How Does Information About Liking or Disliking an Odor Travel Through the Brain?

With the same odor, are the first impression that makes you want to pull away and the more settled judgment you reach after smelling it briefly really part of the same process? In 2024, Frans Nordén and colleagues investigated when, and in which direction, information about liking or disliking an odor travels between the olfactory bulb and the piriform cortex.

The olfactory bulb is a brain structure that receives olfactory information arriving from the nose; the piriform cortex processes odor information further along the pathway. The study suggested that information not only travels from the nose-side structure toward the cortex but also returns from the cortex, with different timing and different rhythms of electrical activity.

Comparing personal odor preferences with brain signals in two experiments

Fifty-five adults took part in the first experiment. After 15 were excluded on criteria including signal quality, 40 were included in the analysis. In the second experiment, 29 participated and 20 were analyzed after nine exclusions. The mean ages of the analyzed participants were approximately 26 and 29 years, respectively.

In the first experiment, four odors were each presented 20 times, for 80 presentations in total. The second broadened the range of pleasantness and unpleasantness, including odors with a neutral impression: six odors were each presented 30 times, for 180 presentations. A device delivered odors in time with nasal inhalation, after which participants rated how pleasant and how intense each odor was on scales from 0 to 100.

Importantly, the researchers did not decide that an odor was pleasant or unpleasant simply from the name of the substance. They compared the highest and lowest thirds of each person’s trial-by-trial pleasantness ratings. This approach accommodates the fact that experience and associations can change how the same substance is perceived.

The rhythms linking two brain regions

Alongside scalp EEG, the researchers recorded signals from electrodes intended to capture olfactory bulb activity. Because these recordings contain activity mixed from many locations, they used a standard head model and electrode positions to estimate activity originating in the olfactory bulb and piriform cortex. This was not an experiment in which electrodes were inserted into the brain to measure activity directly.

First, they analyzed when, and at which frequencies, fluctuations in electrical activity in the two regions were coordinated. They then used machine learning to test whether this relationship could distinguish trials rated pleasant from those rated unpleasant. Rather than simply asking whether brain signals were large or small, the study focused on the relationship between activity in the two locations.

They also tested whether knowing the past of one signal helped predict the next state of the other. This method, called Granger causality, statistically estimates the direction of information flow. Although its name includes the word causality, it differs from an experiment that manipulates one brain region to establish that it directly drives another.

Early, fast rhythms emphasize unpleasantness

At an early point, around 150 milliseconds, or 0.15 seconds, after an odor reached the nose, information distinguishing pleasant from unpleasant appeared in fast fluctuations called the gamma band. At this stage, trials rated unpleasant tended to be easier to classify than pleasant trials.

In a separate directional analysis, gamma-band relationships were consistent with information flowing from the olfactory bulb to the piriform cortex. Combining the timing and direction analyses, the authors interpreted the results as showing that early information, particularly about unpleasantness, is sent from the bulb to the cortex.

The authors suggest that this rapid stage could help distinguish odors quickly and prepare avoidance behavior when needed. However, the study did not measure actual withdrawal movements or the ability to identify hazards. Perceiving an unpleasant odor is also not the same as determining whether a substance is harmful.

Information returning later also includes pleasant impressions

Later, at approximately 600–850 milliseconds, information distinguishing pleasant from unpleasant appeared in the beta band, which fluctuates more slowly than the gamma band. Pleasant-odor trials also became relatively easier to classify in this later phase. The authors see this as broader information encompassing both pleasantness and unpleasantness than was available earlier.

The directional analysis indicated a beta-band relationship returning from the piriform cortex to the olfactory bulb. Similar patterns of timing and frequency were also observed in the second experiment, which used different odors and numbers of presentations.

This supports a two-stage view: an initial broad response is followed by the return of information processed in the cortex. It does not mean that unpleasantness disappears later. Rather, alongside the unpleasantness that stood out early, evaluations including pleasant impressions became easier to capture.

What is the returning information doing?

The authors propose that the later returning information may incorporate the context in which an odor occurs and memories associated with it. Their hypothesis is that this information updates the initial processing in the olfactory bulb and also influences the processing of subsequent odors.

The study did not independently manipulate memories or context to observe such updating directly. Even so, the different timing and directions of signals carrying pleasantness and unpleasantness offer a clue that evaluating an odor may not end with a single judgment of the incoming stimulus.

There are limits to what these signals can tell us

Signals recorded from the scalp and reconstructed to estimate their sources are weaker than signals recorded directly inside the brain and may contain other activity. Predictive relationships between signals provide evidence consistent with a direction of information transfer, but do not establish a biological causal pathway.

There were also substantial individual differences in classification performance. Even the same unpleasantness score need not represent the same sensations or associations for different people. The researchers considered an explanation based on odor intensity and reported a different time–frequency pattern from the main results, but did not isolate every potentially influential factor, such as attention or breathing.

Nor does the difficulty of extracting early information about pleasant odors necessarily mean that such information was not being processed. The authors leave open the possibility that pleasant-odor representations were more complex and harder to capture with this method.

Summary

In this study, information about liking or disliking odors appeared differently in early gamma-band and later beta-band activity. Unpleasantness-related information was prominent early, whereas later information covered pleasantness and unpleasantness more broadly. Directional analyses were consistent with a flow from the olfactory bulb to the piriform cortex and back again.

In the authors’ interpretation, this back-and-forth exchange allows a rapid initial response to be updated into an evaluation informed by context and memory. Liking or disliking an odor may not simply be a fixed value determined in one place and passed onward, but something assembled over time. Capturing this process in human signals is the study’s contribution.

References

Nordén, F., Iravani, B., Schaefer, M., Winter, A. L., Lundqvist, M., Arshamian, A., & Lundström, J. N. (2024). The human olfactory bulb communicates perceived odor valence to the piriform cortex in the gamma band and receives a refined representation back in the beta band. PLOS Biology, 22(10), e3002849. DOI: 10.1371/journal.pbio.3002849 / Full original article.

Original article: © 2024 Nordén et al. CC BY 4.0.