Volatile compounds in blood headspace and nasal breath.

Breath analysis is a form of metabolomics that utilises the identification and quantification of volatile chemicals to provide information about physiological or pathological processes occurring within the body. An inherent assumption of such analyses is that the concentration of the exhaled gases correlates with the concentration of the same gas in the tissue of interest. In this study we have investigated this assumption by quantifying some volatile compounds in peripheral venous blood headspace, and in nasal breath collected in Tedlar bags obtained at the same time from 30 healthy volunteers, prior to analysis by selected ion flow tube mass spectrometry. Some endogenous compounds were significantly correlated between blood headspace and nasal breath, such as isoprene (r p  = 0.63) and acetone (r p  = 0.68), however many, such as propanol (r p  = -0.26) and methanol (r p  = 0.23), were not. Furthermore, the relative concentrations of volatiles in blood and breath varied markedly between compounds, with some, such as isoprene and acetone, having similar concentrations in each, while others, such as acetic acid, ammonia and methanol, being significantly more abundant in breath, and others, such as methanal, being detectable only in breath. We also observed that breath propanol and acetic acid concentrations were higher in male compared to female participants, and that the blood headspace methanol concentration was negatively correlated to body mass index. No relationship between volatile concentrations and age was observed. Our data suggest that breath concentrations of volatiles do not necessarily give information about the same compound in the blood stream. This is likely due to the upper airway contributing compounds over and above that originating in the circulation. An investigation of the relationship between breath volatile concentrations and that in the tissue(s) of interest should therefore become a routine part of the development process of breath-based biomarkers.