Breath analysis as a non-invasive biomarker for the early detection of preclinical type 1 diabetes. 01/04/2021 - 31/03/2022


Type 1 diabetes mellitus (T1DM) is an auto-immune disorder, causing destruction of the insulin producing β-cells. Often, people with T1DM have auto-antibodies on a predisposing genetic background (HLA genotype). In Belgium, new diagnoses are reported to the Belgian Diabetes Registry (BDR), as they offer free risk-stratification (auto-antibody detection and HLA class) and follow-up of patients and first-degree relatives. Autoantibodies can be present months to years before symptomatic onset, however not all individuals with antibodies develop T1DM. Typically, changes in insulin secretion and glucose tolerance occur months to years after autoantibodies development. So despite evidence that the pathophysiological process begins well before hyperglycaemia is present, treatment with insulin is only initiated once the stage of overt hyperglycaemia. Recently, several studies indicated that overt T1DM can be delayed with immunotherapy. This opens the possibility to treat type 1 diabetes in its 'preclinical' phase. Evidently, this means there is a greater urgency for identification of at-risk individuals. Currently, biomarkers with sufficient sensitivity, specificity and easy applicability are lacking. Identifying subjects at high risk for T1DM can nowadays only be done using invasive methods including repetitive blood samples for antibody profiling and HLA genotyping and an oral glucose tolerance test or clamp. However, repeated blood sampling is an invasive way to ensure population screening, particularly in children. Exhaled breath analysis offers the potential to serve as a non-invasive biomarker screening tool as breath contains volatile organic compounds (VOCs). VOCs arise from endogenous biochemical processes, via physiological or disease-specific mechanisms and enter the lung alveoli by gas exchange. Hence, VOCs can serve as biomarkers of health and disease. Several pilot studies investigated VOCs as surrogate markers of blood glucose and/or as diagnostic tool for detection of hyper- or hypoglycaemic events. These studies all indicated an association between glycaemic state and certain VOCs. None of these studies, however, assessed asymptomatic people with a preclinical phase of T1DM, where autoimmunity is already changed and potentially VOCs are altered before the onset of clinically overt disease. Therefore, the aim of this study is to investigate the novel hypothesis that VOCs can serve as a biomarker to identify individuals at high risk for the development of T1DM. As primary objective, this study wants to elucidate if VOCs of individuals with a high risk of progression to T1DM differ significantly from VOCs of individuals with a low risk of progression to T1DM The study population consists of individuals along the spectrum of T1DM development: 1. Individuals aged 6 through 25 years, with a high risk of progression to T1DM 2. Individuals aged 6 through 25 years, with a recent T1DM diagnosis (< 3 months). 3. Individuals aged 6 through 25 years, with an established T1DM diagnosis (> 1 year). 4. Healthy individuals aged 6 through 25 years, with a low risk of progression to T1DM Risk determination will be assessed in cooperation with prof Bart Keymeulen, member of BDR, and based on both genetic background and the absence or presence of several auto-antibodies. BDR enrolled more than 7000 siblings or offspring of people with T1DM. About 7% of them are persistently antibody positive, and consequently at high risk for the development of T1DM. Patients with recent onset or established type 1 diabetes will be recruited via the UZA diabetologists, who have more than 500 patients in follow-up. Participants will be asked to provide information on personal and familial medical history. Capillary glycemia will be checked to ensure VOCs are collected in a euglycemic state. Collection of a breath sample is an easy process that requires breathing into a mouthpiece that is connected to a ion mobility spectrometer, for 5 minutes.


Research team(s)

Project type(s)

  • Research Project