Clinical applicability of breath measurements for detecting disturbances in systemic glucose metabolism: a systematised review

Breath analysis is a candidate non-invasive approach for detecting disturbances in glucose metabolism, but its clinical applicability has not been systematically assessed across analytical platforms and clinical targets. This systematised review evaluates breath-based biomarkers and analytical technologies for the detection of diabetes, prediabetes, insulin resistance, gestational diabetes, and hypoglycaemia, and for blood glucose estimation. A structured PubMed search identified human studies published between 2010 and 2025; 30 studies met the inclusion criteria. Most analysed volatile organic compounds using laboratory platforms such as gas chromatography-mass spectrometry or proton transfer reaction mass spectrometry, a growing number used portable sensor-based systems including electronic noses and single-gas sensors, and a smaller group used isotope-based breath tests measuring labelled carbon dioxide after substrate ingestion. Reported diagnostic performance was highest and most consistent for hypoglycaemia detection in type 1 diabetes mellitus (sensitivity and specificity approximately 90 to 95%) and for isotope-based discrimination of prediabetes and diabetes subtypes (approximately 90 to 100%), whereas qualitative diabetes detection and quantitative blood glucose prediction were more variable. Both findings are based on a small number of studies, and the isotope-based results originate from a single research group and require independent replication. Performance depended strongly on signal complexity, with multivariable and spectral approaches outperforming single-compound methods. However, most studies used small cohorts and internal validation only, and sampling procedures and reporting were poorly standardised. Standardised protocols, larger externally validated cohorts, and platform selection matched to the intended clinical role are required before implementation.