Will You Still Like the Same Smell Three Months Later?

If you find someone’s smell pleasant, does that preference persist for months? In a 2013 study, women’s preference rankings for men’s body odors showed some consistency about three months later. Preferences for the fragrance materials used for comparison agreed more strongly across sessions than preferences for body odor.

The study’s central question was whether evaluations of people through body odor are reproducible over time. Assessments of fragrance materials and facial photographs provided comparisons for the stability of odor preferences.

Collecting odors from the same men again three months later

The raters were 63 heterosexual women aged 18–32 from a UK university. Missing data meant that analyses included 59–63 women depending on the task. Women who were not using oral contraceptives were assessed twice on days 10–14 after menstruation began, with the second session scheduled about three months later.

Odor donors were selected from 92 men aged 18–31, with each woman evaluating six men. The men wore cotton T-shirts for two nights and avoided scented products, strongly odorous foods, and similar exposures from before collection. Each woman’s set included three men with similar MHC genes, a group of genes involved in immunity, and three with dissimilar MHC genes. This sampling arrangement was shared with another, larger study.

Importantly, the T-shirts smelled at the second session were not stored samples from the first. The same men wore new shirts about three months later. This allowed the researchers to examine how well preferences for an individual’s odor persisted while including possible day-to-day changes in body odor.

The women smelled T-shirts presented in jars and used seven-point scales to rate pleasantness and how desirable the odor would be in a long-term partner. They also arranged the six jars in order of preference. Presentation order was reversed in the second session, and the women were told that the samples came from the same men.

Body-odor preferences showed moderate consistency

For each woman, the researchers examined how closely the ordering of evaluations of the six men agreed between sessions. A rank correlation close to 1 indicates strong agreement in order, while a value near 0 indicates no consistent correspondence.

Median correlations were 0.35 for body-odor pleasantness, 0.18 for desirability as a long-term partner’s odor, and 0.34 for directly ranking the jars. All were significantly above zero at the group level. Rather than preferences changing completely each time, there remained a tendency to rate relatively favored men’s odors more highly again later.

Individual correlations nevertheless ranged from negative values to strongly positive ones. The finding does not mean everyone retained the same rankings. Rather, there was overall repeatability despite some individuals’ preferences changing.

Rankings for fragrances and faces were more stable than for body odor

The fragrance task compared six materials: bergamot, lime, lemongrass, myrrh, camphor, and clove. The first six of 15 evaluated materials were selected to match the number of body odors and faces. The women were asked how much they liked each as an ingredient in their own perfume.

The median rank correlation for fragrance ratings was 0.77, significantly higher than for body-odor pleasantness or desirability. At the first session, bergamot had the highest mean rating and myrrh the lowest. However, this was a comparison of a small number of materials using this particular presentation method.

The women also evaluated photographs of the same six men’s faces. Identical photographs were used in both sessions. Median correlations were 0.67 for attractiveness in a short-term relationship and 0.60 for attractiveness in a long-term relationship, both showing greater consistency than the corresponding body-odor evaluations. The difference in stability between fragrances and facial photographs was not statistically clear.

Stable preferences in one domain do not necessarily extend to another

Another finding was that individual differences in stability did not agree across domains. Women with stable body-odor preferences were not found to be particularly stable in their preferences for fragrances or facial photographs.

However, stability in pleasantness and desirability ratings of the same body odors was related, as was stability in short- and long-term facial-attractiveness ratings. This makes it difficult to explain the results through a single personal trait of always having consistent preferences. The authors discuss possible differences in interest or attention between domains.

Separating changes in the odor from changes in preference

Body-odor samples were collected anew, whereas the faces were identical photographs and the fragrance materials were the same types. Thus, the lower consistency for body odor could reflect changes in men’s odors due to diet or lifestyle as well as changes in women’s preferences. It is also possible that large differences between materials such as bergamot and myrrh made their rankings easier to retain.

Because evaluations were repeatable even for newly collected body odors, the authors conclude that some features of an individual’s odor persist over time and may be meaningful as cues in social evaluations of others. This is an interpretation based on perceptual assessments; the study did not measure body-odor chemistry or its genetic causes.

The study also concerned young women and men, with women’s menstrual cycles aligned by day count. Ovulation testing was not performed. The authors discuss the stability of fragrance ratings as potentially relevant to continued use of a favorite perfume, but did not track product purchases or long-term use.

Summary

Across two sessions about three months apart, women’s preferences for men’s body odors showed some repeatability. The central finding was that relative preferences were partly retained even with freshly collected samples from the same men.

Ratings of fragrance materials and facial photographs were more stable than those of body odors, with no clear difference between fragrances and faces. However, people with more stable preferences in one domain did not necessarily have more stable preferences in another. The study highlights the importance of considering changes both in the person smelling and in what is being smelled.

The research was funded by the Wellcome Trust, and support for authors from the Czech Science Foundation and Charles University is also reported. No separate conflict-of-interest statement was found in the article text.

Reference

Roberts SC, Havlíček J, Petrie M. Repeatability of odour preferences across time. Flavour and Fragrance Journal. 2013;28:245–250. doi:10.1002/ffj.3142 (Author-hosted full text)