Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2024 Nov 2;45(16):e70061. doi: 10.1002/hbm.70061

Olfactory Dysfunction and Limbic Hypoactivation in Temporal Lobe Epilepsy

Markus Schmidt 1,2, Tobias Bauer 1,2,3, Marcel Kehl 2, Anna Minarik 2,4, Lennart Walger 1,2, Johannes Schultz 5,6, Martin S Otte 7, Peter Trautner 5, Christian Hoppe 2, Tobias Baumgartner 2, Louisa Specht‐Riemenschneider 8, Florian Mormann 2, Alexander Radbruch 1,3,8, Rainer Surges 2, Theodor Rüber 1,2,3,8,✉
PMCID: PMC11530705  PMID: 39487626

ABSTRACT

The epileptogenic network in temporal lobe epilepsy (TLE) contains structures of the primary and secondary olfactory cortex such as the piriform and entorhinal cortex, the amygdala, and the hippocampus. Olfactory auras and olfactory dysfunction are relevant symptoms of TLE. This study aims to characterize olfactory function in TLE using olfactory testing and olfactory functional magnetic resonance imaging (fMRI). We prospectively enrolled 20 individuals with unilateral TLE (age 45 ± 20 years [mean ± SD], 65% female, 90% right‐handed) and 20 healthy individuals (age 33 ± 15 years [mean ± SD], 35% female, 90% right‐handed). In the TLE group, the presumed seizure onset zone was left‐sided in 75%; in 45% of the individuals with TLE limbic encephalitis was the presumed etiology; and 15% of the individuals with TLE reported olfactory auras. Olfactory function was assessed with a Screening Sniffin’ Sticks Test (Burkhart, Wedel, Germany) during a pre‐assessment. During a pre‐testing, all individuals were asked to rate the intensity, valence, familiarity, and associated memory of five different odors (eugenol, vanillin, phenethyl alcohol, decanoic acid, valeric acid) and a control solution. During the fMRI experiment, all individuals repeatedly smelled eugenol (positively valenced odor), valeric acid (negatively valenced odor), and the control solution and were asked to rate odor intensity on a five‐point Likert scale. We acquired functional EPI sequences and structural images (T1, T2, FLAIR). Compared to healthy individuals, individuals with TLE rated the presented odors as more neutral (two‐sided Mann–Whitney U tests, FDR‐p < 0.05) and less familiar (two‐sided Mann–Whitney U tests, FDR‐p < 0.05). fMRI data analysis revealed a reduced response contrast in secondary olfactory areas (e.g., hippocampus) connected to the limbic system when comparing eugenol and valeric acid in individuals with TLE when compared with healthy individuals. However, no lateralization effect was obtained when calculating a lateralization index by the number of activated voxels in the olfactory system (two‐sided Mann–Whitney U test; U = 176.0; p = 0.525). TLE is characterized by olfactory dysfunction and associated with hypoactivation of secondary olfactory structures connected to the limbic system. These findings contribute to our understanding of the pathophysiology of TLE. This study was preregistered on OSF Registries (www.osf.io).

Keywords: functional MRI, neurological disorder, olfactometry, sense of smell

This study aims to characterize olfactory function in temporal lobe epilepsy using olfactory testing and olfactory functional magnetic resonance imaging. Our findings show olfactory dysfunction associated with hypoactivation in the limbic system of people with temporal lobe epilepsy.

graphic file with name HBM-45-e70061-g001.jpg

Summary.

  • Individuals with temporal lobe epilepsy perceive odors more neutral and less familiar.

  • fMRI hypoactivation in limbic areas in individuals with temporal lobe epilepsy when comparing between positively and negatively valenced odors.

Abbreviations

BOLD

blood oxygenation level‐dependent

EPI

echo‐planar imaging

FLAIR

fluid attenuated inversion recovery

fMRI

functional magnetic resonance imaging

GLM

general linear model

LE

limbic encephalitis

MNI

Montreal Neurological Institute

MRI

magnetic resonance imaging

PC

piriform cortex

PET

positron emission tomography

ROI

region of interest

SD

standard deviation

TLE

temporal lobe epilepsy

1. Introduction

Temporal lobe epilepsy (TLE) is the most common focal epilepsy in adults. Its epileptogenic network contains structures of the mesial temporal lobe, primarily the hippocampus, amygdala, and connected cortical areas (Bernhardt et al. 2013; Engel 1996) with emphasis on the limbic system which integrates olfactory, emotional and behavioral information. These cortical areas include the limbic lobe, orbitofrontal cortex, entorhinal cortex, and also the piriform cortex. In particular, the critical role of the piriform cortex (PC) in the pathogenesis of TLE has been demonstrated by recent research (Cheng et al. 2020; Vaughan and Jackson 2014). Moreover, it has been shown that in surgically treated individuals with TLE, resecting at least half of the PC increases the odds of seizure freedom by a factor of 16 (Borger et al. 2021; Galovic et al. 2019; Hwang et al. 2022). The PC is a part of the primary olfactory cortex and related to odor perception (Gottfried et al. 2002; Gottfried, Winston, and Dolan 2006; Howard et al. 2009). It is therefore not surprising that TLE is in many ways associated with olfactory dysfunction. Previous studies have shown that odor identification and discrimination are impaired in individuals with TLE, while the threshold sensitivity for odor perception is lower in people with TLE than in healthy controls (Doty et al. 2018; Khurshid et al. 2019). Additionally, olfactory dysfunction is likely to appear if the age of onset is older and the seizure foci are bilateral (Motoki et al. 2021). However, the findings about olfactory function in TLE are heterogeneous because olfactory testing has not been standardized yet (Hwang et al. 2020). Olfactory auras are reported in 5.1% of all individuals with focal epilepsy, 81.7% of whom have TLE (Taşcı et al. 2021). In individuals with mesial TLE and unilateral hippocampal sclerosis, approximately 3% experience olfactory auras (Dupont et al. 2015). Despite the clinical association between TLE and olfactory dysfunction, little is known about the neuronal substrates of olfactory dysfunction in TLE. Previous studies utilizing functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) to investigate the response to an olfactory stimulus in TLE identified an asymmetric pattern of activation, with greater activity observed in the hemisphere contralateral to the seizure onset (Aguado‐Carrillo et al. 2021; Ciumas et al. 2008) making it possible to predict the seizure onset zone by activation. These studies included bilateral exposure to one odor such as coffee (Aguado‐Carrillo et al. 2021), or exposure to familiar and unfamiliar odors in comparison (Ciumas et al. 2008). However, less is known about the valence‐dependent stimulation of the olfactory network, although it is expected that the pleasantness of a stimulus could have a deep impact on the modulation of the activation due to the strong association between odors and emotional reactions. To gain a deeper insight, in this study, we use behavioral tests (1) and a specific olfactory fMRI paradigm (2) to investigate the olfactory network in TLE and compare it with healthy individuals. We hypothesized that (1) the olfactory perception (valence, familiarity, and intensity) differs between people with TLE and healthy individuals, (2a) the fMRI activation during olfactory stimulation generally differs between people with TLE and healthy individuals, and (2b) this difference is modulated during olfactory stimulation with either positive or negative valence odors varies between people with TLE and healthy individuals.

2. Methods

2.1. Participants

Individuals with TLE were prospectively recruited at the Department of Epileptology at University Hospital Bonn in 2023. The inclusion criterion was a diagnosis of unilateral TLE or limbic encephalitis (LE) with a predominantly affected hemisphere as extracted from the clinical records. Healthy individuals were recruited via advertisements at the University of Bonn, social media, and from a control database at the Department of Epileptology, University Hospital Bonn. For all participants, further inclusion criteria were age above 18 years, normal or corrected‐to‐normal vision, and ability to undergo MRI scanning for 1 h. Exclusion criteria for all participants were a history of neurological or psychiatric disorders (besides TLE), prior neurosurgery (except for invasive EEG: electrocorticography or stereoelectroencephalography), COVID infection within 1 month prior to MRI (Jafar et al. 2021), daily tobacco use, history of otorhinolaryngological trauma or surgery, MRI contraindications according to clinical standards. All participants had signed an informed consent before inclusion in the study. The study protocol was approved by the institutional review board of the Medical Faculty of the University of Bonn (No. 434/22) and is in accordance with relevant guidelines and regulations. Lateralization of the epileptogenic focus in individuals with TLE (or the predominantly affected hemisphere in LE, respectively) was assessed with the reports of all available diagnostic modalities, including non‐invasive video EEG, MRI, PET, SPECT, and neuropsychological testing (Harms et al. 2023).

2.2. Olfactory Evaluation

2.2.1. Experimental Setup

An olfactometer (Sniff‐0, CyNexo srl, Udine, Italy) and a breathing cycle monitor (Spir‐0, CyNexo srl, Udine, Italy) were used for the olfactory paradigm. The olfactometer consists of 13 + 1 channels including a clear air channel, 6 of them were used for presenting odors. The following odors were used as stimuli: eugenol (0.5 mL), vanillin (5 mg), phenethyl alcohol (0.8 mL), decanoic acid (2 mg), and valeric acid (0.2 mL). All odors were diluted in 10 mL of 1,2‐propanediol, and 10 mL of pure 1,2‐propanediol was used as a control condition. Fresh air was provided by a compressor with a pressure of 2.5–3 bar. Inside the olfactometer, the air was downregulated to two bars. Every channel in use was calibrated to an airflow rate of 2.5 L/min and strengthened with Teflon tape to avoid any air leakage. Teflon tubes connected the olfactometer located at the MRI console with the manifold connected to the participant inside the MRI. The participant wore the manifold at their breast, where it was secured with belts. From there, the odorous air was presented to the participant via two nasal tubes, which are plugged into the nose through nasal adapters. The nasal adapters are attached to nasal tubes, which are connected to the manifold. The manifold distributes odors from different channels to the nasal tubes. Although each nasal adapter is designed to deliver odor to a specific nostril, the manifold sends odors to both nasal tubes simultaneously, resulting in stimulation of both nostrils at the same time. One of the nasal adapters was fixed with a thin fiber optic cable consisting of a sensor at the tip for detecting the participant's airflow and recording it at the control panel. Participants were instructed to breathe only through the nose during the whole experiment in their normal breathing cycle, without sniffing when they felt that odor was released. Furthermore, participants were provided with a controller in each hand for intensity evaluation during the fMRI paradigm. The participants were instructed not to react to the odor stimulation but to breathe regularly and evaluate the intensity of each stimulus only after the pause. An in‐house paradigm presentation software provided by the Human MRI core facility of the Medical Faculty, University of Bonn was used for controlling odor releases, synchronizing them with the MRI scanner, and receiving the participant's input when asked to evaluate odor intensity. Figure 1 shows a schematic of the experimental design.

FIGURE 1.

FIGURE 1

Open in a new tab

Experimental design. After the olfactory pre‐assessment of whether to include the recruited individuals, a pre‐testing follows to assess the intensity, valence, familiarity, and individual memory of the odors (eugenol, vanillin, phenethyl alcohol, decanoic acid, valeric acid) and a control solution. Finally, the olfactory paradigm is executed with eugenol as a pleasant odor, valeric acid as an unpleasant odor, and a control solution, consisting of olfactory stimulation and the subjective intensity rating of this stimulation.

2.2.2. Olfactory Pre‐Assessment

Basic olfactory capacities were assessed using a basic smell questionnaire (see Supporting Information). To screen for hyposmia, the general olfactory function was assessed with a Sniffin’ Sticks Test (Burkhart, Wedel, Germany) (Hummel et al. 2001). Each nostril was tested separately. The participants were asked to first close the right nostril by gently applying pressure with their finger and odors were presented under the left nostril by the examiner. After presenting all odors and shuffling the odor palette, this was repeated for the other nostril. Age‐specific cutoffs for the number of incorrectly identified odors indicative of hyposmia were provided by the manufacturer. When the number of incorrectly identified odors in the nostril ipsi‐/contralateral to the presumed seizure onset zone was compared to healthy individuals on a group level, an equal proportion of left/right nostrils was randomly sampled from the control group.

2.2.3. Olfaction Pre‐Testing

To verify that the selected two odors with positive and negative valence under equivalent conditions compared to the paradigm were consistent across subjects, before the actual MRI scan, each participant was presented with all six stimuli (eugenol, vanillin, phenethyl alcohol, decanoic acid, valeric acid, and the control solution) while lying in the MRI scanner equipped with the full experimental setup. Each stimulus was presented for 4 s. After each stimulus, participants were asked to rate the intensity (no odor, very weak, weak, strong, very strong), the valence (pleasant, unpleasant, neutral), and the familiarity (known, unknown). If the stimulus was rated as known, participants were further asked to describe the closest related memory to the stimulus in their own words. All answers were communicated verbally to the examiner. After all questions were answered, the next stimulus was released.

2.3. MRI Acquisition

All experiments were performed on a 3T MRI scanner (Siemens Magnetom TrioTim, Siemens Healthcare, Erlangen, Germany) using a 32‐channel receiver head coil. For the BOLD signal acquisition during the olfactory paradigm, an echo planar imaging (EPI) sequence with 30° angulation in the plane between the anterior and posterior commissure was used to avoid signal dropout in the olfactory cortex due to the frontal sinus (Deichmann et al. 2003; Gottfried 2015; Gottfried et al. 2002). Sequence parameters were TR = 2000 ms, TE = 28 ms, voxel size = 3.0 × 3.0 × 2.0 mm, FOV = 192 mm, flip angle = 60°. 0.8 mm isotropic T1‐weighted (sequence parameters: TR = 1660 ms, TE = 2.54 ms, voxel size = 0.8 × 0.8 × 0.8 mm, FOV = 256 mm, flip angle = 9°), T2‐weighted (sequence parameters: TR = 3200 ms, TE = 401 ms, voxel size = 0.8 × 0.8 × 0.8 mm, FOV = 256 mm) and FLAIR images (sequence parameters: TR = 5000 ms, TE = 388 ms, voxel size = 1.0 × 1.0 × 1.0 mm, FOV = 256 mm) were acquired.

2.4. Olfactory Paradigm

2.4.1. Piloting

Before recruiting the final study sample, we conducted a pilot study with 10 participants who met the inclusion criteria for healthy participants to establish a solid pipeline. During the piloting phase, all six stimuli were presented and the most pleasant and unpleasant odors were selected based on the ratings of the piloting phase. After the most positive and negative odors were selected and the technical issues that occurred during the piloting phase were resolved, all 10 participants of the pilot phase were re‐examined with the finalized protocol. The data from the second run of the piloting phase were included in the control sample.

2.4.2. Experiment

Each stimulation block started with a black screen on the monitor placed behind the MRI scanner that could be viewed by participants via a mirror placed over the MRI head coil. After a jittered interval of 3–5 s, either eugenol (pleasant odor), valeric acid (unpleasant odor), or the control solution was released through the olfactometer for 1 s, followed by a pause of 2 s. This was followed by another 1 s‐release of the same stimulus and another pause of 2 s. This kind of odor presentation has been found optimal to maximize the signal in the olfactory brain regions and aligns with previously used paradigms (Georgiopoulos et al. 2018). After a pause of 8 s, participants were prompted on the screen to evaluate the intensity of the presented odor by moving a slider with the controllers along a five‐point Likert scale as follows: no odor, very weak, weak, strong, very strong. This was intended to maintain concentration during the experiment. Each stimulation block ended with a pause of 4 s. In sum, odor releases were at least 21 s apart to avoid odor adaptation or stimulus overload (Donoshita et al. 2021; Gottfried 2015). In every run, eugenol, valeric acid, and the control solution were presented six times in a random order without the same odor or control solution being presented consecutively. This resulted in 18 stimulation blocks per run. In all participants, three runs were performed.

2.5. Data Analysis

2.5.1. Preprocessing

First, the dataset was formatted to the Brain Imaging Data Structure (BIDS) (Gorgolewski et al. 2016) and preprocessed with fMRIPrep (Esteban et al. 2019, 2020) (see Supporting Information). The fMRIPrep output of all participants was visually checked for plausibility. In detail, the output was checked for a plausible segmentation of grey matter, white matter, and cerebrospinal fluid. Furthermore, we verified the alignment of participants' scans following transformation into MNI space. Participants with a mean head motion over 5 mm or 3° were excluded. All further fMRI data were analyzed using Python packages from the NiPy ecosystem including Nilearn (Abraham et al. 2014). An individual high pass filter was set as three divided by the average period between two stimuli. A double‐gamma hemodynamic response function (HRF) model according to the Statistical Parametric Model (SPM) was used to convolve each stimulus block of 6 s.

2.5.2. Task‐Driven GLM Analysis

Activation maps of individual participants were assessed using a general linear model including six rigid‐body movement parameters, white matter signal, signal of cerebrospinal fluid, and a drift effect according to a cut‐off period based on the inter‐stimulus duration. Images were smoothed with a 6 mm fill width half maximum (FWHM) Gaussian kernel. The following two contrasts were tested: any odor (eugenol or valeric acid)—control solution and positive (eugenol)—negative (valeric acid). At this point, z‐maps of individuals with a presumed right‐hemispheric seizure onset were flipped along the left/right‐axis to visualize lateralization effects. The z‐maps of an equal number of chosen healthy individuals with age and sex match were as well flipped. Activation maps of olfactory stimulation were compared between people with TLE and healthy individuals using a general linear model including age and sex as covariates. The resulting group‐level z‐maps were thresholded using a false positive rate (FPR) control at FPR(p < 0.05). Activation clusters were identified by voxel size using the Python package AtlasReader. Furthermore, we performed a within‐subject lateralization analysis in which we counted the activated voxels of the thresholded contrast “positive + negative > control” (z ≥ 1.96) with a mask on each hemisphere including the primary olfactory cortex and the amygdala (based on the Havard‐Oxford Atlas). In a second step we calculated a lateralization index as (number of activated voxels in the left mask—number of activated voxels in the right mask)/(number of activated voxels in the left mask + number of activated voxels in the right mask). This approach was chosen in alignment with commonly used lateralization indices for assessing language lateralization through task‐based MRI (Binder et al. 1996).

2.5.3. Statistics

Statistical analyses were performed using the Python packages SciPy and Annotator. To compare behavioral results on Likert scales and the mean values of categorical variables in the pre‐testing, we performed two‐sided Mann–Whitney U tests for unpaired samples and Wilcoxon signed‐rank tests for paired samples. Continuous variables were compared using two‐sided t‐tests. False‐discovery rate was controlled using the Benjamini‐Hochberg method (Benjamini and Hochberg 1995). Furthermore, for the results of the GLM analysis, we performed a two‐sided t‐test and thresholded our results at p < 0.05, taking the false positive rate (FPR) into account (Taylor et al. 2023).

3. Results

3.1. Participants

3.1.1. Smell Questionnaire

Thirty‐three people with TLE who met the inclusion criteria were invited to participate. Eight were excluded because they met exclusion criteria or their performance in the Sniffin’ Sticks Test was indicative of hyposmia. One was excluded due to issues during the pre‐testing (e.g., communication problems). Four were excluded due to technical problems during the scan or because the scan was aborted.

Twenty‐eight healthy individuals who met the inclusion criteria were invited to participate. Three were excluded because they met exclusion criteria or their performance in the Sniffin’ Sticks Test was indicative of hyposmia. None healthy individual was excluded due to issues during the pre‐testing. Four were excluded due to technical problems during the scan or because the scan was aborted. One was excluded due to too much movement which was discovered in the preprocessing.

Finally, 20 individuals with TLE (age 45 ± 20 years [mean ± SD], 65% female, 90% right‐handed) and 20 healthy individuals (age 33 ± 15 years [mean ± SD], 35% female, 90% right‐handed) were included. In the TLE group, 75% had a presumed seizure onset zone on the left hemisphere, 15% had olfactory auras, and 45% had LE (see table 1 in supplement; Graus et al. 2016). All individuals with TLE had unilateral epilepsy (or a predominantly affected hemisphere if they had LE; 56% on the left side, 44% on the right side). The proportion of participants who had an infection with SARS‐CoV‐2 more than 1 month prior to MRI did not differ between people with TLE and healthy individuals (70% vs. 75%, χ 2(1) = 0.0, p = 1.0). Also, the proportion of participants who experienced smell loss during SARS‐CoV‐2 infection did not differ between people with TLE and healthy individuals (25% vs. 35%, χ 2(1) = 0.119, p = 0.730). Olfactory bulbs and olfactory grooves were examined in T1‐weighted and T2‐weighted images by a neuroradiologist. They appeared symmetrical and without any anomalies in all participants.

3.1.2. Sniffin’ Sticks Test

The number of incorrectly identified odors in the nostril ipsilateral to the presumed seizure onset zone was higher compared to the same side of healthy individuals (two‐sided Mann–Whitney U test, U = 102.0, p = 0.007). Comparing the nostril ipsilateral to the nostril contralateral to the presumed seizure onset zone within individuals with TLE, we could not find a lateralization of olfactory dysfunction (Wilcoxon signed‐rank test; p = 0.116; test statistic = 28.5).

3.2. Pre‐Testing

During pre‐testing, the perceived intensity did not differ between people with TLE and healthy individuals for any odor or the control solution (two‐sided Mann–Whitney U tests, all FDR‐p > 0.05). Regarding valence, valeric acid and eugenol were perceived more neutral by individuals with TLE than by healthy individuals (two‐sided Mann–Whitney U tests; U = 118.0, U = 277.5; FDR‐p = 0.037, FDR‐p = 0.046). Vanillin, valeric acid, phenethyl alcohol, and eugenol were less familiar to individuals with TLE than to healthy individuals (two‐sided Mann–Whitney U test; U = 277.5, U = 292.0, U = 293.0, U = 285.5; FDR‐p = 0.040, FDR‐p = 0.017, FDR‐p = 0.017, FDR‐p = 0.017). Additionally, within the TLE group, decanoic acid and phenethyl alcohol were less familiar to individuals with TLE and LE than to individuals with TLE without LE (two‐sided Mann–Whitney U test; U = 80.5, U = 79.5; FDR‐p = 0.044, FDR‐p = 0.044). Comparing individuals with LE with healthy individuals, eugenol, vanillin, phenethyl alcohol, decanoic acid, and valeric acid were perceived as less familiar by individuals with LE (two‐sided Mann–Whitney U test; U = 145.5; U = 131.5; U = 160.5; U = 121.5; U = 145.5; FDR‐p = 0.006; FDR‐p = 0.046; FDR‐p = 0.002; FDR‐p = 0.035; FDR‐p = 0.012). The proportion of individuals who found an odor familiar and were then able to associate a specific memory with it did not differ between people with TLE and healthy individuals (two‐sided Mann–Whitney U tests, FDR‐p > 0.05).

These results suggested that the chosen odors eugenol and valeric acid are eligible for the experiment because they significantly differed in valence compared to each other (two‐sided Mann–Whitney U test, U = 380.0, FDR‐p < 0.001) and compared to the control solution (two‐sided Mann–Whitney U test; U = 106.5, U = 356.0; FDR‐p = 0.025, FDR‐p < 0.001) in healthy individuals. Additionally, they were rated significantly more intense than the control solution (two‐sided Mann–Whitney U test; U = 66.5, U = 36.5; FDR‐p = 0.001, FDR‐p < 0.001) and they both did not differ in intensity in healthy individuals (two‐sided Mann–Whitney U test; U = 125.0, FDR‐p = 0.104).

3.3. Odor Intensity During the Scan

There were no significant intensity differences regarding perceived intensity between individuals with TLE and healthy individuals (two‐sided Mann–Whitney U tests, all FDR‐p > 0.05). Moreover, both odors were perceived as significantly more intense than the control solution (two‐sided Mann–Whitney U test; U = 46.0, U = 31.5; FDR‐p < 0.001, FDR‐p < 0.001) but were perceived as equally intense (two‐sided Mann–Whitney U test; U = 157.5; FDR‐p = 0.307) in healthy individuals.

These results align with the results from the olfactory pre‐testing. Regarding the time needed to evaluate the intensity of an odor, individuals with TLE needed more time to rate the intensity of the control solution (two‐sided Mann–Whitney U test; U = 100.0; FDR‐p = 0.021). There were no differences regarding the time needed to rate the intensity of the odors (two‐sided Mann–Whitney U test; both FDR‐p > 0.05). It is of note that individuals with LE needed more time to rate the intensity of eugenol and the control solution than healthy individuals (two‐sided Mann–Whitney U test; U = 42.0; U = 37.0; FDR‐p = 0.038; FDR‐p = 0.038).

3.4. Task‐Driven GLM Analysis

In individuals with TLE and healthy individuals, areas showing significant activations in response to odors belong to the primary olfactory system including the piriform cortex and its downstream cortical and subcortical areas: bilateral amygdala, insula, and orbitofrontal cortex (see Figure 2A,B). When contrasting people with TLE and healthy individuals, we did not observe any differences regarding the activation pattern (two‐sided t‐test, corrected for false‐positive error rate [FPR]). To test whether there were lateralization effects in activation, we compared the lateralization indices of healthy individuals and individuals with TLE. Healthy individuals did not show lateralization of olfactory activation (one sample t‐test vs. 0; p = 0.065; t(19) = −1.957). When comparing all individuals with TLE, and those with left‐sided and right‐sided TLE separately, with healthy individuals, we did not observe any lateralization of olfactory activation in individuals with TLE (two‐sided Mann–Whitney U test; U = 176.0; U = 136.0; U = 40.0; p = 0.525; p = 0.653; p = 0.530). The contrast positive (eugenol)—negative (valeric acid), however, was different between people with TLE and healthy controls. In healthy individuals, there was more activation in the hippocampus, parahippocampal cortex, temporal fusiform cortex, temporal pole, frontal medial cortex, and thalamus bilaterally associated with the positive odor (eugenol) than with the negative odor (valeric acid) (see Figure 2C). In individuals with TLE, the activation was not different between positive and negative stimuli. When comparing the contrast of positive and negative stimuli between people with TLE and healthy individuals, the pattern observed in healthy individuals was also significant relative to people with TLE (see Figure 2D). This means that there is a significantly different activation pattern between positively and negatively valenced stimuli in healthy individuals, which could not be observed in individuals with TLE.

FIGURE 2.

FIGURE 2

Open in a new tab

BOLD signal response to odors. Activation of the olfactory system in healthy individuals (A) and individuals with temporal lobe epilepsy (TLE) (B) during olfactory stimulation. Two‐sided t‐test, corrected for false‐positive error rate (FPR). In statistical testing, no differences regarding the activation pattern between healthy individuals and people with TLE were observed (not shown) (C) Activation in the hippocampus when comparing the activation between both odors (positive/eugenol, negative/valeric acid) in healthy individuals. (D) Hypoactivation in limbic‐system‐related areas including bilateral hippocampus, temporal pole, thalamus, fusiform, medial frontal, and parahippocampal cortices in individuals with TLE when comparing the activation between both odors (positive/eugenol, negative/valeric acid) (IL = ipsilateral, CL = contralateral). Please note that a proportion of healthy individuals are flipped according to the presumed seizure onset zones in individuals with TLE.

4. Discussion

In this study, we show that individuals with TLE have a reduced ability to discriminate odors and assess their valence and familiarity. The bilateral activation pattern to olfactory stimuli in TLE does not differ from that of healthy individuals. However, the response to positive and negative stimuli differs in healthy individuals, which could not be observed in people with TLE.

4.1. Validity of the Olfactory Paradigm

The pattern of olfactory activation in healthy individuals revealed by our analyses aligns with the results of the automatic meta‐analysis of activated areas included in studies using the term “olfactory” on the Neurosynth platform (https://neurosynth.org/analyses/terms/olfactory/). This confirms that the olfactory paradigm used in this study is suitable for investigating the olfactory system.

4.2. Age/Sex‐Related Challenges

Due to challenges in the recruitment of healthy individuals, the control group was not ideally matched concerning age and sex (fewer female and younger participants in the control group). This somewhat imbalanced composition may have an impact on the study results. Indeed, olfactory function decreases with a higher age which is known as presbyosmia (Hüttenbrink et al. 2013). It is caused by atrophy of the olfactory epithelium. About 25% of the population over 50 years old suffer from presbyosmia (Brämerson et al. 2004; Murphy et al. 2002). Our group with individuals with TLE is 45 ± 20 years old [mean ± SD] which means that potentially 25%–50% of the individuals in this group could be affected by presbyosmia. However, only 10% of our cohort reported that they had experienced smelling issues and no individual had hyposmia according to the Sniffin’ Sticks Test. Furthermore, according to the data from population studies, we expected a higher proportion with impairment in olfaction (Hummel et al. 2001) but our results from the pre‐assessment suggest that this is not the case. Another reason, why the effects of presbyosmia are not relevant in our study is the fact that there were no differences according to the perceived intensity of the presented odors. With presbyosmia we would expect to see hypointensity in the pre‐testing in the group with individuals with TLE and also hypoactivation in the experiment in the primary olfactory cortex due to the atrophy of the olfactory epithelium but this was not observed. According to sex‐related effects, the sex‐specific differences are not resolvable in the Sniffin’ Sticks Test (Hummel et al. 2001; Kamath et al. 2024) and a meta‐analysis shows that there are differences in olfaction between women and men but the effect sizes are very small (Sorokowski et al. 2019). Due to the small sample size and the small resolution of the fMRI analysis, it can be expected that the uneven distribution of women and men between both groups does not interfere with our study.

4.3. Lateralization of Olfactory Dysfunction

The lower performance of people with TLE in different domains (discrimination, valence, familiarity) of the behavioral testing confirms previous observations that olfactory dysfunction is a pertinent feature of TLE (Carroll, Richardson, and Thompson 1993; Desai et al. 2015). At the behavioral level, we did not observe a lateralization of olfactory dysfunction using the Sniffin’ Sticks test with respect to the seizure onset side. Also, our imaging results show a bilateral activation of the olfactory system in TLE not different from healthy individuals. Other imaging studies using fMRI (Aguado‐Carrillo et al. 2021) and PET (Ciumas et al. 2008) showed, in disagreement with our results, a certain degree of lateralization towards more activation contralateral to the epileptogenic focus. This may be explained by the fact that our cohort includes a larger proportion of TLE cases with a less clear lateralization of the epileptogenic focus, for instance those with LE. It should, additionally, not be neglected that the Sniffin’ Sticks Test was conducted separately on both nostrils, while it was technically not feasible to separately test the two nostrils in the scanner.

Whether or not the odor discrimination test can be used to localize the side of the epileptogenic focus cannot be said with absolute certainty based on our data and would have to be evaluated in a future study.

4.4. Qualitative Versus Quantitative Olfactory Dysfunction

When evaluating an olfactory stimulus, quantitative and qualitative characteristics can be distinguished. Quantitative characteristics are odor intensity and duration of exposure, qualitative attributes are, for instance, perceived valence and familiarity of an odor or the discrimination between odors. A dysfunction regarding the evaluation of quantitative characteristics is mostly associated with the primary olfactory cortex and upstream structures of the olfactory system, such as the olfactory bulb. An impairment regarding the evaluation of qualitative characteristics is more likely to be related to secondary olfactory areas, such as the hippocampus and other projection areas of the limbic system. This is to be expected due to the high proportion of individuals with LE in our study. Regarding our behavioral results, odors are generally perceived as more neutral and as less familiar by individuals with TLE, the latter especially in individuals with LE. In addition, individuals with TLE had difficulties distinguishing between different odors. The perceived intensity, however, was not different from healthy individuals. This speaks for a qualitative rather than a quantitative dysfunction of the olfactory system in TLE. Together with the imaging results, this provides a coherent picture: The overall activation in the piriform cortex did not differ between people with TLE and healthy individuals. However, we observed that positive and negative valence odors showed different activation patterns in healthy individuals which could not be found in people with TLE. The differences in activation between positive and negative valence odors were further located in secondary olfactory areas linked to the limbic system, which can finally be taken as a correlate of the behaviorally measured quantitative olfactory dysfunction.

4.5. Limitations

From a technical perspective, the limitations of this study are inherent in the complexity of the fMRI paradigm. Although numerous efforts were made to ensure that the stimulation was as objective as possible, for example by monitoring the respiratory cycle, it is not possible to determine exactly when the stimulation of the olfactory system occurred and how strong it was. From a clinical point of view, our TLE sample is heterogeneous with regard to disease stages, medical treatments, and disease entities within the TLE spectrum such as different LE subtypes according to the antibodies. Additionally, almost a half of our individuals with TLE had LE. This reduces the specificity of our results and impairs the generalizability of our study to all individuals with TLE.

Furthermore, the question of the origin of olfactory auras which occurred in three of the individuals with TLE included in this study cannot be answered based on our data. Besides the entorhinal cortex, the orbitofrontal cortex, the amygdala, the insulae and many more, it turned out that the olfactory bulb is the most promising candidate to generate olfactory auras (Sarnat and Flores‐Sarnat 2016). The olfactory bulb itself can be anomalous due to dysgenesis which is associated with agenesis or hypoplasia (Sarnat and Yu 2016). This might be one of many plausible reasons for the appearance of olfactory auras. Another important factor in transmitting these auras could be fibers mainly from the anterior olfactory nucleus which connect the nucleus with the contralateral olfactory bulb via the anterior commissure. Regarding to functional connectivity especially between those structure and the overall brain, we could not find any differences in olfactory activation between the individuals with olfactory auras and the remaining individuals with TLE. This may be due to our limited sample size (n = 3). Moreover, we could not find any structural anomalies in the olfactory bulb of individuals with TLE. To assess the integrity of connecting fiber tracts within the olfactory system, further studies using diffusion tensor imaging are strongly encouraged.

Despite these technical and clinical limitations, this study provides novel insights into the interplay between olfactory dysfunction and TLE by combining dedicated behavioral testing including a valence‐dependent spectrum of olfactory stimuli with a sophisticated olfactory fMRI paradigm presenting eugenol (positive) and valeric acid (negative) as both extreme ends of this spectrum.

5. Conclusions

Altogether, our results suggest a rather qualitative than quantitative dysfunction of the olfactory system in TLE. As a correlate of this, we found that in TLE, unlike in healthy individuals, the activation in secondary olfactory areas as projections of the limbic system did not differ between odors with positive and negative valence. Our behavioral and imaging results showed no lateralization of olfactory activation and the question of whether testing of the olfactory system may inform localization of the side of the epileptogenic focus remains open. Furthermore, these novel insights help to categorize olfactory symptoms as a rather systemic impairment on a network level. Finally, our findings also contribute to the understanding of the physiological process of olfactory perception in healthy individuals.

Ethics Statement

The study protocol was approved by the institutional review board of the Medical Faculty of the University of Bonn (No. 434/22) and is in accordance with relevant guidelines and regulations.

Consent

All participants provided written consent.

Conflicts of Interest

Markus Schmidt, Tobias Bauer, Marcel Kehl, Anna Minarik, Lennart Walger, Johannes Schultz, Martin S. Otte, Peter Trautner, Christian Hoppe, Tobias Baumgartner, Louisa Specht‐Riemenschneider, and Florian Mormann declare no conflicts of interest. Alexander Radbruch serves on the scientific advisory boards for GE Healthcare, Bracco, Bayer, Guerbet, and AbbVie; has received speaker honoraria from Bayer, Guerbet, Siemens, and Medscape; and is a consultant for, and has received institutional study support from, Guerbet and Bayer. Rainer Surges has received fees as speaker or for serving on the advisory board from Angelini, Arvelle, Bial, Desitin, Eisai, Janssen‐Cilag GmbH, LivaNova, Novartis, Precisis GmbH, UCB Pharma, UnEEG and Zogenix and grants from the Deutsche Forschungsgemeinschaft (DFG), the Bundesministerium für Bildung und Forschung (BMBF), the Bundesministerium für Gesundheit, and the Marga and Walter Boll Stiftung. These activities were not related to the content of this manuscript. None of the previously mentioned activities were related to the content of this manuscript. Theodor Rüber has received fees as a speaker from Eisai. He declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The results were presented on a poster at the BONFOR‐Symposium 2024 in Bonn, Germany, at the DGfE 2024 in Offenburg, Germany, and at the EEC 2024 in Rome, Italy.

Supporting information

Data S1.

HBM-45-e70061-s001.pdf (619KB, pdf)

Acknowledgments

We would like to thank the Core Facility human 3T MRI of the Medical Faculty at the University of Bonn for providing support and instrumentation.

Funding: Markus Schmidt and Tobias Bauer were supported by the BONFOR Research Commission of the Medical Faculty of the University of Bonn (Grant Numbers 2022‐4‐31 and 2022‐1A‐21).

Markus Schmidt and Tobias Bauer contributed equally to this study.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

  1. Abraham, A. , Pedregosa F., Eickenberg M., et al. 2014. “Machine Learning for Neuroimaging With Scikit‐Learn.” Frontiers in Neuroinformatics 8: 14. 10.3389/fninf.2014.00014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aguado‐Carrillo, G. , Saucedo‐Alvarado P. E., Cuellar‐Herrera M., et al. 2021. “Olfactory Function in Patients With Temporal Lobe Epilepsy; Correlation of Functional Magnetic Resonance Imaging for Olfaction With the Laterality of the Epileptic Focus.” Epilepsy Research 178: 106807. 10.1016/j.eplepsyres.2021.106807. [DOI] [PubMed] [Google Scholar]
  3. Benjamini, Y. , and Hochberg Y.. 1995. “Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing.” Journal of the Royal Statistical Society: Series B: Methodological 57, no. 1: 289–300. 10.1111/j.2517-6161.1995.tb02031.x. [DOI] [Google Scholar]
  4. Bernhardt, B. , Hong S.‐J., Bernasconi A., and Bernasconi N.. 2013. “Imaging Structural and Functional Brain Networks in Temporal Lobe Epilepsy.” Frontiers in Human Neuroscience 7: 624. 10.3389/fnhum.2013.00624. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Binder, J. R. , Swanson S. J., Hammeke T. A., et al. 1996. “Determination of Language Dominance Using Functional MRI: A Comparison With the Wada Test.” Neurology 46, no. 4: 978–984. 10.1212/wnl.46.4.978. [DOI] [PubMed] [Google Scholar]
  6. Borger, V. , Schneider M., Taube J., et al. 2021. “Resection of Piriform Cortex Predicts Seizure Freedom in Temporal Lobe Epilepsy.” Annals of Clinical Translational Neurology 8, no. 1: 177–189. 10.1002/acn3.51263. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brämerson, A. , Johansson L., Ek L., Nordin S., and Bende M.. 2004. “Prevalence of Olfactory Dysfunction: The Skövde Population‐Based Study.” Laryngoscope 114, no. 4: 733–737. 10.1097/00005537-200404000-00026. [DOI] [PubMed] [Google Scholar]
  8. Carroll, B. , Richardson J. T. E., and Thompson P.. 1993. “Olfactory Information Processing and Temporal Lobe Epilepsy.” Brain and Cognition 22, no. 2: 230–243. 10.1006/brcg.1993.1036. [DOI] [PubMed] [Google Scholar]
  9. Cheng, H. , Wang Y., Chen J., and Chen Z.. 2020. “The Piriform Cortex in Epilepsy: What We Learn From the Kindling Model.” Experimental Neurology 324: 113137. 10.1016/j.expneurol.2019.113137. [DOI] [PubMed] [Google Scholar]
  10. Ciumas, C. , Lindström P., Aoun B., and Savic I.. 2008. “Imaging of Odor Perception Delineates Functional Disintegration of the Limbic Circuits in Mesial Temporal Lobe Epilepsy.” NeuroImage 39, no. 2: 578–592. 10.1016/j.neuroimage.2007.09.004. [DOI] [PubMed] [Google Scholar]
  11. Deichmann, R. , Gottfried J. A., Hutton C., and Turner R.. 2003. “Optimized EPI for fMRI Studies of the Orbitofrontal Cortex.” NeuroImage 19, no. 2: 430–441. 10.1016/S1053-8119(03)00073-9. [DOI] [PubMed] [Google Scholar]
  12. Desai, M. , Agadi J. B., Karthik N., Praveenkumar S., and Netto A. B.. 2015. “Olfactory Abnormalities in Temporal Lobe Epilepsy.” Journal of Clinical Neuroscience 22, no. 10: 1614–1618. 10.1016/j.jocn.2015.03.035. [DOI] [PubMed] [Google Scholar]
  13. Donoshita, Y. , Choi U.‐S., Ban H., and Kida I.. 2021. “Assessment of Olfactory Information in the Human Brain Using 7‐Tesla Functional Magnetic Resonance Imaging.” NeuroImage 236: 118212. 10.1016/j.neuroimage.2021.118212. [DOI] [PubMed] [Google Scholar]
  14. Doty, R. L. , Tourbier I., Neff J. K., et al. 2018. “Influences of Temporal Lobe Epilepsy and Temporal Lobe Resection on Olfaction.” Journal of Neurology 265, no. 7: 1654–1665. 10.1007/s00415-018-8891-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dupont, S. , Samson Y., Nguyen‐Michel V.‐H., et al. 2015. “Are Auras a Reliable Clinical Indicator in Medial Temporal Lobe Epilepsy With Hippocampal Sclerosis?” European Journal of Neurology 22, no. 9: 1310–1316. 10.1111/ene.12747. [DOI] [PubMed] [Google Scholar]
  16. Engel, J. 1996. “Introduction to Temporal Lobe Epilepsy.” Epilepsy Research 26, no. 1: 141–150. 10.1016/S0920-1211(96)00043-5. [DOI] [PubMed] [Google Scholar]
  17. Esteban, O. , Ciric R., Finc K., et al. 2020. “Analysis of Task‐Based Functional MRI Data Preprocessed With fMRIPrep.” Nature Protocols 15, no. 7: 2186–2202. 10.1038/s41596-020-0327-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Esteban, O. , Markiewicz C. J., Blair R. W., et al. 2019. “fMRIPrep: A Robust Preprocessing Pipeline for Functional MRI.” Nature Methods 16, no. 1: 111–116. 10.1038/s41592-018-0235-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Galovic, M. , Baudracco I., Wright‐Goff E., et al. 2019. “Association of Piriform Cortex Resection With Surgical Outcomes in Patients With Temporal Lobe Epilepsy.” JAMA Neurology 76, no. 6: 690–700. 10.1001/jamaneurol.2019.0204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Georgiopoulos, C. , Witt S. T., Haller S., et al. 2018. “Olfactory fMRI: Implications of Stimulation Length and Repetition Time.” Chemical Senses 43, no. 6: 389–398. 10.1093/chemse/bjy025. [DOI] [PubMed] [Google Scholar]
  21. Gorgolewski, K. J. , Auer T., Calhoun V. D., et al. 2016. “The Brain Imaging Data Structure, A Format for Organizing and Describing Outputs of Neuroimaging Experiments.” Scientific Data 3, no. 1: 160044. 10.1038/sdata.2016.44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gottfried, J. A. 2015. “Structural and Functional Imaging of the Human Olfactory System.” In Handbook of Olfaction and Gustation, edited by Doty R. L., 279–304. Hoboken, New Jersey: John Wiley & Sons, Ltd. 10.1002/9781118971758.ch13. [DOI] [Google Scholar]
  23. Gottfried, J. A. , Deichmann R., Winston J. S., and Dolan R. J.. 2002. “Functional Heterogeneity in Human Olfactory Cortex: An Event‐Related Functional Magnetic Resonance Imaging Study.” Journal of Neuroscience 22, no. 24: 10819–10828. 10.1523/JNEUROSCI.22-24-10819.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Gottfried, J. A. , Winston J. S., and Dolan R. J.. 2006. “Dissociable Codes of Odor Quality and Odorant Structure in Human Piriform Cortex.” Neuron 49, no. 3: 467–479. 10.1016/j.neuron.2006.01.007. [DOI] [PubMed] [Google Scholar]
  25. Graus, F. , Titulaer M. J., Balu R., et al. 2016. “A Clinical Approach to Diagnosis of Autoimmune Encephalitis.” Lancet Neurology 15, no. 4: 391–404. 10.1016/S1474-4422(15)00401-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Harms, A. , Bauer T., Witt J.‐A., et al. 2023. “Mesiotemporal Volumetry, Cortical Thickness, and Neuropsychological Deficits in the Long‐Term Course of Limbic Encephalitis.” Neurology Neuroimmunology & Neuroinflammation 10, no. 4: e200125. 10.1212/NXI.0000000000200125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Howard, J. D. , Plailly J., Grueschow M., Haynes J.‐D., and Gottfried J. A.. 2009. “Odor Quality Coding and Categorization in Human Posterior Piriform Cortex.” Nature Neuroscience 12, no. 7: 932–938. 10.1038/nn.2324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hummel, T. , Rosenheim K., Konnerth C.‐G., and Kobal G.. 2001. “Screening of Olfactory Function With a Four‐Minute Odor Identification Test: Reliability, Normative Data, and Investigations in Patients With Olfactory Loss.” Annals of Otology, Rhinology and Laryngology 110, no. 10: 976–981. 10.1177/000348940111001015. [DOI] [PubMed] [Google Scholar]
  29. Hüttenbrink, K.‐B. , Hummel T., Berg D., Gasser T., and Hähner A.. 2013. “Olfactory Dysfunction: Common in Later Life and Early Warning of Neurodegenerative Disease.” Deutsches Ärzteblatt International 110, no. 1–2: 1–7. 10.3238/arztebl.2013.0001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hwang, B. Y. , Mampre D., Penn R., Anderson W. S., Kang J., and Kamath V.. 2020. “Olfactory Testing in Temporal Lobe Epilepsy: A Systematic Review.” Current Neurology and Neuroscience Reports 20, no. 12: 65. 10.1007/s11910-020-01083-x. [DOI] [PubMed] [Google Scholar]
  31. Hwang, B. Y. , Mampre D., Tsehay Y. K., et al. 2022. “Piriform Cortex Ablation Volume Is Associated With Seizure Outcome in Mesial Temporal Lobe Epilepsy.” Neurosurgery 91, no. 3: 414. 10.1227/neu.0000000000002041. [DOI] [PubMed] [Google Scholar]
  32. Jafar, A. , Lasso A., Shorr R., Hutton B., and Kilty S.. 2021. “Olfactory Recovery Following Infection With COVID‐19: A Systematic Review.” PLoS One 16, no. 11: e0259321. 10.1371/journal.pone.0259321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kamath, V. , Chen H., Shrestha S., et al. 2024. “Normative Data for the 12‐Item Sniffin’ Sticks Odor Identification Test in Older Adults.” Archives of Clinical Neuropsychology 39, no. 3: 335–346. 10.1093/arclin/acad080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Khurshid, K. , Crow A. J. D., Rupert P. E., et al. 2019. “A Quantitative Meta‐Analysis of Olfactory Dysfunction in Epilepsy.” Neuropsychology Review 29, no. 3: 328–337. 10.1007/s11065-019-09406-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Motoki, A. , Akamatsu N., Fumuro T., et al. 2021. “Characteristics of Olfactory Dysfunction in Patients With Temporal Lobe Epilepsy.” Epilepsy & Behavior 125: 108402. 10.1016/j.yebeh.2021.108402. [DOI] [PubMed] [Google Scholar]
  36. Murphy, C. , Schubert C. R., Cruickshanks K. J., Klein B. E. K., Klein R., and Nondahl D. M.. 2002. “Prevalence of Olfactory Impairment in Older Adults.” JAMA 288, no. 18: 2307–2312. 10.1001/jama.288.18.2307. [DOI] [PubMed] [Google Scholar]
  37. Sarnat, H. B. , and Flores‐Sarnat L.. 2016. “Might the Olfactory Bulb Be an Origin of Olfactory Auras in Focal Epilepsy?” Epileptic Disorders 18, no. 4: 344–355. 10.1684/epd.2016.0869. [DOI] [PubMed] [Google Scholar]
  38. Sarnat, H. B. , and Yu W.. 2016. “Maturation and Dysgenesis of the Human Olfactory Bulb.” Brain Pathology (Zurich, Switzerland) 26, no. 3: 301–318. 10.1111/bpa.12275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sorokowski, P. , Karwowski M., Misiak M., et al. 2019. “Sex Differences in Human Olfaction: A Meta‐Analysis.” Frontiers in Psychology 10: 242. 10.3389/fpsyg.2019.00242. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Taşcı, İ. , Balgetir F., Müngen B., and Gönen M.. 2021. “Epileptic Olfactory Auras: A Clinical Spectrum.” Neurological Sciences 42, no. 8: 3397–3401. 10.1007/s10072-020-04999-x. [DOI] [PubMed] [Google Scholar]
  41. Taylor, P. A. , Reynolds R. C., Calhoun V., et al. 2023. “Highlight Results, Don't Hide Them: Enhance Interpretation, Reduce Biases and Improve Reproducibility.” NeuroImage 274: 120138. 10.1016/j.neuroimage.2023.120138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Vaughan, D. N. , and Jackson G. D.. 2014. “The Piriform Cortex and Human Focal Epilepsy.” Frontiers in Neurology 5: 259. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1.

HBM-45-e70061-s001.pdf (619KB, pdf)

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES