between sedentary lifestyle and IR measured in the fasting state is partly explained by adipositas, whereas the associations between IR and insulin secretion based on Oral Glucose Tolerance Test is independent of adipositas (22). Studies also show higher insulin sensitivity if patients are physical active just before a meal (23) and increased fitness associates to increased level of glucose transporters (GLUT4) and thereby the exploit of insulin (24).

Our research group has been involved in studies of the immune system, risk factors for T1D and the remission phase for several years. We have already proven that families are interested in changing diet if there is a chance it prolong remission.

Methods and material Design: A prospective study including all newly diagnosed children with type 1 diabetes will be instructed and supervised by a trained dietician to eat a gluten-free- and low-glycaemic diet within 2 month after diagnosis and followed by a dedicated team of a medical doctor, a nurse and a dietician at entry in the protocol and at 6 and 12 month after diagnosis.

The diet: Families will be instructed by a dietician in either

• eating gluten-free
• normal guidelines concerning carbohydrate in diabetic children The diet will be estimated based on validated food frequency questionnaire - filled online before study entry and after 6, 12 months.

Primary outcome: Insulin adjusted HbA1c (IDAA1c) (25); Stimulated C-peptide; Insulin dose per kg Secondary outcome: The alterations in gut microbiome composition and function and metabolic, immune and inflammatory biomarkers.

Boost- test (meal-stimulation test) The meal test is done after 8 hours fasting in the morning. Boost (Med. Johnson, Evansville, Indiana, USA; 237 ml = 8 ounce (OZ); 33 g carbohydrate, 15 g protein and 6 g fat, 240 kcal): 6 ml/kg (max.: 360 ml). Blood samples and blood glucose are taken at 0 minutes and again 90 minutes after intake of Boost. The Boost-test is performed at inclusion plus 6 and 12 month post diagnosis.

Fear of hypoglycemia and Quality of life: We will use validated questionnaire for fear of hypoglycemia and quality of life.

Gut microbiota: Three stool samples are taken following standard operation procedures (SOP)s) at the indicated time points (inclusion and 6 month). Stool samples are sent for quantitative metagenomics and analysis of microbial DNA at Novo Nordisk Foundation Center for Basic Metabolic Research, Copenhagen University.

Inflammation:

The level of master transcription factors are responsible for the shift of Cluster Differentiation (CD) 4+ T cells into different phenotypes: T-bet for Th1 cells, Trans-acting T-cell-specific transcription factor GATA-3 for Th2, Foxp3 for T-regs, RORγt (retinoic acid receptor-related orphan nuclear receptor gamma) for Th17, and DX5 as a marker of Natural Killer cells. Isolated lymphocytes are saved and resuspended in 500 µL Trizol (Invitrogen) reagent for RNA extraction and stored in -80 ˚C until use. The Bartholin Institute, Rigshospitalet will be responsible for the study of diet induced proportional changes in the above mentioned T cell populations.

Statistics and power calculations Statistics: Multiple linear regression will be used to find other measures of insulin sensitivity and test factors associated with c-peptide decline and IdaA1c. Relevant confounders such as age, c-peptide at onset, diabetes duration and gender will be adjusted for in the models.

Power: We need 100 individuals, where 50% complete the glutenfree diet, to find a different in IdaA1c of 1 with a power of 80% and significance of 0.05. The mean in previous cohorts was 10,7 after 12 months and Standard Deviation = 2,0. 27. To find a difference in c-peptide we need 128 individuals in each group with an 80% power and significance level of 0.05. If we can use previous cohorts as controls we only need 58 patients in the intervention group.

Feasibility Paediatric Department at Herlev Hospital has at least 60 new children with T1D each year, and approximately 50% now agrees to participate in the pilot study with glutenfree diet. We have access to two cohorts of 270 and 129 children respectively both followed for at least 12 month with boost test at 1,6 and 12 month. Paediatric Department is the largest center for childhood diabetes in Denmark, we have previously been responsible for follow-up studies in newly diagnosed children during remission and describe genes and biomarkers (Zink autoantibodies, Interleukin-1ra, adiponectin, glucagon) associated with remission. Furthermore we have investigated cytokines and immuno globulins associated with risk of T1D. We have also found increasing chemokines and unchanged vitamin D in newly diagnosed children.

Collaborators:

Henrik B. Mortensen, professor and research leader at the Paediatric Department, has been working with remission phase through decades and published numerous studies on the subject.

Professor Karsten Buschard, Rigshospitalet, and his postdoc Julie Antvorskov have in animal models studied different T-cell populations as well as the cytokine pattern of T-cells and regulatory T-cells, and have the capacity to test human cells as well.

Professor Oluf Borbye-Pedersen, University of Copenhagen, is experienced in leading studies in gut microbiota and is involved in numerous projects concerning gut microbiota.

Ethical aspects The study concerning intervention with glutenfree diet has been approved by the regional scientific ethical committee, and the project including the stool samples has been approved in January 2013.

Inclusion Criteria:

• Newly diagnosed with type 1 diabetes (duration < 3 month)

Exclusion Criteria:

• Diabetes duration > 3 months,
• not type 1 diabetes