Blood ammonia is routinely used in clinical settings to assess systemic ammonia in hepatic encephalopathy and urea cycle disorders. Despite its drawbacks, blood measurement is often used as a comparator in breath studies because it is a standard clinical test. We sought to evaluate sources of measurement error and potential clinical utility of breath ammonia compared to blood ammonia.
We measured breath ammonia in real time by quartz enhanced photoacoustic spectrometry and blood ammonia in 10 healthy and 10 cirrhotic participants. Each participant contributed 5 breath samples and blood for ammonia measurement within 1 h. We calculated the coefficient of variation (CV) for 5 breath ammonia values, reported medians of healthy and cirrhotic participants, and used scatterplots to display breath and blood ammonia.
For healthy participants, mean age was 22 years (±4), 70% were men, and body mass index (BMI) was 27 (±5). For cirrhotic participants, mean age was 61 years (±8), 60% were men, and BMI was 31 (±7). Median blood ammonia for healthy participants was within normal range, 10 μmol L−1 (interquartile range (IQR), 3–18) versus 46 μmol L−1 (IQR, 23–66) for cirrhotic participants. Median breath ammonia was 379 pmol mL−1 CO2 (IQR, 265–765) for healthy versus 350 pmol mL−1 CO2 (IQR, 180–1013) for cirrhotic participants. CV was 17 ± 6%.
There remains an important unmet need in the evaluation of systemic ammonia, and breath measurement continues to demonstrate promise to fulfill this need. Given the many differences between breath and blood ammonia measurement, we examined biological explanations for our findings in healthy and cirrhotic participants. We conclude that based upon these preliminary data breath may offer clinically important information this is not provided by blood ammonia.