Could Cotton Swab Reveal Disease? Lasers Uncover Clues Hidden in Body Odour

Researchers used lasers to analyse body odour collected with cotton swabs from 24 people, finding patterns associated with Parkinson’s, COVID-19 and mild cognitive impairment. Five wavelengths preserved group separation, but larger studies are needed to establish reliability and diagnostic accuracy.

A cotton swab from the nose, ear or belly button could carry chemical clues that help researchers understand disease, offering a possible route towards tests that are easier on patients. In "Detection of disease-associated VOC signatures with laser-based photoacoustic spectroscopy (LPAS)," published in npj Scientific Reports, Thorsten Graunke, Teresa Scholz, Michal Pieniak and colleagues explored whether laser analysis could reveal differences between samples from healthy people and people with Parkinson's disease, COVID-19 or mild cognitive impairment. Their small laboratory study found promising patterns in body odour, providing an early demonstration of the approach rather than evidence of a test ready for clinical use.

How Body Odour Becomes a Measurable Signal

The body releases tiny airborne chemicals called volatile organic compounds, or VOCs, through processes associated with metabolism, sweat and skin oils. Illness can influence these chemical mixtures, creating patterns that sensitive instruments may detect even when the differences escape the human nose. Finding useful signals is difficult because the compounds occur at low concentrations and form complex mixtures, with some chemicals appearing across several conditions rather than identifying one particular disease.

The researchers used laser-based photoacoustic spectroscopy, a technique that turns the absorption of light into a measurable sound signal. Gas molecules absorb pulses of infrared light, producing tiny pressure changes that an extremely sensitive sensor records. Scanning across different wavelengths produces a chemical fingerprint of the gases released by a swab, allowing researchers to compare overall patterns without identifying every compound in the mixture.

Cotton swabs offer a practical way to collect these odours, with straightforward handling and storage compared with some gas-sampling methods. The team chose the front of the nasal cavity, the navel and the bowl-shaped outer ear because these areas are relatively sheltered from perfume, soap and other personal care products. Such sampling could become useful for non-invasive screening if larger studies confirm that the patterns reliably reflect disease.

What the Swabs Revealed in 24 People

The study involved 24 people, divided into four groups of six: healthy people, people with Parkinson's disease, people with mild cognitive impairment, and people with COVID-19 confirmed by a PCR test. Researchers collected odour samples from each person's nose, belly button and outer ear using sterile cotton swabs, producing 72 samples. Each area was rubbed for about 10 seconds, and the swabs were frozen until testing. Participants with COVID-19 collected their own samples by following detailed instructions to reduce infection risk.

In the laboratory, researchers used three lasers to examine the gases released by the swabs, with each measurement taking about 12 minutes. A clean, unused swab helped them account for background signals that did not come from the participants. They used a statistical method to simplify the measurements and look for patterns, examining samples from each body area separately.

Samples from all three areas showed some differences between the groups, with the nose and belly button giving the clearest results, especially when comparing healthy people with those who had mild cognitive impairment. Some patterns overlapped, including those from people with Parkinson's disease and COVID-19. The findings suggest that body odour may contain clues about disease, but they do not show that the method can reliably diagnose an individual.

Three more Parkinson's samples collected two weeks later showed patterns similar to those in the original Parkinson's group, offering an early sign that the results might be consistent over time. The researchers did not test the same samples repeatedly or carry out this follow-up check for every group, so more testing is needed

Five Wavelengths Could Help Shrink the Technology

Researchers found differences between the groups at 241 of the 730 wavelengths tested, or about one-third, after adjusting their calculations for the many comparisons. No single signal clearly separated all four groups, suggesting that a combination of signals could be more useful than looking for one chemical clue.

Five carefully chosen wavelengths separated the groups almost as well as scanning a much wider range, producing scores of 1.823 and 1.827, respectively. These scores describe how distinct the groups appeared in this study, not how accurately the method could diagnose disease. Using only four wavelengths did not work as well, and the results depended on how researchers selected and processed the measurements.

Using fewer wavelengths could help researchers develop smaller, simpler screening devices, although this study did not build or test a portable device for patients. Studying a mixture of signals may also be useful because the same chemical can appear in several health conditions, making it difficult to identify a disease from one chemical alone.

Why a Useful Screening Test Needs More Evidence

Six participants per group is a very small foundation for a diagnostic tool, and the study explored patterns without developing or validating a classification model. It did not establish how accurately the method would identify disease in new patients or how often it would produce false alarms. Selecting the most effective wavelengths and processing settings within such a small dataset makes independent testing particularly important.

Age remains a major unresolved influence because it was associated with the statistical patterns, and Parkinson's disease and mild cognitive impairment tend to affect older people. The researchers could not clearly separate age-related effects from disease-related effects in this cohort. Gender, height, weight and body mass index showed no consistent associations with the patterns, but the small sample cannot establish that these influences are absent.

Larger studies involving more diverse participants, better age matching and everyday clinical conditions will need to examine environmental chemicals, personal hygiene and individual differences. The present findings show that ordinary cotton swabs can capture measurable odour patterns associated with different health conditions under controlled conditions. Turning those patterns into a reliable screening tool will depend on demonstrating that they remain useful beyond this small laboratory study.