Diagnosing disease using a patient's breath

How the compounds in our breath might hold the key to a more accurate diagnosis
25 November 2022

Interview with 

Dr Michael Wilde, University of Plymouth


A breath full of disease


A new way to diagnose certain diseases by analysing a patient's exhaled breath has been demonstrated by UK researchers. As he explains to Chris Smith, Plymouth University's Michael Wilde and his colleagues have identified unique combinations of chemicals that are produced in different amounts by different diseases and are released from the lungs alongside the air we breathe out. So if you analyse the composition of that air and measure these chemicals, you can identify the underlying medical problem, even if the patient has more than one thing wrong with them at the time. At the moment it's at the proof of concept stage, but the hope is that the technology could be shrunk to provide a rapid diagnostic system that could screen patients at the hospital doors, or even in their own homes…

Michael - More than one in eight of all emergency admissions to our hospital patients presenting with acute breathlessness and currently the diagnostic markers used to identify the underlying diseases blood tests and radiological procedures such as x-rays. And they have poor discriminatory power in patients with different presentations and also delays in blood sample processing in a triage.

Chris - Because a person could be breathless because they have a chest infection. They could also have heart failure. They could have both.

Michael - Exactly. Despite the same presenting symptom, the underlying causes of acute breathness are highly varied and patients presenting with that symptom will have different disease progressions and treatment options.

Chris - And you think you can improve on what we've already got?

Michael - Yes. So the scope of the study was to develop a new noninvasive method based on breath analysis. And the advantage of a breath test over say a blood test for instance, is not only for those of us who don't like needles because it's non-invasive, it allows for repeated and frequent measurements. So breath is much more readily available than blood and this minimizes the rest of the patient and allows us to monitor those markers or disease a lot quicker.

Chris - How does it work then?

Michael - We know the main constituents of air are carbon dioxide, oxygen, nitrogen, and so most of us would expect exhaled breath to also comprise of slightly altered composition of these main gases. However, it might be surprising to hear that alongside these respiratory gases, our breath also contains hundreds if not thousands of chemicals known as volatile compounds. These volatile compounds are coming from complex chemical reactions happening inside our body. And when we have a disease or a certain condition, this can disturb the levels of these small chemicals in our blood, which then partition to our breath. And so if we can detect these volatile chemicals in breath, we can use them in the non-invasive diagnosis and prognosis of different diseases.

Chris - So different diseases would be represented by a different fingerprint, as it were, of changes in what would be the normal levels of a cluster or constellation of those chemicals?

Michael - Precisely, yes.

Chris - And if you can detect what's changed and by how much, you've got a way of effectively putting your finger on what the diagnosis is, just from sniffing what the person's breath smells like.

Michael - Yeah, precisely.

Chris - Does it work? Is it any good?

Michael - It worked very well. Um, so we were able to identify breath chemicals that differentiate acute cardio exacerbations and the underlying disease subgroups. So we are able to measure and detect about 805 different volatile chemicals across 277 patients. We identified a set of these chemicals that were able to differentiate acute breathlessness, and then within that set of chemicals we also identified smaller subsets, which were able to diagnose acute asthma, chronic obstructive pulmonary disease, known as COPD, community required pneumonia, and acute heart failure.

Chris - How did you find these things in the first place? Because if we envisage an enormous chemical haystack, which is all of the chemicals in the human body, many of which are gonna come out in the breath, how did you find the needles in that haystack that are representative of those different diseases, which are the really sensitive and specific ones that point the finger reliably at the underlying condition each time?

Michael - So that's where I come in as an analytical chemist. To trap the volatile chemicals and breath, we need to pass the breath through a small tube containing a solvent material. So you can think of these tubes as a chemical sponge or a filter. We can send those tubes back to the lab and use an advanced analytical technique called gas chromatography. It's a bit of a mouthful, but most people will have performed chromatography at school. When you place a ink dot on a piece of paper and then run water up for paper, it separates the ink into the different colored pigments. Using this technique, we are able to visualize the separation of hundreds of these volatile chemicals. So we effectively have a molecular lens through which we're able to take a chemical photograph to see the chemicals present in your breath.

Chris - Now obviously you have at your disposal a very good analytical laboratory to do this kind of thing. An A&E department doesn't. So is the idea that having found those needles in the biochemical haystack, you now say, well, we will make small shrunken tests that are just as good on a small scale at finding those particular molecules and that will give us a test?

Michael - Precisely. So in the first instance, we've demonstrated how breath biomarker platforms and how these noninvasive breath tests can be used in an acute care setting. But in the future, now we have an understanding of what these molecules are, what we need to look for, we can start to think about translating these breath signatures, these chemicals onto portable sensors. And then I think most of us could readily envisage in the future a sensor built into a wearable technology or your phone for instance. And you have a constant health status fed back based on the volatile chemicals that you're emitting from your body.


Add a comment