A Pleasant Scent with an Unpleasant Ingredient: How Does the Brain Respond?

Even an odor that is unpleasant on its own can feel pleasant overall when mixed with another scent. A 2007 study used a jasmine-like fragrance mixed with indole to investigate how the brain handles this combination of pleasantness and unpleasantness.

The results were not as simple as every part of the brain reducing the mixture to a single overall score. Regions associated with pleasantness responded to the mixture rather like the jasmine-like fragrance, whereas regions associated with unpleasantness showed responses intermediate between that fragrance and indole.

A jasmine model with a defined composition, rather than natural flowers

The experimental fragrance was a jasmine-like mixture of eight ingredients, including linalool, benzyl acetate, hedione, and jasmone. The researchers first made a model without indole, then added indole so that they could investigate what changed when an unpleasant component was introduced into the same floral scent.

They compared four conditions: the model fragrance, indole alone, a mixture of the two, and odorless air. The mixture was adjusted to keep the concentrations of the eight model-fragrance ingredients the same as in the original condition. Natural jasmine also contains indole, but this experiment measured responses to a laboratory fragrance with a controlled composition.

Indole was set at 6% in the initial formulation, or 4% for some participants, and each stimulus was further diluted 1:5 in dipropylene glycol. These values describe preparation of the liquid samples, not the airborne concentrations reaching the nose.

Studying 14 people who found indole unpleasant alone but pleasant in the mixture

The participants were 14 healthy adults, seven men and seven women, with a mean age of 26. They were selected beforehand for finding indole alone unpleasant and its mixture with the jasmine model pleasant. Because people differ in their responses to indole, the study compared brain responses in people who met these criteria.

Inside an fMRI scanner, participants smelled stimuli delivered near their noses for four seconds, then rated pleasantness and intensity separately. They were not told the names or number of stimulus types, and each of the four conditions was presented nine times in random order. Brief presentations of limonene and clean air were interspersed between trials.

The analysis examined relationships between participants’ ratings on each trial and the fMRI BOLD signal, which is related to blood flow. Responses to the three odors were compared in regions identified as candidates from earlier research, including the orbitofrontal cortex, anterior cingulate cortex, piriform cortex, and insula.

The mixture remained pleasant, but not everyone liked it more

In participants’ ratings, both the jasmine model and the mixture were more pleasant than indole alone. However, mean pleasantness did not differ significantly between the model fragrance and the mixture. Two of the 14 participants rated the mixture as more pleasant, but the group results did not show that adding indole made the scent better liked.

Intensity ratings also showed no clear differences among the three stimuli. Under these conditions, the comparison therefore made it easier to focus on differences in pleasantness rather than comparing stimuli with greatly different intensities. A nonsignificant difference does not, however, guarantee that every person experienced exactly the same intensity.

Regions associated with pleasantness responded more like they did to the floral scent

In the medial orbitofrontal cortex, higher pleasantness ratings were associated with higher BOLD signals. Here, the mixture produced a greater response than indole alone, and also a higher response than the jasmine model.

Another, more anterior and lateral region of the orbitofrontal cortex also showed a relationship with pleasantness. In that region, however, responses to the mixture and model fragrance did not differ significantly, and both differed from indole alone. From the perspective of regions associated with pleasantness, the mixture was treated more like the pleasant floral scent than the unpleasant indole.

The authors discussed the possibility that the higher response to the mixture in the medial region reflected features such as the fragrance’s fullness that simple pleasantness ratings might not fully capture. This did not establish that the brain knew participants’ true preferences better than their own reports. Rather, it offered a lead for further investigation of the relationship between ratings and brain signals.

Regions associated with unpleasantness showed intermediate responses

By contrast, a posterior mid-orbitofrontal region, the dorsal anterior cingulate cortex, and part of the insula showed the opposite relationship: lower pleasantness ratings were associated with greater responses. In these regions, responses to the mixture lay between those to indole alone and the jasmine model.

The individual comparisons nevertheless differed. In the posterior mid-orbitofrontal region, the mixture produced a greater response than the model fragrance. In the dorsal anterior cingulate cortex and insula, indole alone produced greater responses than either the mixture or model fragrance, while the difference between the latter two was not significant.

The researchers also compared, across regions, how close the mixture’s signal was to that of the model fragrance or indole. They confirmed that this relative position differed between regions associated with pleasantness and those associated with unpleasantness. This is the paper’s central finding: even when someone calls a mixture pleasant overall, different brain regions need not represent the same average score.

Separate responses associated with intensity

In part of the piriform cortex, which receives olfactory information, trial-by-trial intensity ratings were positively related to the BOLD signal, but pleasantness ratings were not. Mean response magnitudes did not differ significantly among the three stimuli.

Another part of the insula showed an inverse relationship with intensity ratings, again without significant differences in mean responses to the three stimuli. By distinguishing responses associated with pleasantness, unpleasantness, and intensity, the authors showed that the evaluation of an odor mixture differs across brain regions.

Summary

The jasmine-like mixture remained pleasant to the selected participants even though it contained indole that they found unpleasant alone. In brain regions associated with pleasantness, the mixture elicited responses closer to the floral scent; in regions associated with unpleasantness, responses were intermediate; and in regions associated with intensity, response magnitudes were similar across the three stimuli.

The authors interpreted the findings as evidence that pleasant and unpleasant aspects can be represented in parallel in the brain, and discussed how this information might be used in making choices. Actual choice behavior was not measured, but the study suggests that a single answer about liking or disliking a scent can sit atop processes handling several different aspects.

This was a small experiment involving 14 people selected beforehand for their responses to indole and the mixture. The research received support from the Gottlieb-Daimler and Karl Benz Foundation, and two authors were affiliated with Firmenich at the time. No separate conflict-of-interest statement was found in the paper.

Reference

Grabenhorst, F., Rolls, E. T., Margot, C., da Silva, M. A. A. P., & Velazco, M. I. (2007). How Pleasant and Unpleasant Stimuli Combine in Different Brain Regions: Odor Mixtures. The Journal of Neuroscience, 27(49), 13532–13540. Original paper