Physicians have known for centuries that unusual breath odors may signify disease. The complex chemistry of a healthy human body produces a myriad of compounds, many of which readily equilibrate between blood and inspired air in the lungs. Some disease states alter that chemistry and leave a clear signature in breath composition. People suffering from uncontrolled diabetes, for example, may have a sweet, fruity breath odor due to acetone, while poor kidney function produces the sour smell of amines.1
Although breath analysis is perhaps the least invasive of all diagnostic methods, it is neither accepted nor commonly used in diagnosis.2 The substantial difficulty is that most breath components are present in exceedingly low concentrations, from parts per billion to parts per trillion by volume.
Acetone concentration is an exception, however. As a normal metabolite, its presence in breath levels ranges from a relatively high 0.5ppmv for healthy individuals to hundreds of ppmv for critically ill, ketoacidotic diabetics. Acetone is one of three principal ketone bodies produced endogenously from the breakdown of fatty acids. Because it readily equilibrates throughout the body and can even be detected emanating from the skin,3 breath measurements represent a uniquely nonintrusive probe. Increased acetone production can be a normal metabolic response in healthy individuals due to physical exercise. They can also represent the caloric imbalances frequent with low-carbohydrate diets or indicate the presence of diseases such as diabetes.4
Southwest Sciences has recently demonstrated a diode laser- based acetone measurement method capable of quantifying acetone at concentrations typically present in human breath.5,6 Our measurement method takes advantage of a gas– solid chemical reaction of acetone with hydroxylamine hydrochloride (HA) to quantitatively produce hydrogen chloride gas (HCl). The reaction is depicted in Figure 1. The advantage of our method is that HCl is a diatomic molecule with simple spectroscopy and strong optical absorptions accessible with commercially available near IR diode lasers. Breath acetone concentrations can be readily calculated from measured HCl and a predetermined acetone-to- HCl conversion efficiency for the HA reaction. Typical conversion efficiencies are greater than 20%, corresponding to acetone detection limits on the order of 100ppbv.
Figure 1. Acetone detection method, chemical conversion to hydrogen chloride (HCl) gas.
Our acetone measurement technique has enough sensitivity to measure normal levels of breath acetone, and we have demonstrated this with a small, compact sensor platform (see Figure 2). Breath measurements are made by first collecting the latter part of an exhalation in a Tedlar bag. A small diaphragm pump forces the breath sample from the bag through a calcium chloride desiccant column to remove water. The dried sample then flows through a reactor tube containing powdered hydroxylamine hydrochloride (HA). HCl gas released from the acetone– HA reaction passes from the reactor to an optical cell where HCl concentration is determined. The volume of gas from a single exhalation is more than sufficient for this purpose.
Figure 2. Prototype breath acetone sensor. The hydroxylamine hydrochloride (HA) reactor is in the foreground, hydrogen chloride gas (HCl) optical detection cell, and control electronics at the rear of the photograph.
Shown in Figure 3 are breath acetone measurements for ‘Steve’ and ‘Andre’ performed with our prototype sensor. The measurements were made using a baseline nitrogen gas flow to illustrate an important aspect of the method. HA materials are slightly reactive to oxygen, as indicated by the response with acetone removed. Such reactivity can be treated as a fixed offset with a small added error and breath measurements made against an air background. Results indicate acetone concentrations of 2.5ppm and 0.7ppm. Notably, the author (‘Steve’) had fasted for 17 hours and recorded a slightly high breath acetone value. When the sensor is recently calibrated and has been optimized properly, acetone sensitivity for breath measurements is conservatively estimated at several tenths ppmv, and it is appropriate for breath acetone measurements of healthy, metabolically stressed, and diseased individuals.
Figure 3. The sensor response over time for two breath acetone measurements (labeled ‘Steve’ and ‘Andre’). During the course of measurement, the gas flow through the HA reactor changed from nitrogen (A), breath with acetone removed (B), and breath dried with calcium chloride alone (C).
Over the short term, we plan to focus on issues such as improved gas flow and temperature regulation of the HA reactor, and increasing calibration lifetime for the breath acetone sensor. The latter issue will be addressed by investigating oxygen reactivity and stability of other HA materials. We anticipate that this technology will enable the medical community to evaluate breath acetone levels for diabetes screening and maintenance, to monitor calorie restriction weight-loss diets, and for use in exercise physiology.
This work was supported by the US Army Research Institute of Environmental Medicine, Natick MA, SBIR contract #DAMD17-03- C-0032.
Southwest Sciences, Inc.
Santa Fe, NM
Steve Massick is a Senior Research Scientist at Southwest Sciences and the principle investigator of the breath acetone work. He received his PhD in Physical Chemistry at the University of Utah in 1996 for spectroscopic studies of neutral metal atom-rare gas van der Waals complexes. From 1997 to 2001 he was a postdoctoral research associate at the University of California, Santa Barbara where he studied solution phase homogeneous catalysts using time-resolved infrared spectroscopy.