When Tatami and Osmanthus Scents Bring Past Scenes to Mind

The scent of osmanthus brings to mind a field where someone played with friends as a child. The smell of tatami recalls a scene of talking with a grandmother. Older participants in a 2021 study described specific memories like these.

The researchers examined brain activity while people smelled odors connected with personal memories, and how that activity related to memory vividness and feelings of comfort. By selecting an odor suited to each person and comparing it with a control odor less strongly linked to memory, they explored the experience of a scent bringing a scene to mind.

Choosing a personally meaningful memory-associated odor

The researchers recruited 30 older adults who lived independently and reported concerns such as forgetfulness. A neurologist’s assessment classified them as having normal cognitive function. After exclusions based on medical history, 27 proceeded to MRI scanning. Excluding participants affected by technical problems or who stopped because of discomfort during scanning left 24 for analysis: 12 men and 12 women, with a mean age of 74.4 years.

On the first day, the researchers assessed olfactory and cognitive function and presented five odors in random order: tatami, osmanthus, baby powder, citrus, and incense. Participants reported whether the odors brought to mind a person, place, or event, their approximate age at the time of the memory, and its pleasantness, vividness, emotional intensity, and other qualities.

For scanning, the researchers used the odor that elicited the strongest autobiographical memory in each participant. Thus, not everyone smelled tatami or osmanthus; one personally relevant odor was selected for each individual. The control was β-phenylethyl alcohol, which has a rose-like smell.

Alternating odor and unscented air in 30-second blocks

fMRI was performed about two weeks later. Participants smelled the odors through a nasal mask, drawing them in with their own inhalations. Each session alternated 30 seconds of odor with 30 seconds of unscented air, repeating each block five times. The memory-associated odor and control odor were measured in separate sessions.

fMRI detects changes in brain activity through signals related to blood oxygenation. The study first compared each odor with unscented air, then directly compared the memory-associated and control odors. The researchers also modeled relationships between the activity of different brain regions and subjective ratings of memory and emotion.

Differences in comfort and a sense of returning to the past, rather than odor pleasantness

The authors reported that memory-associated odors produced higher ratings of comfort, memory vividness, degree of memory retrieval, and the sense of returning to the past than the control odor. There were no significant differences in the pleasantness of the odor itself or in arousal.

This is worth distinguishing from a simple account in which a pleasant smell improves someone’s mood. Even without a clear difference in pleasantness, the experience of recalling a personal scene differed. However, some z values and p values reported for the subjective ratings are inconsistent. The discrepancy is particularly large for the sense of returning to the past, so the reported direction of the means must be distinguished from rigorous confirmation of statistical significance.

Connections between visual regions and the orbitofrontal cortex

Compared with unscented air, memory-associated odors were accompanied by increased activity in the left posterior orbitofrontal cortex, right inferior frontal cortex, and fusiform gyri on both sides. The control odor also activated both fusiform gyri. Because the fusiform gyrus is involved in visual processing, not all of this activity was unique to memory-associated odors.

In the direct comparison between the memory-associated and control odors, the left posterior orbitofrontal cortex showed greater activity for the memory-associated odor. In that condition, connections were also detected between activity in this region and regions including the left fusiform gyrus.

Regions involved in vision were engaged even though no photographs were shown, and participants described scenes involving people and places. Together, these observations led the authors to consider the possibility of coordination between odor-evoked mental imagery and emotions associated with memory. The study did not decode the content of imagined scenes from brain images.

A model linking a sense of returning to the past, memory, and brain activity

The study’s second aim was to examine relationships between subjective experience and brain activity in a single path model. In the memory-associated odor condition, the retained paths linked a stronger sense of returning to the past with higher retrieval ratings, the degree of retrieval with activity in the left fusiform gyrus, and that activity with activity in the left posterior orbitofrontal cortex. Comfort and vividness were also associated with other subjective ratings.

For the control odor, relationships were found among subjective ratings, but there was no corresponding significant path to the left fusiform gyrus. The authors interpreted the results as a model bringing together activity associated with memory, emotion, and visual imagery.

However, arrows in the model do not establish the actual sequence of causes and effects. Paths were narrowed down using the same data, and the sample comprised only 24 people. The direct path to pleasantness was not significant, while the prose describing indirect paths as significant conflicts with the table’s p values and path labels. The model cannot be treated as establishing the pathways through which pleasantness arises.

The possibility of stimulating memory in later life, and what this study established

Expected activity in regions such as the hippocampus and amygdala did not remain significant after correction for multiple comparisons. The authors suggested possible contributions from habituation to repeated odors and attention to the recalled scenes. These are questions for further testing, not mechanisms directly established in this study.

The authors also discussed whether stimulating memory and emotion with personally meaningful odors might benefit cognitive function in later life. But the study measured responses during a short experiment; it did not investigate long-term cognitive improvement or dementia prevention. Further research needs to include more people, including those with cognitive decline, while assessing the content of memories in greater detail.

Summary

Odors matched to participants’ personal memories produced higher ratings of memory retrieval, comfort, and related experiences than a control odor, together with greater activity in the left posterior orbitofrontal cortex. The study also found connections with the left fusiform gyrus, which is involved in visual processing, and associations with the strength of memory retrieval.

A scent can lead to the experience of recalling a past scene as a mental image. This study examined that experience through older adults’ subjective ratings and brain activity, proposing a model linking memory, emotion, and visual imagery. Causal relationships within the model and applications to health remain to be tested.

Research context and disclosures: The study was funded by KAKENHI grant 20K03482 and Kao Corporation. At the time of the research, coauthor Haruko Sugiyama was affiliated with Kao and contributed to the design, experiments, and analysis of subjective measures. The paper states that the data analysts were blinded to group information.

Reference

Masaoka, Y., et al. (2021). Odors Associated With Autobiographical Memory Induce Visual Imagination of Emotional Scenes as Well as Orbitofrontal-Fusiform Activation. Frontiers in Neuroscience, 15, 709050. Original paper and open full text