Exhaled breath acetone in predicting the presence and severity of respiratory failure

Respiratory failure (RF) has a high mortality rate and poor prognosis, making the development of novel non-invasive biomarkers crucial. Hypoxia promotes lipolysis, increasing free fatty acid (FFA) and ketones. Exhaled breath acetone (EBA), a volatile component of ketone bodies, may be linked to the presence and severity of RF. In this study, 156 patients were enrolled and categorized based on arterial blood gas analysis into RF group (<i>N</i>= 74) and control group (<i>N</i>= 82). The EBA was compared between the two groups. RF patients were classified by PaO₂/FiO₂(P/F): high P/F (200 ⩽ P/F < 300 mmHg;<i>N</i>= 42) and low P/F (P/F < 200 mmHg;<i>N</i>= 32), and subsequently EBA was compared. Multivariate and multiple-model logistic regression analyses were employed to investigate the impacts of EBA on the RF. Additionally, receiver operator characteristic curve was utilized to evaluate the diagnostic efficacy of EBA. The RF group presented a significantly higher EBA [1.61 (0.98-2.57) vs 1.24 (0.86-1.69) ppm,<i>P</i>= 0.001], compared to the control group. The EBA within the low P/F group was higher than within the high P/F group [2.43 (1.57-3.23) vs 1.37 (0.91-1.83) ppm,<i>P</i>< 0.001]. EBA was conspicuously negatively correlated with PaO₂/FiO₂, and positively correlated with beta-hydroxybutyrate and FFA. Logistic regression analyses demonstrated that EBA was correlated with the presence and severity of RF. The area under curve of EBA in the diagnosis of RF and low P/F were 0.651 (95% CI: 0.564-0.738,<i>P</i>= 0.001) and 0.763 (95% CI: 0.652-0.875,<i>P</i>< 0.001). EBA can serve as a valuable predictor for the presence and severity of RF.