Do More Odor Precursors Mean Stronger Underarm Odor? The Role of Bacteria

Does underarm odor become stronger simply because sweat contains more of the materials from which odors are made? In a study of 24 people, the concentrations of the measured precursors showed no clear correspondence with the intensity of the odor people actually smelled. The relative proportions of different underarm bacteria, however, showed patterns associated with odor.

A distinctive feature of this 2015 study was its use of DNA collected from skin to examine bacterial communities, rather than relying solely on growing bacteria in culture. Combining sweat chemistry, bacterial quantity and composition, and human odor assessments, it explored the conditions in which body odor develops.

Examining 24 people’s underarms in the morning and afternoon

The participants were white men and women aged 25–50 living around Geneva, Switzerland. Of 26 people recruited, 24 completed the full schedule and were included in the analysis. Thirteen used antiperspirants and 11 did not.

Participants used the same shower gel, showered each morning on study days, and wore cotton T-shirts. Restrictions were also placed on perfume use and underarm shaving. On each of the first two days, they attended a morning session and an afternoon session about six hours later, making four measurement sessions in total.

Four trained assessors rated odor from both underarms on a 0–10 scale. In addition to overall intensity, they assessed characteristics such as sulfurous, acid-spicy, fatty, and onion-like notes. Skin pH was measured on the left underarm, while bacterial DNA was measured from swabs of the right.

On the third day, underarm sweat produced during 10 minutes in a sauna was collected and analyzed for four precursors of acid-related and sulfur-related odorants. It is worth noting that precursor measurements were not taken at the same time as the odor and bacterial assessments.

Odor and bacterial load differed with antiperspirant-use status

The median overall sweat odor score was 3.11 among nonusers and 1.57 among antiperspirant users. Bacterial load estimated by quantitative PCR of bacterial DNA was also about 50 times higher on average among nonusers. Skin pH was lower among users.

These were observational comparisons between existing users and nonusers. The study did not randomly assign the same people to conditions with and without antiperspirant. The result therefore cannot be reinterpreted as a product effect showing that anyone who uses antiperspirant will reduce their bacteria to one-fiftieth of the original level.

Overall odor and sulfurous and acid-spicy notes generally became stronger from morning to afternoon. Among nonusers, men also tended to have stronger overall odor and some odor characteristics, but differences were not significant for every measure or time point.

Bacteria in the same broad group can have different relationships with odor

Detailed DNA sequencing of bacterial composition was carried out on 44 samples from the 11 nonusers. Samples from antiperspirant users contained too little bacterial DNA for adequate analysis. This part of the study should therefore be distinguished from an equally detailed microbiome comparison of all 24 participants.

An average of 96% of the analyzed sequences belonged to three genera: Staphylococcus, Corynebacterium, and the group called Propionibacterium in the original paper. Even so, community composition differed between individuals, and these between-person differences were greater than changes within the same person over about 30 hours.

The relative abundance of Corynebacterium was positively correlated with overall odor intensity. The relative abundance of Propionibacterium, in contrast, showed a negative correlation. When Staphylococcus was considered as a whole genus, no clear correspondence with odor emerged.

Differences appeared, however, when bacteria with similar DNA sequences were divided into smaller units called OTUs. Several OTUs assigned to Staphylococcus hominis were positively correlated with onion-like or acid-spicy odor and other characteristics. Some OTUs assigned to S. epidermidis, in the same genus, showed weak negative correlations.

An OTU assigned to Corynebacterium tuberculostearicum was associated with overall odor and the onion-like characteristic, while an Anaerococcus OTU was associated with the fatty characteristic. Broad statements that all bacteria in a particular genus intensify odor can hide these differences.

Precursor quantity alone did not explain odor intensity

Sweat precursor concentrations varied substantially between individuals. Men tended to have higher ratios of acid-related to sulfur-related precursors, but there was no clear direct correlation between the measured precursor concentrations and odor intensity.

This supports an interpretation in which the presence of odor-forming materials is only part of the picture: the quantity and combination of bacteria able to transform them also matter. Alongside relative abundance, the researchers analyzed an abundance measure incorporating quantitative PCR information and found broadly similar associations.

Some bacterial groups increased together, whereas others showed an inverse relationship. The authors considered possible cooperation or competition. However, relative-abundance data are constrained by their fixed total, and correlations alone cannot establish interactions between bacteria.

An exploratory map for choosing bacteria to investigate

Only 11 people were included in the detailed microbiome analysis, and no statistical correction was applied for the many comparisons. The four samples were also repeated samples from the same individuals. These associations are candidates for testing in larger populations and in experiments that culture bacteria and examine their functions directly.

The authors propose possibilities for odor control that focus on microbiome composition, such as encouraging bacteria that compete with odor-producing organisms. This is a direction for future research, not a trial demonstrating the deodorizing effects or safety of adding bacteria to skin. A negative correlation with odor does not necessarily mean that a bacterium itself has a deodorizing action.

Summary

In this study, measured sweat precursor concentrations alone did not explain the intensity of underarm odor. In contrast, higher proportions of Corynebacterium and some S. hominis OTUs were associated with stronger odor, while higher proportions of Propionibacterium were associated with weaker odor.

Nonusers of antiperspirants had stronger odor and higher estimated bacterial loads than users, and detailed microbiome profiles showed individual characteristics. Understanding body odor requires looking together at sweat’s raw materials, total bacterial load, and the finer identities of the bacteria. This study used culture-independent methods to provide clues to that combined picture.

The study was conducted at Firmenich, and the paper lists four authors as affiliated with the company. No separate funding section was identified; the authors declared no competing interests.

References

Troccaz M, Gaïa N, Beccucci S, et al. Mapping axillary microbiota responsible for body odours using a culture-independent approach. Microbiome. 2015;3:3. doi:10.1186/s40168-014-0064-3