Genetic and Environmental Risk Factors of Type 1 Autoimmune Diabetes and Its Early Complications (ISIS-DIAB)

Still recognized in youth only at a stage of complete beta-cell mass destruction and insulin deficiency, autoimmune T1D remains a source of major morbidity through daily life and chronic angiopathic complications despite a better glycemic control. T1D onset is now predominantly pediatric, since its incidence shows a rapid and continuous increase in young European children, due to unknown emerging environmental changes, creating a major need for discoveries in the environmental field. Finding avoidable E factors can allow T1D prevention in the whole children population. Lack of infectious exposures ("the hygiene hypothesis"), viruses, early nutrition, or other factors have been suspected, but E causes of T1D remain a black box, as for most human diseases, that should now be approached more systematically and with respect to gene-environment interactions.

The aim of our study is to complete the identification of genetic factors (G) and to undertake the search of environmental factors (E) predisposing to type 1 diabetes (T1D) by constituting a cohort of 3500 T1D patients and a control cohort. We will use the base of G analysis (whole genome genotyping done once a patient using methods conually updated at Centre National de Genotype) and innovative E analysis to develop the following long term objectives :

• Identify G and E factors influencing the process of remaining beta cells' destruction during the first 3 years after diagnosis (subgroup of T1D patients with a 0-3 years diabetes duration);
• Identify G factors (pharmacogenomics) of the individual response to insulin using the effective insulin dose as a phenotype over a period of 2 years (subgroup of T1D patients with negative C-peptide and well managed diabetes);
• Undertake a prospective research of G and E risk factors of "death in bed" syndrome in diabetic adolescents;
• Undertake a prospective research of G and E risk factors of incipient glomerular microangiopathy (regule measurement of microalbuminuria).

We propose to constitute a French multicentric cohort of T1D patients, well phenotyped by 3 data types : genetic, environmental and clinical data. The data collection scheme includes at entry a comprehensive 850 items environmental questionnaire for all subjects and a full genotyping with at least 500,000 SNPs until whole-genome sequencing can be deployed by CNG-CEA. Every 6 months, a standardized clinical assessment is made in patients (a WEB application ensuring this standardization has already been developed). Personal address(es) will be collected and geocoded, then mapped with environmental geographic information systems (GIS).

With environmental modelling, the high dimensionality analyses (HDA) constitutes one of the main originality of ISIS-DIAB approaches of translational research. HDA will face not only a massive mass of data, but data of a remarkable diversity (genomic variants, environmental items from questionnaires, environmental data bases mapped to patient address, space-time items, characteristics of social environment, clinical phenotypes etc). A given genotype (defined by many genomic variants) will predispose to T1D only in a given environmental context (defined possibly by a number of factors) and induce a given type of autoimmune process (age of onset, rate of destruction, biomarkers). Since T1D is both multifactorial and heterogeneous, causal factors may interact in a considerable number of scenarii, thus platforms which study these factors should obviously interact. Without HDA, each platform would be left faced with its own data. The chef d'orchestre has to be HDA, to integrate the massive amounts of data and draw networks of causality. Technological advances in HDA developed in other fields of sciences, business and economics (forecasting technology) will be transferred to biomedical research through ISIS-DIAB. French have a strong tradition of high-level maths in this area. We designed the ISIS-DIAB cohort and collection of data to feed HDA with multidisciplinary data. In ISIS-DIAB program, HDA will identify the variables that have the most predictive value on several outcomes (not confined to T1D causality).