Neuroticism in Temporal Lobe Epilepsy is Associated with Altered Limbic-Frontal Lobe Resting State Functional Connectivity

Abstract

Neuroticism, a core personality trait characterized by a tendency towards experiencing negative affect, has been reported to be higher in people with temporal lobe epilepsy (TLE) compared to healthy individuals. Neuroticism is a known predictor of depression and anxiety, which also occur more frequently in people with TLE. The purpose of this study was to identify abnormalities in whole brain resting state functional connectivity in relation to neuroticism in people with TLE, and to determine the degree of unique versus shared patterns of abnormal connectivity in relation to elevated symptoms of depression and anxiety. 93 individuals with TLE (55-females) and 40 healthy controls (18-females) from the Epilepsy Connectome Project (ECP) completed measures of neuroticism, depression and anxiety which were all significantly higher in people with TLE compared to controls. Resting state functional connectivity was compared between controls and high and low neuroticism TLE groups using analysis of variance (ANOVA) and t-test. In secondary analyses the same analytics were performed using measures of depression and anxiety and the unique variance in resting state connectivity associated with neuroticism independent of symptoms of depression and anxiety identified. Increased neuroticism was significantly associated with hyposynchrony between the right hippocampus and Brodmann area 9 (region of prefrontal cortex) (p<0.005), representing a unique relationship independent of symptoms of depression and anxiety. Hyposynchrony of connection between the right hippocampus and Brodmann area 47 (anterior frontal operculum) was associated with high neuroticism but also with higher depression and anxiety scores (p<0.05), making it a shared abnormal connection for the three measures. In conclusion, increased neuroticism exhibits both unique and shared patterns of abnormal functional connectivity with depression and anxiety symptoms between regions of the mesial temporal and frontal lobe.

1. Introduction

Population and community-based investigations as well as numerous clinical cohort studies have demonstrated elevated psychiatric comorbidity in people with epilepsy including temporal lobe epilepsy (TLE) [1-4]. Much less work has been directed to differences in normal personality traits, particularly those composing the so-called “Big 5” traits (agreeableness, conscientiousness, extraversion, openness and neuroticism). Of the Big 5 traits, neuroticism, characterized by a tendency towards experiencing negative affect, has been investigated the most in the epilepsy literature and reported to be higher in individuals with TLE than healthy individuals [5,6], In persons with epilepsy it has also been established that neuroticism is a predictor and risk factor for depression and anxiety [7,8] or their symptoms [9,10], is negatively correlated with post-operative subjective cognitive self-appraisal and psychological adjustment [11,12], and is positively associated with increased perceived stigma [13] as well as poorer social well-being [14]. In the general population, elevated neuroticism also has implications for suicidal ideation [15] and behavioral dysregulation, which refers to behavioral strategies that are harmful such as drinking alcohol as a coping mechanism, binge eating, among others [16].

The neurobiological correlates of neuroticism have been investigated predominantly in healthy populations where a relationship between neuroticism and altered brain structure has been demonstrated. Specifically, salient findings indicate reduced volume in prefrontal regions affecting the left superior frontal gyrus [17], dorsomedial PFC, and other regions including the medial cingulate gyrus, precentral gyrus and the medial temporal lobe including the hippocampus [18], with reduced cortical thickness in the orbitofrontal cortex with extension into the anterior medial frontal cortex [19] in association with elevated neuroticism scores. In epilepsy, only one investigation has examined the structural correlates of neuroticism where a pattern similar to the aforementioned healthy controls was observed with increased neuroticism scores in TLE associated with significant volumetric reductions in the left superior and middle frontal gyrus, left anterior insula, left precentral gyrus, left precuneus, left fusiform gyrus, left lateral parietal-occipital cortex, left amygdala and bilateral hippocampi [6]. Taken together, these studies suggest partial but not complete overlap of identified structural abnormalities associated with elevated neuroticism scores in controls and individuals withTLE.

Regarding patterns of functional connectivity, in the general population high neuroticism scores are characterized by a “less than optimal functional network organization”, showing patterns of functional disconnection in the frontoparietal network, somatosensory-motor network, and visual subnetwork—with a positive association with the efficiency of the affective subnetwork [20]. Seed based functional connectivity analysis has revealed disconnection between the amygdala and the temporal poles, insula and superior temporal gyrus with increased connectivity between the amygdala and precuneus [21].

More generally, behavioral and cognitive disorders are now felt to not be merely secondary effects of epilepsy, but rather complications that may share common mechanisms with epilepsy [22,23]. It has been argued that an important task for the field is to use basic neuroscience and neuroimaging research to better understand the precise network abnormalities underlying neurocognitive and behavioral comorbidities of epilepsy markers, hopefully contributing to future early identification and treatment strategies to improve quality of life [23]. Alterations in functional brain connectivity associated with increased neuroticism in epilepsy generally and TLE in particular have not been investigated to date. Our findings in structural differences in people with TLE with high neuroticism scores [24], led us to consider functional components to this abnormality. Recent findings on resting functional connectivity are revealing its potential to be used as individualized “fingerprint” [25]. A functional connectivity approach can help determine the presence and nature of abnormal underlying circuitry to better understand elevated neuroticism in people with TLE. Therefore, the first aim of this investigation was to identify patterns of disrupted functional connectivity associated with neuroticism scores in people with TLE. We used whole brain resting-state functional magnetic resonance imaging (rs-fMRI) to identify differences in resting functional connectivity patterns between controls and people with TLE exhibiting low versus high neuroticism scores. Second, given the strong relationship between performance on measures of neuroticism, depression and anxiety in both the general population [26,27] and individuals with TLE [7,10,28,29], and the high incidence of depression and anxiety symptoms and formal psychiatric diagnoses in TLE [2,30-32], secondary analyses were performed with measures of depression and anxiety with special interest in identifying both unique and shared patterns of anomalies in whole-brain resting state brain functional connectivity in relation to neuroticism scores.

We address the results in terms of synchrony between regions, with synchrony defined as the temporal coincidence of two regions [33]. Hypersynchronous connectivity implies high temporal correlation (either positive or negative) between two brain regions, while hyposynchronous suggests functional disconnect, or low temporal correlation between regions. In this case, hypersynchrony does not refer to seizure activity, but increased synchronization of resting state brain signals.

2. Methods and Materials

2.1. Participants

Study participants included 93 individuals with TLE (38.2 ± 10.9 years old, 55 females) and 40 healthy controls (33.9 ± 10.2 years old, 18 females) from the Epilepsy Connectome Project (ECP) [34,35]. ECP is a multi-site research project involving the Medical College of Wisconsin (MCW) and the University of Wisconsin-Madison. The Institutional Review Board (IRB) at MCW reviewed and approved the project. All participants provided written informed consent.

Individuals with TLE enrolled in the ECP are between the ages of 18 and 60, demonstrate estimated full-scale Intelligence quotient (IQ) at or above 70, speak English fluently, and have no medical contraindications to MRI. They have a diagnosis of TLE supported by two or more of the following: 1) described or observed clinical semiology consistent with seizures of temporal lobe origin, 2) electroencephalogram (EEG) evidence of either temporal intermittent rhythmic delta activity (TIRDA) or temporal lobe epileptiform discharges, 3) temporal lobe onset of seizures captured on video EEG monitoring study, or 4) MRI evidence of mesial temporal sclerosis or hippocampal atrophy. Individuals with TLE and any of the following were excluded: 1) Presence of any lesions other than mesial temporal sclerosis causative for seizures on 3 Tesla MRI, 2) an active infectious/autoimmune/inflammatory etiology of seizures, either suspected by treating clinician or documented through laboratory testing or response to immunosuppressive therapy.

Healthy controls are between the ages of 18 and 60. Exclusion criteria for the healthy controls included: Edinburgh Laterality (Handedness) Quotient less than +50; primary language other than English; history of any learning disability, brain injury or illness, substance abuse, or major psychiatric disorders (major depression, bipolar disorder, or schizophrenia); current use of vasoactive medications; and any medical contraindications to MRI.

2.2. Behavioral Testing

All individuals with TLE and 28 healthy controls completed the NEO Five-Factor-Inventory (NEO-FFI) to measure neuroticism. Age- and gender-corrected standardized T-scores were used for all analysis. The NEO Five Factor Model [36] establishes a T-score scale with a mean of 50 based on raw scores from a healthy population, with suggested cutoffs for very low, low, average, high and very high at z = −1.5, −0.5, +0.5, +1.5 standard deviations away from the mean respectively.

All participants also completed the Achenbach Adult Self-Report (ASR) to measure DSM (Diagnostic and Statistical Manual of Mental Disorder) oriented depression and anxiety symptoms using the Depressive Problems and Anxiety Problems scales. The TLE participant group was divided into high vs. low scoring groups based on each corrected T-score, and Table 1 summarizes the information on the three groups (including healthy controls) within each variable. ASR depression and anxiety T-scores of 64 and under are considered normal, 65-69 are classified as borderline clinical range, and at or above 70 are classified as clinical. With most participants falling below clinical levels of anxiety (90%) and depression (83%), our results represent connectivity differences detected in sub-clinical depression or anxiety. Thus, when referring to depression and anxiety throughout this paper we are referring to symptoms of those disorders as opposed to formal diagnoses. That said, secondary analyses were undertaken using ASR clinical cut points.

Table 1.

For each test score, the temporal lobe epilepsy (TLE) group was split into high-and low-scoring (normal) groups based on the median score. This table summarizes the demographics of each group. The mean age and gender ratio did not significantly differ (p > 0.15) between groups within each variable.

Score	Group	N	Age (years)	Sex (M / F)	Scores
Neuroticism	High TLE	42	35.5 ± 11.3	19 / 23	68.5 ± 5.6
	Low TLE	41	38.7 ± 10.6	18 / 23	45.7 ± 8.5
	Control	28	34.8 ± 11.6	17 / 11	50.0 ± 9.4
Depression	High TLE	43	37.2 ± 10.8	18 / 25	67.4 ± 7.3
	Low TLE	43	37.0 ± 10.4	18 / 25	51.3 ± 1.7
	Control	40	33.9 ± 10.2	22 / 18	54.5 ± 6.5
Anxiety	High TLE	40	37.3 ± 11.4	17 / 23	64.1 ± 7.2
	Low TLE	43	36.3 ± 9.7	17 / 26	50.8 ± 1.1
	Control	40	33.9 ± 10.2	22 / 18	54.0 ± 5.5

2.3. MRI Image Acquisition

MRI was performed on 3T GE (General Electric) 750 scanners at both institutions. T1-weighted structural images were acquired using magnetization prepared gradient echo sequence (MPRAGE, TR/TE=604ms/2.516ms, TI=1060.0ms, flip angle=8°, FOV=25.6cm, 0.8mm isotropic). Rs-fMRI images were acquired using whole-brain simultaneous multi-slice (SMS) imaging [37] (8 bands, 72 slices, TR/TE=802ms/33.5ms, flip angle=50°, matrix=104×104, FOV=20.8cm, 2.0mm isotropic) and a Nova 32-channel receive coil. The participants were asked to fixate on a white cross at the center of a black screen during the scans for better reliability [38].

2.4. Image Processing

MRI images were preprocessed using the Human Connectome Project (HCP) minimal processing pipelines [39] which are primarily based on FreeSurfer [40] and FSL (Functional MRI of the brain Software Library) [41]. In brief, the function of this pipeline is to non-linearly register T1-weighted images to the MNI (Montreal Neurological Institute) space, segment the volume into predefined structures, reconstruct white and pial cortical surfaces, and perform FreeSurfer's standard folding-based surface registration to a surface atlas (the “fsaverage” template). Then, it is used to remove nonlinear spatial distortions in the rs-fMRI images using spin echo unwarping maps, realign volumes to compensate for subject motion, register to the structural images, reduce the bias field, normalize the 4D image to a global mean, mask the data with the final brain mask and map the voxels within the cortical gray matter ribbon onto the native cortical surface space. More details on the HCP processing pipelines can be found in Glasser et al. [39].

Additional pre-processing was performed on the rs-fMRI images using AFNI (Analysis of Functional Neuro-Images) [42]. This included motion regression using 12 motion parameters, regression-based removal of signal changes in the white matter, cerebrospinal fluid (CSF), global signal, and band-pass filtering (0.01-0.1Hz) [43].

Using the Connectome Workbench (version 1.1.1) [44], time-series data from four 5-minute rs-fMRI scans acquired in two sessions were concatenated. 360 time-series from Glasser Parcellation [45] plus 19 FreeSurfer subcortical regions [46] were extracted per subject. MATLAB (Matrix Laboratory) R2018a was used to calculate pairwise Pearson correlations between 379 time-series and to perform Fisher-z transformations for generating resting state connectivity matrices.

2.5. Group Comparison

Resting state connectivity matrices were compared across the three groups (individuals with TLE with high and low neuroticism scores and healthy controls), first with one-way analysis of variance (ANOVA), then for connections that showed significance, two-sample two-sided t-tests to confirm differences between the two median-split TLE groups. We only focused on connections showing significant differences both between the two median-split TLE groups, and also between control and high-scoring TLE groups. Benjamini-Hochberg false discovery rate adjustment was used to correct p-values for multiple comparisons in all analyses [47].

Secondary analyses were performed using the depression and anxiety scores following the same analytic method described for neuroticism. Furthermore, to consider the collinearity that exists between the measures (neuroticism, depression and anxiety), the unique variance of neuroticism score was calculated that was orthogonal to both depression and anxiety scores using a generalized linear model (neuroticism ~ depression + anxiety [+ residual or unique variance of neuroticism]). Then the same analysis was performed using this variance as a score. The mean age and gender ratio were statistically matched between groups in comparison per score (p>0.15).

In addition, in order to confirm the robustness of the findings across specific cut-points, the previously described analysis was performed using suggested standardized cutoffs [36]: between individuals with average or lower (z< +0.5, T-score < 55) versus high (z> +1.0, T-score > 60) neuroticism scores. Also, for anxiety and depression, clinical cutoffs were tested, comparing those with borderline or above (score ≥ 65) versus normal groups (score ≤ 55) individuals. These analyses revealed similar results as the median split approach attesting to the robustness of effect (Supplementary Figure 1).

3. Results

3.1. Behavior Scores

Lilliefors test showed that neuroticism scores for individuals with TLE were not normally distributed (p < 0.03), thus non-parametric tests were used. Individuals with TLE showed significantly higher neuroticism (p=0.008, N=111, Wilcoxon rank sum test) compared to controls (Figure 1). Neuroticism T-scores were highly correlated with depression and anxiety scores both in healthy controls (p<0.01, N=28, Spearman correlation,) and TLE participants (p<0.001, N=93).

Figure 1.

Boxplots of the three behavior scores. Scores of individuals with temporal lobe epilepsy (TLE) were significantly higher than healthy controls in all three scores (p’s < 0.05, Wilcoxon rank sum test). Median split was used to separate the individuals with TLE into high-scoring vs. low-scoring groups.

Depression and anxiety scores had a minimum T score of 50, which caused the distributions to be non-normal (p<0.001). Non-parametric tests were used for subsequent analysis. Individuals with TLE showed significantly higher depression (p=0.006, N=126, Wilcoxon rank sum test) and anxiety (p=0.04, N=123) scores compared to controls (Figure 1). None of the scores were significantly associated with laterality of EEG abnormality, the number of anti-epileptic drugs (AED) (p’s>0.15, Kruskal-Wallis test), or the age of recurrent seizure onset (p’s>0.15, Spearman correlation).

3.2. Comparisons of Resting Connectivity

Individuals with TLE with high neuroticism showed significantly (corrected p<0.05) decreased resting connectivity (hyposynchrony) in seven connections, compared to both individuals with TLE with low neuroticism and healthy controls (Table 2). Two of these were right hippocampal connections.

Table 2.

Resting state connections that showed significant hyposynchrony (corrected p < 0.05) in individuals with temporal lobe epilepsy (TLE) with high scores compared to both those with low scores, and healthy controls are summarized. The regions are based on the Glasser Parcellation (Glasser et al., 2016).

Significant Connections (Neuroticism)		p (FDR)
*[L] Area 9 Anterior (medial prefrontal)	[R] Hippocampus	0.003
[L] Presubiculum	[R] Area 9 Middle (prefrontal)	0.016
[L] Lateral Area 7P (region in medial superior parietal lobule)	[L] Area Lateral Intraparietal Dorsal	0.016
[L] Area 9 Middle (prefrontal)	[L] Presubiculum	0.016
*[R] Area 47m (orbitofrontal)	[R] Hippocampus	0.027
[R] Area 8Av	[R] Area OP4/PV	0.028
[L] Ventral Area 6	[R] Area 8Av (dorsolateral prefrontal)	0.040
Significant Connections (Depression)		p (FDR)
*[L] Area 9 Anterior (medial prefrontal)	[R] Hippocampus	<0.001
*[R] Area 47m (orbitofrontal)	[R] Hippocampus	0.018
[L] Area TG Dorsal (dorsal temporal pole)	[R] Area 47m (orbitofrontal)	0.021
[L] Area 9 Posterior (prefrontal)	[R] Hippocampus	0.023
[L] Area TG Dorsal (dorsal temporal pole)	[L] Area FST (temporal/occipital junction)	0.025
[L] Area IFSa (inferior frontal cortex)	[R] Primary Auditory Cortex	0.039
[L] Area Frontal Opercular 5	[R] Area 33 Prime (posterior cingulate)	0.039
[L] Area Anterior 10p (orbitofrontal)	[R] Hippocampus	0.039
Significant Connections (Anxiety)		p (FDR)
*[R] Area 47m (orbitofrontal)	[R] Hippocampus	0.012

^*Connections appear more than once.

In the secondary analysis of depression scores, eight connections showed significant differences (corrected p<0.05), all of which were hyposynchronous in people with TLE with high depression scores. Four of these were right hippocampal connections. The strongest difference for both neuroticism and depression scores was found for the connection between the right hippocampus and left area 9 anterior (a part of Brodmann area [BA] 9—a region of medial prefrontal cortex) (corrected p<0.005). This connection was also the only connection showing significant hyposynchrony in individuals with TLE with high unique variance of neuroticism (corrected p=0.01), which was orthogonal to depression and anxiety scores. Its strength was not significantly associated with left or right hippocampus volume (p=0.70, 0.19, Pearson correlation), age of recurrent seizure onset (p=0.29), laterality of EEG abnormality (p=0.11, one-way ANOVA), nor the number of AEDs (p=0.22) in ourTLE participants.

In the secondary analysis of anxiety scores, one connection showed a significant difference (corrected p<0.05) with hyposynchrony in individuals with TLE with high anxiety scores, which was between the right hippocampus and right area 47m (a part of BA 47 in orbitofrontal cortex). This connection was found in all three comparisons (Table 2, Figure 3). Its strength was not significantly associated with left or right hippocampus volume (p=0.68, 0.69, Pearson correlation), age of recurrent seizure onset (p=0.52), laterality of EEG abnormality (p=0.57, one-way ANOVA) nor the number of AEDs (p=0.09) in individuals with TLE.

Figure 3.

Connection strength (Fisher z-transformed) between right hippocampus to left area 9 anterior (A) showed the most significant hyposynchrony based on neuroticism and depression scores. Connection strength between right hippocampus and right area 47m (B) showed significance based on all three behavior scores. Hyposynchrony in these connections are associated with high neuroticism and depression in individuals with temporal lobe epilepsy (TLE).

The main findings of this analysis were invariant (Supplementary Figure 1) when clinically relevant scores were used as cutoffs. Mean T scores using this approach were: (1) Neuroticism: high (>60, N=41, 68.7±5.6) vs. average or lower (<55, N=41, 43.6±7.0), (2) Depression: borderline or above (≥65, N=30 [32%], 71.2±5.7) vs. normal (≥55, N=42, 51.2±1.6) and (3) Anxiety: borderline or above (≥65, N=16 [17%], 71,7±4.1) vs. normal (≤55, N=56, 5112±1.5).

4. Discussion

This investigation found significantly elevated scores across measures of neuroticism, depression and anxiety in participants with TLE compared to controls, elevations that were associated with disrupted patterns of mesial temporal-frontal lobe resting state connectivity. Neuroticism, the core focus of this study, exhibited both unique and as well as shared patterns of disrupted connectivity with depression and anxiety scores. These points will be addressed in turn.

4.1. Neuroticism and Its Association with Depression and Anxiety Scores

The dimension of neuroticism, a core Big 5 personality trait, is also referred to as negative emotionality or negative affectivity. The present study, in agreement with previous results [5,6], showed increased neuroticism scores in TLE compared to healthy individuals using the NEO-FFI.

Neuroticism has been positively associated with depression and anxiety symptoms in both the general population [26,27] and in people with epilepsy [29]. Although scale overlap for symptoms of depression and anxiety scores with symptoms of neuroticism suggests incomplete independence between measures, a number of studies found significant associations between neuroticism and depression even after controlling for shared test items and concurrent depressive states [48]. In secondary analyses in this investigation, higher depression and anxiety symptoms in individuals with TLE compared to controls was found as expected, and depression and anxiety scores were significantly positively correlated with neuroticism both in healthy controls (p<0.01) and individuals with TLE (p<0.001). In agreement with previous studies, our results suggest a positive relationship between neuroticism and depression and anxiety scores in individuals with TLE.

4.2. Frontal Lobe Contributions to Behavioral Comorbidities in TLE

Early studies examining personality and behavioral changes as well as DSM/ICD (International Classification of Diseases) psychiatric diagnoses in TLE focused predominantly on the mesial temporal lobe as a window to understand the neurobiology of disordered behavior and emotional regulation. However, considerable research now shows that extratemporal anomalies contribute to behavioral, personality and psychiatric disorders in TLE [6,49-51], especially frontal regions. Specific to epilepsy, pre-surgical cognitive measures of frontal lobe dysfunction (e.g., tests of higher level executive ability) have been found to moderate the magnitude and direction of mood change after TLE surgery [52]. Furthermore, people with TLE and a history of depressive symptoms exhibit pre-operative focal hypometabolism in the ipsilateral orbitofrontal cortex compared to those without such a history [53]. Individuals with TLE who had a Major Depressive Episode postoperatively exhibited hypometabolism in the same region [51]. Taken together, these studies demonstrate the involvement of extratemporal regions, particularly regions of the frontal lobe, in the psychiatric comorbidities or their symptoms in TLE, pointing to the relevance of a whole-brain functional connectivity approach. Furthermore, the studies suggest that functional phenotypes can be used to predict treatment outcomes.

Previous research in our group [6] also showed that increasing neuroticism scores were associated with significantly decreased volumes of the left superior/middle frontal cortex, anterior insula, precentral gyrus, lateral parietal-occipital cortex, superior temporal gyrus, precuneus and fusiform gyrus in individuals with TLE. Here we demonstrate for the first time that increased neuroticism scores in TLE is associated with patterns of abnormal resting state connectivity between mesial temporal and frontal regions, some abnormal connections unique to neuroticism and others shared with elevated symptoms of depression and anxiety. This suggests that people with TLE present both temporal and extra-temporal structural and functional abnormalities in relation to scores on measures of neuroticism, depression and anxiety.

4.3. Resting State Hyposynchrony

Bear’s early sensory limbic hyperconnection (SLH) theory [54] suggested that changes in diverse personality and behavior traits in TLE were associated with hyperconnectivity (increased synchronization to be more precise) between limbic and diverse cortical areas, a theory that garnered little empirical attention compared to the large number of papers that addressed the proposed behavioral changes associated with TLE. The present findings do support a pattern of altered connectivity linked to a core personality trait (neuroticism), but hypo- as opposed to hypersynchrony, between mesial temporal and cortical regions, with prominence of hyposynchrony between hippocampus and frontal lobe regions with increased neuroticism, as well as the unique variance of neuroticism independent of symptoms of depression and anxiety.

More generally, altered patterns of connectivity have been demonstrated in relation to symptoms of depression and anxiety in epilepsy. Garcia et al. [55] examined people with idiopathic generalized epilepsy, demonstrating decreased functional connectivity between the visuospatial/dorsal attention and the salience and default mode networks (DMN), and the DMN and the left executive control network in relation to symptoms of depression and anxiety in epilepsy. Similar research specific to TLE using the hippocampus as a seed region showed reduced connectivity between the hippocampus and anterior PFC as a strong contributor to depressive symptoms in left TLE [50]. Chen et al. [56] performed rs-fMRI in individuals with treatment-naive TLE with depressive symptoms, finding decreased connectivity between the limbic system, temporal lobe and frontal lobe. Furthermore, individuals with TLE with major depressive disorder also present alterations in resting state brain activity within the prefrontal-limbic system [57], particularly, the bilateral PFC, precuneus, angular gyrus and the right hippocampal gyrus and right temporal lobe. Our results suggest that, as in healthy individuals, people with TLE exhibit fronto-limbic functional disconnection in relation to depression and anxiety symptoms.

We now turn to regions found here to exhibit functional changes in relation to neuroticism, depression and anxiety scores in individuals with TLE.

4.4. Area 9 Hyposynchrony in Relations to Neuroticism and Depression Scores

Different regions of BA9 were found to be hyposynchronous to regions in the temporal lobe in relation to both neuroticism and depression scores, with left area 9a to right hippocampus showing the strongest effect (Table 2). This connection was also significant when the unique variance of neuroticism was examined, independent of depression and anxiety scores. Low cortical folding, as well as lesions to the dorsolateral prefrontal cortex (DLPFC) have been previously associated with higher neuroticism [58,59]. The Left BA9 has also been implicated in response to negative emotional distracters during high working memory load [60], self-criticism, which strongly correlates to depressive symptoms [61], and reaction to unpleasant stimuli [62]. BA9 also shows negative connectivity with the amygdala in relation to emotion expression/appraisal [63]. Taken together, these results suggest a dysregulation of brain regions that modulate emotion.

4.5. Hippocampal-Prefrontal Cortex (PFC) Hyposynchrony

Hippocampal-prefrontal cortex (PFC) connections have been known to play a role in behavioral and cognitive functions, as well as of major interest in understanding the neurological mechanisms of psychiatric disorders such as Post Traumatic Stress Disorder (PTSD) and Major Depression Disorder (MDD) [64]. The hippocampus and the PFC are both directly and indirectly connected. A monosynaptic projection directly connects the two, and multi-synaptic projections involve the thalamus, the nucleus accumbens, the ventral tegmental area and the basolateral amygdala [64]. Reduced fractional anisotropy, suggesting impaired structural connectivity, has been observed in tracts connecting the temporal and frontal lobes in association with high neuroticism [65], and generalized anxiety disorders [66]. Functional disconnection between the hippocampus and the PFC has been previously reported in MDD [67,68]. Functional connectivity in neuroticism and anxiety, however, has been more commonly associated with functional disconnection or impairment between the amygdala and the PFC [21,69]. This could represent a difference in populations between primary personality and psychiatric disorders secondary to TLE.

4.6. Area 47m and hippocampus hyposynchrony in all behavioral measures

The connection between right hippocampus and right area 47m (a part of BA47) was associated with all three behavioral measures (Table 2, Figure 3). Ongur et al. [70] subdivided Brodmann area 47 (or area 47/12 in Ongur et al.) into four regions based on cell body stains and Glasser et al. [45] confirmed these boundaries based on structural and functional MRI images. Right BA47 has been associated with affective prosody [71,72], perception of angry facial expressions [73] and behavioral inhibition [74,75] from many fMRI task activation studies. According to Glasser et al. [45], area 47m is one of the more distinctive areas of the cortex due to its heavy myelination.

Furthermore, they show that compared to its surrounding areas, 47m is more activated during an emotion processing task.

Functional hyposynchrony was observed in regarded to elevated neuroticism between a region in right dorsolateral prefrontal cortex (BA8Av) and regions in the somatosensory cortex (OP4) [76] and premotor area (Ventral area 6) [77].

Moreover, for depression symptoms, hyposynchrony was observed between the left fundus of the superior temporal area (temporal/occipital junction), a region involved in motion-sensitive vision [78], and the left dorsal temporal pole, a paralimbic region involved in social and emotional processing, auditory and visual aspects of facial recognition, emotional processing of auditory, olfactory and visual stimuli and theory of mind [79]. These results suggest widespread functional disconnections in the brain of people with TLE in relation to self-reported neuroticism, depression and anxiety, inferring the disruption of multiple functional networks.

4.7. Network Hypothesis

We hypothesize that it is the loss of connectivity (hypoconnectivity) between the limbic system and frontal cortex that gives epilepsy its propensity to increased neuroticism and the related symptoms of depression and anxiety [52,80]. Neuroticism is a stable core (Big 5) personality feature compared to symptoms and/or diagnoses of anxiety or depression which can be more episodic but also have a direct relationship with disruption of the frontal-limbic systems. Depression and anxiety symptoms may manifest from the combined effects of neuroticism, other neurobiological processes as well as the important associated psychosocial impact of seizures. Disruption of frontal lobe/limbic connection before temporal lobectomy predicts which patients are most likely to exhibit depressive symptoms after surgery [52] and many patients, even after temporal lobectomy, continue to experience depression and anxiety symptoms(even if seizure free) [52,81]. Potentially it may be the patients with persistent neuroticism who continue to experience depression and anxiety symptoms after epilepsy surgery while patients with improved mood had depressive symptoms at least in part related to the psychological stress of epilepsy. Thus, epilepsy surgery may relieve depression related to psychological stress of seizures but cannot improve the fronto-limbic hypoconnectivity which causes neuroticism. Indirectly supporting this hypothesis is the current study, which finds a greater number of hypoconnected regions for patients with elevated depression symptoms than for neuroticism. This may be because depressive symptomatology is not only related to elevated neuroticism but also the factors underlying seizure intractability. Thus, depression is mediated via both seizure intractability and neuroticism. To evaluate this hypothesis further studies are needed that measure neuroticism particularly before and after epilepsy surgery and as well as longitudinally are needed to ensure neuroticism is a stable metric in this patient population.

4.8. Limitations

One limitation of this study is that, per protocol, the measure of neuroticism (NEO-FFI) was not administered to every control that participated in the ECP; therefore, our sample size for neuroticism scores in controls is lower compared to the group with TLE. Second, to maximize the sample size we did not separate people with TLE based on the laterality of EEG abnormality, which is a common procedure in neuroimaging analysis of epilepsy, however, laterality was considered in the secondary analyses. Since our population fell mostly below the clinical threshold for self-reported anxiety and depression, longitudinal analysis would be helpful in the future to help understand whether connectivity changes lead to clinical psychiatric manifestations.

4.9. Conclusion

Epilepsy is a disease characterized by aberrant connectivity [35]. Within the primary epileptic circuit, conceptualized as the “epileptogenic zone”, hypersynchrony is required for seizure generation and propagation. However, corollary to this network specific hypersynchrony may be disruption of normal connectivity to regions functionally associated with the epileptogenic zone, which in turn causes hyposynchrony of signals and secondary neuro-psychiatric dysfunction. In this case, we find a loss of synchronization between the frontal and temporal regions in relation to increased reports of neuroticism, depression, and anxiety.

Supplementary Material

1

Supplementary Figure 1. Analyses were performed using clinically relevant cutoffs from previous literatures in order to verify the validity of the median split. When using those thresholds, the groups ended up as (1) Neuroticism: high (>60, N=41) vs. average or lower (<55, N=41), (2) Depression: borderline or above (≥65, N=30) vs. normal (≥55, N=42) and (3) Anxiety: borderline or above (≥65, ISM 6) vs. normal (≤55, N=56).

Figure 2.

Resting state connections that showed significant hyposynchrony (corrected p < 0.05) in individuals with temporal lobe epilepsy (TLE) with high scores compared to both those with low scores, and healthy controls, based on neuroticism (left) and depression (right) are plotted. The background brain images were generated with the Connectome Workbench and from Glasser et al. (2016)

Highlights.

• TLE patients showed significantly higher neuroticism compared to controls.

• Neuroticism T-scores were highly correlated with depression and anxiety, both in healthy controls and TLE patients.

• Decreased synchronization between mesial temporal and frontal cortical regions was associated with increased neuroticism in TLE patients.

• Hyposynchrony between the right hippocampus and right area 47m (a part of BA 47 in orbitofrontal cortex) was associated with increased neuroticism, depression and anxiety scores in TLE patients.

Abbreviations

AED

Anti-Epileptic Drug

ANOVA

Analysis of Variance

ASR

Adult Self Report

DSM-V

Diagnostic and Statistical Manual of Mental Disorders, 5^th edition

ECP

Epilepsy Connectome Project

EEG

Electroencephalogram

FOV

Field Of View

HCP

Human Connectome Project

MCW

Medical College of Wisconsin

NEO-FFI

Neuroticism-Extraversion-Openness Five-Factor-Inventory

PFC

Pre-Frontal Cortex

rs-fMRI

Resting State Functional Magnetic Resonance Imaging

SLH

Sensory Limbic Hyperconnection

TLE

Temporal Lobe Epilepsy