Can Body Odor Change Sensitivity to Other Smells?

After noticing someone’s body odor, might we become more sensitive to other smells around us? In a study published in 2025, Wunder and colleagues used sweat collected during an anxiety-provoking situation to investigate recipients’ sensitivity to faint odors.

The focus was not on how strongly people perceived the sweat itself, but on detecting another odor smelled afterward. Alongside behavioral experiments, the researchers combined human assessments of the sweat’s smell with chemical analysis of its constituents. The study explored whether bodily chemical cues related to emotion are associated with the way we take in information from our surroundings.

Comparing sweat collected during dental treatment with sweat from the night before

The sweat donors were 12 women with dental anxiety. They collected underarm sweat on pads during treatment and, for comparison, also collected sweat while sleeping the night before. The study called these the anxiety and neutral conditions, respectively.

The effects of this sweat were tested in 36 healthy adults with normal olfaction, comprising 18 women and 18 men. On four separate days, everyone experienced four conditions: anxiety sweat, neutral sweat, a pad without sweat, and a horror film. The order of testing days was randomized to reduce bias.

Sweat was presented as pooled samples containing pieces of pads from four donors, and participants were not told what kind of sample they were receiving. Samples were presented for 3–5 minutes before odor testing and again between the two tests. The horror-film condition was included to compare becoming anxious oneself while watching a film with receiving odors emitted by another person’s body.

Participants detected a rose-like odor at lower concentrations

The test substances were phenylethyl alcohol (PEA), which has a rose-like smell, and n-butanol. Both are used in olfactory research. By varying their concentrations, the researchers measured how dilute an odor could be while still being detected.

For PEA, detection sensitivity was higher after presentation of anxiety sweat than after the sweat-free pad. Mean test scores were 12.3 versus 10.7, a statistically significant difference. In this test, higher scores mean that lower concentrations can be detected. For n-butanol, there were no significant differences among the four conditions.

Importantly, neither substance showed a significant difference between anxiety sweat and neutral sweat. These results alone therefore do not separate an effect specific to anxiety sweat from the effects of body odor more generally.

Questionnaires confirmed increased anxiety in the horror-film condition, but an increase in odor detection sensitivity was not clear. However, the direct difference between anxiety sweat and the horror film was not significant either. A difference appearing in one comparison but not another does not, on its own, establish that the two mechanisms differ.

Participants also rated the intensity and pleasantness of concentrated samples, but anxiety sweat did not consistently make both substances smell stronger or more pleasant. Sensitivity to a faint odor needs to be considered separately from the intensity or appeal of a clearly perceptible smell.

The overall impression of sweat is not the same as its chemical composition

Another part of the study examined the smell and composition of the sweat used in the experiment. This served both to search for candidate substances involved in emotional communication and to establish what participants had been exposed to.

Nine trained assessors described the samples using sweaty, citrus-like, grapefruit-like, soapy, waxy, and cotton-like impressions. There were no statistically significant differences between anxiety and neutral samples in overall odor profile, intensity, or pleasantness. This does not mean that their constituents were identical.

The researchers used a method that separated constituents instrumentally and also checked their odors with the human nose. They then progressively diluted the samples to determine which constituents remained detectable. Dodecanoic acid and a constituent associated with sweaty odor emerged as candidates contributing relatively more to the anxiety sample’s smell, whereas tetradecanoic acid and patchouli alcohol did so in the neutral sample. Some identifications were tentative.

Mass spectrometry of samples from individual donors also provided clues to differences in volatile composition. However, the odor-separation analysis used only one pooled sample per condition. Individual constituents were not presented to recipients to reproduce the change in sensitivity, and no molecule that communicates anxiety was identified.

When smelled as a mixture rather than separately, constituents may strengthen or weaken one another’s odor impressions. The authors proposed such mixture effects, along with whether individual constituents reached perceptible concentrations, as possible reasons why compositional differences appeared despite no difference in the overall impression.

A hypothesis about preparing for danger, and questions that remain

The authors consider the possibility that receiving chemical cues related to another person’s anxiety makes it easier to take in environmental information and could help prepare for threats. The increased sensitivity observed for PEA is consistent with that idea. However, the study did not examine whether danger was detected sooner or escape behavior changed.

Why a difference appeared only for PEA also remains unresolved. Possible explanations include differences in sniffing related to pleasantness, variability in measurements, interactions between sweat constituents and test odors, and adaptation. Mechanisms involving recipients’ hormones or brain activity were not directly tested in this study either.

Because comparison sweat was collected the night before treatment, donors may already have been anxious at that point. Sampling duration and environment also differed between sleep and treatment, and residues from scented products could have influenced composition. Cotton pads may retain some constituents more readily than others, so the sampling method itself requires investigation.

The study was limited to 12 female donors, 36 recipients, and two test odors. The authors propose increasing the numbers of participants and odors and matching factors such as sampling duration. Testing odors associated with fire or spoiled food might address the threat-preparation hypothesis more directly.

Summary

After exposure to sweat collected during dental treatment, recipients detected a faint rose-like odor more readily than in the sweat-free condition. However, there was no difference for the other test odor, and the difference from sweat collected the previous night was not clear.

The study demonstrated an approach for examining chemical cues emitted by other people’s bodies together with sensitivity to surrounding odors. Checking the constituents of the odors actually presented, rather than behavior alone, provides a foundation for linking the relationship between emotion and olfaction to specific molecules and conditions.

References

Wunder, A., Bürkel, N., Guder, I., Zorzin, J. I., Mühle, C., Loos, H. M., & Freiherr, J. (2025). Enhanced sensitivity to odors due to chemosignals associated with anxiety. Communications Chemistry, 8, 129. DOI: 10.1038/s42004-025-01512-3 / Full text. Original article: © The Author(s) 2025, CC BY 4.0.