Type 1 diabetes occurs when the body's immune system turns against itself so that, in a very specific and targeted way, it destroys the pancreatic islet β-cells, the only cells in the body that produce the vital hormone insulin. This autoimmune destruction is irreversible and the disease incurable. If new pancreas or islets are transplanted they too are destroyed, unless heavy immunosuppression is applied. Recent genetics and mechanistic studies in humans are indicating major pathways that play roles in early disease development, especially in the interleukin-2 (IL-2) and type 1 interferon systems, probably in concert with the immune response to preproinsulin [¹,²].

Longitudinal studies of children born with high genetic risk of Type 1 diabetes showed that circulating autoantibodies to islet antigens can be seen as early as 6 months. A large proportion of childhood cases has seroconverted by age 3 years, and almost all cases by age 7 years, even although median age-at-diagnosis is approximately 12 years [³,⁴]. Children who will go on to develop Type 1 diabetes in the future must be born with the first precursors of the disease. A child aged under age 5 years with a family history of Type 1 diabetes, carrying the highest risk human leukocyte antigen (HLA) class II genotypes and persistently positive for two or more autoantibody types, has a more than 90% chance of being diagnosed with the disease [⁵]. By making the connection between gene variant and phenotype, genetics provides a way to identify pathways leading to Type 1 diabetes, including the earliest physiological events after birth, and the earliest events in the disease process, including clues about what may be environmental factors or accelerators of disease [⁴,⁶].

National diabetes registries, especially those with available biosamples, provide Europe with important population-based research resources, for studying disease epidemiology, or for testing hypotheses about disease initiation and progression. Registry data where available provide evidence that the incidence of Type 1 diabetes in children under age 15 years has increased. However, there is a shortage of data on Type 1 diabetes over age 15, although at least half of new cases occur after age 15. Data from registries should be accessible to all bona fide researchers, provided that necessary confidentiality protections are in place. Common standards regarding data fields would add value and also aid data access.

The appearance of autoantibodies, the first of which chronologically are usually against insulin itself, is a secondary or downstream event in the autoimmune process. Identification and validation of biomarkers that precede autoantibody appearance are urgently required. Incentives to provide access to these collections should be implemented [⁷]. Useful information to provide possible investigators with includes limitations of the informed consent, types and amounts of material available and storage conditions. Efforts should be made to extend collections of samples of at-risk individuals and persons with newly diagnosed diabetes. Funding opportunities should be implemented to encourage investigators to develop assays that could be used retrospectively, to employ new technologies and to use smaller assay volumes for studying markers.

Despite the limited success of previous prevention trials, they have provided valuable lessons, combined with correctly interpreted data from the animal models. Predictions of the natural history of disease based on combined genetic (HLA) and serological markers (autoantibodies) are accurate. C-peptide is a surrogate marker for loss of β-cell function and progression to diabetes; and the oral glucose tolerance test is a sensitive screening tool for detecting individuals with diabetes before fasting blood sugar or HbA_1c levels become abnormal. Screening for risk of diabetes can avoid suffering in children, as diabetic ketoacidosis prevalence was 4% in children enrolled in prevention trials compared with 15–20% in new-onset children in the general population [⁸]. Interventions at later stages of disease, such as in newly diagnosed Type 1 diabetes, have used mainly non-antigen-specific drugs, with the promising studies having used doses at the edge of safety. Successfully preventing the onset of autoimmunity—true primary prevention— must include antigen-specific tolerance approaches [²]. It is important to begin now to accurately establish doses, timing and best route of administration, whilst analysing mechanistic effects.