Speedy Breath Test Detects Infections in Minutes

Our breath contains hundreds of compounds produced by metabolic processes, and not all of them come from our own cells. Researchers are increasingly exploring whether these chemical traces can serve as indicators of disease.

The original inspiration

Previous studies of bacterial metabolism have shown that certain labeled sugars and sugar alcohols are readily metabolized by bacteria but largely ignored by human cells. When paired with imaging techniques such as positron emission tomography (PET), these compounds can act as tracers that help detect infections and track how they respond to treatment. However, imaging methods require specialized equipment and are not always practical in time‑sensitive clinical settings.

Can the same idea work for respiration?

Motivated by these metabolism‑targeted tracers, a collaborative research team in the United States and Belgium explored whether a similar strategy could be adapted into a noninvasive breath test. In work published in ACS Central Science, the researchers report a prototype approach that uses carbon‑13 (¹³C), a stable, nonradioactive isotope of carbon, to detect bacterial infections through exhaled breath.

The concept builds on the long‑standing ¹³C urea breath test used to diagnose Helicobacter pylori stomach infections. In the new study, the researchers administered ¹³C-enriched metabolites that human cells metabolize poorly but bacteria readily break down. If bacteria are present, this metabolism produces ¹³C-labeled carbon dioxide ([¹³C]CO₂) that can be detected in exhaled breath using portable, relatively inexpensive spectroscopy tools.

The specific metabolites tested included urea, l-arabinose, d-sorbitol, d-xylose, maltose, maltotriose, and d-mannitol substrates. In laboratory experiments, several of these compounds were efficiently converted to [¹³C]CO₂ by multiple bacterial species, while separate in vivo studies showed minimal [¹³C]CO₂ production in uninfected mice.

When tested in animal models, the approach successfully detected several types of bacterial infection, including pneumonia, bloodstream infections, muscle infections, and bone infections. Breath signals typically appeared within the first several minutes after administration, distinguishing infected animals from uninfected controls.

In one model of Escherichia coli myositis, ¹³C signals decreased as antibiotic treatment reduced bacterial burden, suggesting the method could also help monitor how infections respond to therapy. In additional experiments, breath‑based measurements correlated with the behavior of metabolism‑targeted PET tracers in certain bacterial strains, pointing to potential synergy between breath testing and imaging approaches.

Looking ahead

The authors emphasize that this work represents an early, preclinical step. Further studies will be needed to determine which metabolites perform best in humans, how background signals vary across patients, and how the method compares with existing diagnostic tools. Still, because the tested metabolites have been administered safely in humans and the detection instruments are portable, the researchers suggest that breath‑based ¹³C testing could one day complement current strategies for identifying invasive bacterial infections.

"In designing this study, we were motivated by a developing trend in clinical practice, whereby patients and providers want answers right away that will inform treatment decisions,” notes David Wilson, a corresponding author of the study, in a recent ACS press release.

As scientists continue to explore what our breath can reveal about health and disease, metabolism‑based detection may offer a promising route toward faster, more targeted diagnosis.

Explore related articles in ACS journals:

Ultrahigh Flow Regulated Discharge Coupling with Targeted Mass Spectrometry Analysis: Real Time Capturing Biomolecules in Exhaled Breath
Wenzhao Zhou*, Wenting Wang, Mengmeng Jiang, Lanrui Cao, Shichun Shao, Yanna Le*, and Xudong Fu
DOI: 10.1021/acs.analchem.5c00195

A Pt/Pd-Decorated SnO₂-Based Electronic Nose for High-Precision Breath Analysis in Diagnosing SIBO
Xinxin He, Ping Guo, Yinhua Hu, Xuyang An, Shuai Liang, Tiezhu Liu*, Tong Zhang, Guohua Cai*, Yanjie Chu*, and Jia Zhang*
DOI: 10.1021/acssensors.6c00086

Early Screening and Subtype Identification of High-Risk Lung Nodules via Breathprint by Graphene eNose Platform: A Large Cohort Study
Xingyu Zhu, Qiaofen Chen, Jiajing Sun, Lichen Zhang, Zhengwei Huang, Jingwei Xu, Haichuan Hu, Yuqi He, Zhao Chen, Xiaogang Ye, Xueyin Chen, Aotian Guo, Sheng Lu, Tao Shen, Jianmin Wu*, and Zhengfu He*
DOI: 10.1021/acssensors.5c00314

Measurement of Oro-Cecal Transit Time in LPS-Treated Pigs Fed High and Low Fiber Diets Using the Lactose-¹³C-Ureide Test in Breath and Saliva Samples
Mariagrazia Cavalleri, Miriama Sciascia, Solvig Görs, Andreas Vernunft, Henry Reyer, Klaus Wimmers, Jürgen Zentek, Jeannette Kluess, Sven Dänicke, and Cornelia C. Metges*
DOI: 10.1021/acs.jafc.5c00534

Rh and Ir-Doped PtS₂ Monolayers as Promising Sensors for Liver Disease Biomarker Detection in Exhaled Breath: A First-Principles Study
Yungeng Liu, Xiulin Xiao*, and Huihui Xiong
DOI: 10.1021/acsomega.5c06170

High-Precision Synchronous Detection of Breath CO, CH₄, and CO₂ Using an NIR-WMS Sensor Based on a Dual-Channel Multipass Cell
Xue Ou, Hongjiang Dong, Yiyun Gai, Zhaoyue Huang, Guotao Zhang, Peng Liu, and Xin Zhou*
DOI: 10.1021/acs.analchem.5c08129