What Changes in the Brain and Blood When a Scent Brings Back Memories?

A scent associated with a personal memory can bring back a scene and shift our mood along with it. A 2013 study examined brain activity and immune-related substances in the blood during this experience.

Among 10 preselected participants, a personally meaningful scent was followed by greater nostalgia and memory recall, and lower levels of some blood markers, than a control scent. The central question was how these subjective, brain, and bodily responses were related. Although small, the study examined the experience of recalling one’s own past through scent from several perspectives.

Using a perfume associated with the person’s own memories

The participants were 10 healthy adults aged 20–35: three men and seven women. Rather than simply recruiting people who liked fragrances, the researchers used a questionnaire to preselect people who could clearly recall personal odor-related events, including when and where they had happened.

The day before the experiment, each participant chose a perfume that evoked a personal memory. The comparison used two unmarketed perfumes that had been unlikely to evoke nostalgia or autobiographical memories in a preliminary test with 54 other people. An air-only condition was also included for the brain-imaging comparisons.

This differs from a design in which everyone smells the same supposedly nostalgic scent. By using scents linked to each person’s experiences, the researchers sought to examine situations in which odor-evoked memories actually occurred.

Two minutes of scent, brain blood flow, and a blood sample immediately afterward

Participants lay on the PET scanner table and smelled a scent delivered by a device for two minutes. Regional blood flow in the brain was measured with PET for 60 seconds during that period. Blood was collected after the scent exposure ended, and mood and other ratings were completed within two minutes.

Each of the three conditions, personally meaningful scent, control scent, and air, was presented twice, for six scans in total. The order was counterbalanced across participants, with rest periods between sessions.

Subjective ratings covered odor intensity, nostalgia, memory recall, current pleasant mood, and a scale ranging from sleepiness to alertness. Blood measurements covered six cytokines, signaling substances in the immune system: IL-2, IL-4, IL-6, IL-10, TNF-α, and IFN-γ.

The main mood and blood-marker findings below compare conditions after scent exposure. They should be distinguished from measurements of how much a person’s state improved from before exposure.

Greater nostalgia and recall, with lower levels of two blood markers

The personally meaningful scents produced significantly higher nostalgia and memory-recall ratings than the control scents. Perceived odor intensity did not differ significantly.

The authors also report higher pleasant-mood and alertness ratings than in the control condition. However, for pleasant mood, the p value obtained from a conventional two-sided test using the reported t statistic and degrees of freedom is inconsistent with the stated p < 0.05. The figure also shows a higher average mood rating, but the reported significance requires caution. The alertness scale runs toward wakefulness rather than sleepiness, so this is not a finding of sedation.

Of the six cytokines, TNF-α and IFN-γ were significantly lower after the personally meaningful scent than after the control scent. No significant decrease was shown for the other four. The decrease in IL-2 found in the authors’ earlier study was not reproduced here.

Cytokines are diverse substances that regulate immune function, so a lower concentration cannot simply be reinterpreted as stronger immunity. What was measured here was the blood concentration immediately after healthy people had a particular experience.

Associations with brain regions implicated in memory and pleasant experiences

The PET comparisons with the control scent and air conditions showed increased activity in the medial orbitofrontal cortex and the precuneus/posterior cingulate cortex during the personally meaningful scent condition. Drawing on earlier research, the authors interpret the former as a candidate region for processing pleasant value and the latter as a candidate for recalling personal memories.

In the personally meaningful scent condition, people with greater activity in these two regions tended to have lower IFN-γ concentrations, a negative correlation. Activity in the two brain regions was also positively correlated with each other. The same correlations between brain activity and cytokines were not found in the control-scent condition.

One sentence at the beginning of the original paper’s discussion describes the relationship with IFN-γ as positive, but the abstract, numerical results, and figure show a negative relationship. This article follows the consistent results and figure.

From this combination, the authors propose that emotions accompanying memories may be associated with interactions between the brain and immune system. However, because the study measured correlations, it did not establish that these brain regions directly suppressed cytokine secretion.

The short time frame and personally meaningful scents are especially important

The authors themselves identified the short, two-minute exposure as a major limitation. Earlier cytokine studies often examined changes at least 30 minutes after stimulation, so the changes observed here may have occurred too quickly. The authors suggest rapid changes in the distribution of circulating immune cells as one possible explanation, but cell distribution was not measured in this study.

The personally meaningful scents may also have been familiar and pleasant to the participants. The study could not separate whether memory recall, pleasantness, or familiarity accounted for the relationships between brain activity and blood markers.

The sample was also limited to 10 people with vivid odor-related memories. Brain imaging used exploratory thresholds uncorrected for multiple comparisons, and no multiple-comparison correction is described for the tests of several cytokines and subjective measures. Although the authors discuss possible applications to health and well-being, this was not a trial of lasting health effects or improvement in disease. Those possibilities remain hypotheses for future testing.

Summary

In 10 preselected participants, scents that evoked personal memories were accompanied by stronger nostalgia and memory recall than control scents. In the same condition, TNF-α and IFN-γ levels were lower, activity was higher in candidate regions involved in memory and pleasant experiences, and that activity was negatively correlated with IFN-γ.

The study’s contribution was to examine scent-evoked recollection of the past not only through subjective experience, but also in relation to brain and blood measurements. Within the limits of a short-term response and a small sample, it offers clues for further study of the relationships among memory, emotion, and the body.

Research context and disclosures: The study reports partial support from JSPS Grant-in-Aid for Scientific Research 22700683 and no additional external funding. Four coauthors were Kao employees at the time of the research and were involved in study design and the provision of materials and tools. Kao also provided the control scents. The paper states that the public funding body had no role in study design, data collection and analysis, the decision to publish, or manuscript preparation.

References

Matsunaga, M., et al. (2013). Brain–Immune Interaction Accompanying Odor-Evoked Autobiographic Memory. PLOS ONE, 8(8), e72523. Original study and open full text