Back

Biomarkers of the Microbiome-Skin-Brain Axis in Stress and Depression: Fingerprinting of Highly Volatile Compounds in Axillary Sweat via Gas Chromatography-Ion Mobility Spectrometry

Tungkijanansin, N.; Kulsing, C.; Tunvirachaisakul, C.; Sriswasdi, S.; Kerr, S. J.; Hanvoravongchai, J.; Thewaran, N.; Sirinara, P.; Maes, M.

2025-05-07 psychiatry and clinical psychology
10.1101/2025.05.06.25327121 medRxiv
Show abstract

BackgroundDifficulty in the diagnosis of high stress and depression has been recognized conventionally depending on the observation of patient symptoms and psychiatrist diagnosis. These approaches are time-consuming and cannot respond to the excessive demands for large-scale tests with the increasing populations worldwide. This study thus developed an alternative approach to perform fast stress screening, which is based on fingerprinting of highly volatile compounds in axillary sweat. MethodsSweat samples were collected from 227 firefighters, comprising 65 with high stress, 14 with depression, and 148 healthy volunteers. High stress and depression were determined using the standardized Thai versions of the Perceived Stress Scale (PSS-10) and the Beck Depression Inventory II (BDI-II), in conjunction with psychiatric interviews. The samples were collected by placing cotton rods under the axillaries, then analyzed using gas chromatography- ion mobility spectrometry (GC-IMS). The potential marker peaks were selected based on accuracy data. Principal component analysis (PCA) and logistic regression with machine learning were also performed to select significant composite markers. MVOC 3.0, Amibase and Metaboanalyst 6.0 databases were applied to predict the possible metabolomic pathways. ResultsAnalysis against genuine standard compound injections identified acetonitrile, ammonia, diethyl ether, formaldehyde, and octane as potential biomarkers for both high stress and depression, with butane, dimethylamine and pentane additionally observed for high stress. Receiver operating characteristic (ROC) curves demonstrated accuracies of 81.3% for stress screening and 82.8% for depression screening. ConclusionThe biomarkers delineated here indicate the participation of particular metabolic pathways and commensal skin bacteria in the stress response.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.