Breath Acetone Measurement-Based Prediction of Exercise-Induced Energy and Substrate Expenditure

The purpose of our study was to validate a newly developed breath acetone (BrAce) analyzer, and to explore if BrAce could predict aerobic exercise-related substrate use. Six healthy men ran on a treadmill at 70% of maximal oxygen consumption (VO(2max)) for 1 h after two days of a low-carbohydrate diet. BrAce and blood ketone (acetoacetate (ACAC), beta-hydroxybutyrate (BOHB)) levels were measured at baseline and at different time points of post-exercise. BrAce values were validated against blood ketones and respiratory exchange ratio (RER). Our results showed that BrAce was moderately correlated with BOHB (r = 0.68, p < 0.01), ACAC (r = 0.37, p < 0.01) and blood ketone (r = 0.60, p < 0.01), suggesting that BrAce reflect blood ketone levels, which increase when fat is oxidized. Furthermore, BrAce also negatively correlated with RER (r = 0.67, p < 0.01). In our multiple regression analyses, we found that when BMI and VO(2max) were added to the prediction model in addition to BrAce, R(2) values increased up to 0.972 at rest and 0.917 at 1 h after exercise. In conclusion, BrAce level measurements of our BrAce analyzer reflect blood ketone levels and the device could potentially predict fat oxidation.