A study of volatile compounds in the breath of children with type 1 diabetes

TL;DR

This pilot study analyzed exhaled volatile compounds in children with type 1 diabetes to explore potential correlations with blood glucose levels, finding no consistent pattern across individuals or compounds.

Contribution

It provides a comprehensive analysis of breath VOCs in children with type 1 diabetes using GC-MS and SIFT-MS, highlighting the lack of strong correlation with blood glucose levels.

Findings

74 volatile compounds identified in patient samples

No significant overall correlation between VOCs and blood glucose levels

Some individual correlations observed but no consistent pattern

Abstract

A pilot study of exhaled volatile compounds and their correlation with blood glucose levels in eight children with type 1 diabetes is reported. Five paired blood and breath samples were obtained from each child over a 6 hour period. The blood glucose concentration ranged from 41.4 to 435.6 mg/dL. Breath samples were collected in Tedlar bags and immediately evacuated through thermal desorption tubes packed with Carbopack B and C. The VOCs were later recovered by thermal desorption and analysed using gas chromatography mass spectrometry. The study identified 74 volatile compounds present in at least 10% of the patient samples. Of these 74 volatiles 36 were found in all patient samples tested. Further analysis of the 36 compounds found that none showed significant overall correlation with blood glucose levels. Isoprene showed a weak negative correlation with blood glucose levels. Acetone was found to have no correlation with blood glucose levels for the patients studied. Some patients showed significant individual correlation between the relative peak areas of certain compounds and blood glucose levels. However, there was no consistent pattern observed within these results across all 8 patients. Additional breath samples were collected in Tedlar bags and analysed using SIFTMS for 3 of the patients and a healthy control. The levels of 24 volatiles are reported and were found to be generally consistent with previously reported SIFT-MS data. In agreement with the GCMS data, no compounds exhibited a significant overall correlation with blood glucose level.