These authors contributed equally to this work.
Studies have been conducted to identify brain structural alterations related to high impulsivity in psychiatric populations. However, research on healthy subjects is relatively less extensive. Therefore, we aimed to investigate the correlation between the cortical thickness of whole brain regions and the impulsivity level in a healthy population.
We included 100 healthy participants aged 19–65 years. Their T1-weighted magnetic resonance images and the 23-item Barratt Impulsiveness Scale (BIS) score were obtained. The patients were divided into high and low impulsivity groups according to the 75th percentile score of the BIS in the sample. The thickness of each cortical region was calculated using the FreeSurfer, and the difference in cortical thickness of the whole brain between the high and low impulsivity groups was analyzed using one-way analysis of covariance including age, sex, education level, and total intracranial cavity volume as covariates.
The high impulsivity group showed significant cortical thinning in the left pars opercularis. The cortical thickness of the left pars opercularis significantly correlated negatively with the total, attention, and motor scores of the BIS scale.
Our findings suggest that prefrontal cortex thinning may play an important role in the development of high impulsivity in healthy adults.
Impulsivity is known as the tendency to react immediately to a stimulus and to behave in an unplanned manner without carefully reckoning the outcome of the action [
The neural mechanisms of impulsivity have not been fully explained; several studies have attempted to identify the brain regions responsible for impulsivity and identify the related neural circuitry. Brain lesion studies have identified that frontal lobe lesions, especially prefrontal cortex (PFC) lesions, lead to an increase in impulsivity [
Extensive structural neuroimaging studies have investigated the correlation between the PFC and impulsivity in terms of cortical thickness, gray matter volume, and white matter integrity. The cortical thickness of the OFC, superior frontal gyrus, and middle frontal gyrus correlates negatively with the impulsivity level in healthy subjects [
As mentioned above, many studies have attempted to clarify the neural correlates of impulsivity since it constitutes a significant axis of personality traits in human psychology. However, the characteristic neurobiology of impulsivity is not completely understood. Accumulating evidence suggests a core role of the PFC in impulsivity control [
Our a priori hypothesis is that non-clinical individuals in the high impulsivity group will have thinner cortical gray matter in the PFC compared to those in the low impulsivity group. Furthermore, we aimed to identify a more specific sub-region in the PFC that shows cortical thinning in the high impulsivity group.
The present study included 100 healthy participants aged 19–65 years recruited from the community using an advertisement between February 2010 and December 2017. The participants were assessed by two board-certified psychiatrists (Han KM and Ham BJ), who confirmed that none of the participants had any current or previous psychiatric diseases. The exclusion criteria were: 1) any psychiatric disorders (including personality and substance use disorders) on the DSM-IV-TR (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text-revised); 2) history of taking psychotropic medications within the last 6 months; 3) primary neurological illness (e.g., Parkinson’s disease, epilepsy, or cerebrovascular disease); 4) any contraindication for magnetic resonance imaging (MRI), including claustrophobia. The severity of depressive symptoms in all subjects was assessed using the 17-item Hamilton Depression Rating Scale (HDRS) [
The participants’ degree of impulsivity was measured using the 23-item Korean version of the BIS. The original version of the BIS, 11th edition [
In the present study, T1-weighted images were obtained parallel to the anterior commissure-posterior commissure line using a 3.0-T Siemens Trio whole-body imaging system (Siemens Healthineers, Munich, Germany), using T1-weighted magnetization-prepared rapid gradient-echo with the following parameters: repetition time, 1,900 ms; echo time, 2.6 ms; field of view, 220 mm; matrix size, 256×256; slice thickness, 1 mm; coronal slices without gap, 176; voxels, 0.86×0.86×1 mm3; flip angle, 16°; and number of excitations, 1.
From the T1 image data obtained from the participants, we calculated the cortical thickness using the three-dimensional model of the cortical surface reconstructions implemented in the FreeSurfer version 5.3 (Laboratory for Computational Neuroimaging, Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, USA;
The comparison of cortical thickness between the high and low impulsivity groups was performed using one-way analysis of covariance (ANCOVA), including cortical thickness as the dependent variable, group (high vs. low impulsivity group) as independent variables, and age, sex, education level, and TICV as covariates. The extracted cortical thickness values of 76 cortical gyri in the bilateral hemispheres from automated procedures of the FreeSurfer were used in the analysis. We applied the Bonferroni correction to the multiple comparisons in the main analysis to avoid type I error [i.e., p<0.05/76=0.000657 (76 cortical regions in the bilateral hemispheres)]. To investigate the correlation between the BIS score and cortical thickness, Pearson’s partial correlation analysis, including age, sex, education level, and TICV as covariates, was performed for cortical regions with significant differences between the two groups (p<0.05). All statistical analyses were performed using IBM SPSS Statistics for Windows, Version 24.0 (IBM Corporation, Armonk, NY, USA).
Information regarding age, sex, education level, total and sub-score of BIS, HDRS score, and TICV is shown in
The high impulsivity group showed significant cortical thinning in the pars opercularis (inferior frontal gyrus) in the left hemisphere compared to the low impulsivity group (F(1, 94)=13.380, p=4.19×10-4) (
As an exploratory analysis, we examined whether sex and age have moderating effects on the association between impulsivity level and cortical thickness in healthy adults. The total sample was sub-grouped based on sex and age (i.e., 19–39 years vs. ≥40 years), and the cortical thickness was compared between the high and low impulsivity groups. We found that cortical thinning of the left pars opercularis was only observed in female (F(1, 38)=11.420, p=0.002) and younger (F(1, 51)=10.510, p=0.002) participants, but not in male or older participants (both p>0.1). However, these findings were not significant after Bonferroni correction. Detailed information and findings of other cortical regions are presented in the Supplementary Materials (in the online-only Data Supplement) (
As an exploratory analysis, we performed Pearson’s partial correlation analyses on the left pars opercularis and aforementioned 16 cortical regions with significant differences in thickness between the two groups at a level of p<0.05 (
Our results show cortical thinning in the pars opercularis (inferior frontal gyrus) in the left hemisphere in the high impulsivity group, which was significant after Bonferroni correction. We also observed that the cortical thickness of the left pars opercularis had significant negative correlations with the total, attention, and motor scores of the BIS scale. Moreover, the BIS total score correlated negatively with the cortical thickness of the left parahippocampal gyrus, left inferior temporal gyrus, right subcentral gyrus, right anterior cingulate gyrus, and right orbital gyrus, which showed cortical thinning in the high impulsivity group, but was not significant after Bonferroni correction compared to the low impulsivity group. In the exploratory analysis, we also observed that sex and age have moderating effects on the association between impulsivity level and cortical thickness; however, they were not significant after Bonferroni correction.
Previous studies have reported that the cortical thickness of the PFC is inversely correlated with impulsivity levels in both healthy and clinical populations [
Further study may be required to explain the gray matter cortical thinning in the left vlPFC alone. This may be explained by the lateralization of hemispheric functions, since with dominance, the left and right hemispheres have different roles [
There are possible explanations for the decrease in cortical thickness of the pars opercularis in individuals with high impulsivity. Several biological mechanisms may account for this association. According to the Hebbian law, neurons “fire together wire together” [
Our research, as a cross-sectional study, successfully identified the cortical region involved in impulsivity in healthy adults; however, it still has several limitations. First, dividing the subjects into high and low impulsivity groups by the 75th percentile of the BIS-11 total score may seem arbitrary. Based on the recommendation of Stanford et al. [
In conclusion, we observed that the healthy adults with high impulsivity traits showed a thinner cortex in the left pars opercularis compared to those with low impulsivity traits. We also observed that cortical thickness of the left pars opercularis had a significant negative correlation with the total, attention, and motor scores of the BIS scale. As left pars opercularis is significant in motor control, human mirror neuron system and the language production, our research suggests that the highly impulsive individuals may have some functional alteration with left pars opercularis cortical thinning. Further studies should be conducted to confirm the suggested functional alteration in the highly impulsive individuals with left pars opercularis thinning.
The online-only Data Supplement is available with this article at
This research was supported by a Korea University Grant (K1925151).
The authors have no potential conflicts of interest to disclose.
Conceptualization: Kyu-Man Han. Data curation: Kyu-Man Han. Formal analysis: Kyu-Man Han. Funding acquisition: Kyu-Man Han. Investigation: Ji-Eun Lim, Seoyeon Kim, Surin Seo. Methodology: Kyu-Man Han. Project administration: Wooyoung Kang, Youbin Kang, Aram Kim. Resources: Byung-Joo Ham, Wooyoung Kang, Youbin Kang, Aram Kim. Software: Kyu-Man Han. Supervision: Byung-Joo Ham, Woo-Suk Tae. Validation: Kwan Woo Choi. Visualization: Wooyoung Kang, Youbin Kang. Writing—original draft: Ji-Eun Lim, Seoyeon Kim, Surin Seo, Kyu-Man Han. Writing—review & editing: Ji-Eun Lim, Seoyeon Kim, Surin Seo, Kyu-Man Han.
Cortical thinning of the left pars opercularis in the high impulsivity group. (A) shows the left pars opercularis (inferior frontal gyrus) in the Destrieux atlas (Red colored brain region). (B) shows a comparison of the cortical thickness of the left pars opercularis between the high and low impulsivity groups. The error bar represents one standard deviation.
Correlation between BIS score and cortical thickness of the left pars opercularis. (A-C) show scatter plots of the correlations of cortical thickness of the left pars opercularis with the total, attention, and motor scores of the BIS, respectively. BIS: Barratt Impulsiveness Scale.
Sociodemographic and clinical characteristics of participants
Characteristics | Total (N=100) | High (N=28) | Low (N=72) | p value (t, χ2) |
---|---|---|---|---|
Age | 38.84±13.33 | 39.54±15.12 | 38.57±12.67 | 0.766 (t=0.300) |
Sex | ||||
Female | 57 | 12 | 45 | 0.075 (χ2=3.174) |
Male | 43 | 16 | 27 | |
Education level | ||||
Elementary and middle school | 14 | 2 | 12 | 0.463 (χ2=1.542) |
High school or college/university | 72 | 22 | 50 | |
Above graduate school | 14 | 4 | 10 | |
HDRS score | 1.53±1.82 | 1.29±1.51 | 1.63±1.92 | 0.405 (t=-0.837) |
BIS (total) | 47.20±8.63 | 58.00±4.16 | 43.00±5.78 | <0.001 (t=12.511) |
BIS (attention) | 13.89±2.46 | 16.32±1.83 | 12.94±1.99 | <0.001 (t=7.805) |
BIS (motor) | 13.87±3.60 | 18.00±2.31 | 12.26±2.60 | <0.001 (t=10.223) |
BIS (non-planning) | 19.44±4.02 | 23.68±2.74 | 17.79±3.144 | <0.001 (t=8.703) |
TICV (cm3) | 1506.53±137.01 | 1516.89±164.22 | 1502.50±125.94 | 0.640 (t=0.470) |
Data are presented as mean±standard deviation for age, HDRS-17 scores, and BIS scores. The p values for the distribution of sex and education level were obtained using a chi-squared test. High group: BIS=53 or higher; Low group: BIS=52 or lower (75%ile score was 53). BIS: Barratt Impulsiveness Scale, HDRS: 17-item Hamilton Depression Rating Scale, TICV: total intracranial cavity volume
Comparison of cortical thickness between high and low impulsivity group
Cortical regions | High group (N=28) |
Low group (N=72) |
F(1, 94) | p-value | ||
---|---|---|---|---|---|---|
Mean | SD | Mean | SD | |||
Left hemisphere | ||||||
L Frontomarginal gyrus | 2.341 | 0.137 | 2.359 | 0.170 | 0.804 | 0.372 |
L Inferior occipital gyrus | 2.493 | 0.184 | 2.570 | 0.195 | 4.226 | 0.043 |
L Paracentral lobule | 2.366 | 0.210 | 2.424 | 0.170 | 0.704 | 0.404 |
L Subcentral gyrus | 2.693 | 0.198 | 2.729 | 0.170 | 0.807 | 0.371 |
L Transverse frontopolar gyrus | 2.589 | 0.208 | 2.672 | 0.208 | 2.101 | 0.151 |
L Anterior cingulate gyrus | 2.723 | 0.161 | 2.789 | 0.147 | 4.181 | 0.044 |
L Anterior mid-cingulate gyrus | 2.679 | 0.154 | 2.731 | 0.159 | 1.525 | 0.220 |
L Posterior mid-cingulate gyrus | 2.560 | 0.221 | 2.595 | 0.185 | 0.826 | 0.366 |
L Dorsal posterior cingulate gyrus | 2.654 | 0.485 | 2.797 | 0.425 | 2.544 | 0.114 |
L Ventral posterior cingulate gyrus | 2.490 | 0.287 | 2.570 | 0.286 | 3.825 | 0.053 |
L Cuneus | 1.861 | 0.318 | 1.842 | 0.357 | 0.295 | 0.588 |
L Pars opercularis | 2.711 | 0.217 | 2.841 | 0.154 | 13.380 | 4.19×10-4 |
L Pars orbitalis | 2.682 | 0.308 | 2.806 | 0.214 | 3.625 | 0.060 |
L Pars triangularis | 2.666 | 0.190 | 2.718 | 0.176 | 0.815 | 0.369 |
L Middle frontal gyrus | 2.748 | 0.233 | 2.763 | 0.169 | 0.068 | 0.794 |
L Superior frontal gyrus | 3.010 | 0.268 | 3.054 | 0.242 | 0.232 | 0.631 |
L Long insular gyrus | 2.907 | 0.292 | 3.007 | 0.258 | 2.184 | 0.143 |
L Short insular gyrus | 3.173 | 0.564 | 3.359 | 0.492 | 2.241 | 0.138 |
L Middle occipital gyrus | 2.542 | 0.185 | 2.603 | 0.158 | 2.489 | 0.118 |
L Superior occipital gyrus | 2.082 | 0.245 | 2.134 | 0.201 | 1.079 | 0.302 |
L Lateral occipito-temporal gyrus | 2.739 | 0.327 | 2.783 | 0.240 | 0.828 | 0.365 |
L Lingual gyrus | 1.999 | 0.258 | 1.946 | 0.223 | 0.932 | 0.337 |
L Parahippocampal gyrus | 2.866 | 0.275 | 2.977 | 0.224 | 4.977 | 0.028 |
L Orbital gyrus | 2.650 | 0.171 | 2.694 | 0.172 | 0.764 | 0.384 |
L Angular gyrus | 2.593 | 0.182 | 2.678 | 0.164 | 4.329 | 0.040 |
L Supramarginal gyrus | 2.606 | 0.314 | 2.749 | 0.250 | 5.515 | 0.021 |
L Superior parietal lobule | 2.429 | 0.202 | 2.441 | 0.147 | 0.014 | 0.907 |
L Postcentral gyrus | 2.180 | 0.167 | 2.245 | 0.177 | 1.627 | 0.205 |
L Precentral gyrus | 2.778 | 0.209 | 2.840 | 0.204 | 0.998 | 0.320 |
L Precuneus | 2.467 | 0.181 | 2.531 | 0.145 | 2.657 | 0.106 |
L Straight gyrus | 2.534 | 0.299 | 2.653 | 0.337 | 3.779 | 0.055 |
L Subcallosal gyrus | 2.469 | 0.295 | 2.500 | 0.414 | 0.001 | 0.978 |
L Anterior transverse temporal gyrus | 2.416 | 0.431 | 2.309 | 0.397 | 2.898 | 0.092 |
L Lateral superior temporal gyrus | 2.903 | 0.261 | 3.035 | 0.263 | 4.258 | 0.042 |
L Planum polare | 3.233 | 0.355 | 3.377 | 0.295 | 5.249 | 0.024 |
L Planum temporale | 2.569 | 0.200 | 2.653 | 0.207 | 2.597 | 0.110 |
L Inferior temporal gyrus | 2.748 | 0.291 | 2.917 | 0.266 | 9.316 | 0.003 |
L Middle temporal gyrus | 2.960 | 0.316 | 3.043 | 0.248 | 2.443 | 0.121 |
Right hemisphere | ||||||
R Frontomarginal gyrus | 2.339 | 0.220 | 2.396 | 0.173 | 1.731 | 0.191 |
R Inferior occipital gyrus | 2.629 | 0.214 | 2.714 | 0.211 | 5.123 | 0.026 |
R Paracentral lobule | 2.395 | 0.197 | 2.408 | 0.210 | 0.009 | 0.923 |
R Subcentral gyrus | 2.653 | 0.166 | 2.766 | 0.179 | 7.433 | 0.008 |
R Transverse frontopolar gyrus | 2.616 | 0.193 | 2.645 | 0.181 | 0.130 | 0.719 |
R Anterior cingulate gyrus | 2.698 | 0.157 | 2.775 | 0.139 | 6.199 | 0.015 |
R Anterior mid-cingulate gyrus | 2.778 | 0.161 | 2.816 | 0.124 | 1.118 | 0.293 |
R Posterior mid-cingulate gyrus | 2.632 | 0.135 | 2.638 | 0.198 | 0.029 | 0.864 |
R Dorsal posterior cingulate gyrus | 2.697 | 0.479 | 2.830 | 0.401 | 2.621 | 0.109 |
R Ventral posterior cingulate gyrus | 2.586 | 0.313 | 2.668 | 0.313 | 3.302 | 0.072 |
R Cuneus | 1.911 | 0.350 | 1.863 | 0.316 | 0.877 | 0.352 |
R Pars opercularis | 2.775 | 0.170 | 2.835 | 0.174 | 2.312 | 0.132 |
R Pars orbitalis | 2.721 | 0.209 | 2.827 | 0.218 | 3.647 | 0.059 |
R Pars triangularis | 2.684 | 0.158 | 2.752 | 0.162 | 3.379 | 0.069 |
R Middle frontal gyrus | 2.772 | 0.258 | 2.809 | 0.201 | 0.189 | 0.665 |
R Superior frontal gyrus | 3.031 | 0.260 | 3.050 | 0.220 | 0.000 | 0.993 |
R Long insular gyrus | 3.005 | 0.388 | 3.102 | 0.315 | 0.897 | 0.346 |
R Short insular gyrus | 3.053 | 0.552 | 3.215 | 0.438 | 2.118 | 0.149 |
R Middle occipital gyrus | 2.585 | 0.168 | 2.631 | 0.167 | 1.690 | 0.197 |
R Superior occipital gyrus | 2.112 | 0.210 | 2.174 | 0.176 | 1.632 | 0.205 |
R Lateral occipito-temporal gyrus | 2.753 | 0.301 | 2.794 | 0.197 | 1.194 | 0.277 |
R Lingual gyrus | 2.072 | 0.264 | 1.994 | 0.233 | 2.205 | 0.141 |
R Parahippocampal gyrus | 2.969 | 0.349 | 3.053 | 0.276 | 1.597 | 0.209 |
R Orbital gyrus | 2.637 | 0.194 | 2.745 | 0.180 | 4.753 | 0.032 |
R Angular gyrus | 2.587 | 0.196 | 2.667 | 0.147 | 3.867 | 0.052 |
R Supramarginal gyrus | 2.564 | 0.323 | 2.717 | 0.269 | 5.242 | 0.024 |
R Superior parietal lobule | 2.382 | 0.182 | 2.426 | 0.150 | 1.002 | 0.319 |
R Postcentral gyrus | 2.126 | 0.184 | 2.210 | 0.177 | 2.434 | 0.122 |
R Precentral gyrus | 2.768 | 0.239 | 2.825 | 0.171 | 1.601 | 0.209 |
R Precuneus | 2.516 | 0.199 | 2.549 | 0.185 | 0.368 | 0.545 |
R Straight gyrus | 2.520 | 0.321 | 2.659 | 0.267 | 4.755 | 0.032 |
R Subcallosal gyrus | 2.616 | 0.484 | 2.616 | 0.497 | 0.009 | 0.926 |
R Anterior transverse temporal gyrus | 2.407 | 0.416 | 2.414 | 0.382 | 0.091 | 0.763 |
R Lateral superior temporal gyrus | 2.887 | 0.316 | 3.040 | 0.280 | 5.056 | 0.027 |
R Planum polare | 3.118 | 0.359 | 3.159 | 0.289 | 0.570 | 0.452 |
R Planum temporale | 2.600 | 0.193 | 2.632 | 0.212 | 0.087 | 0.769 |
R Inferior temporal gyrus | 2.796 | 0.299 | 2.923 | 0.256 | 4.282 | 0.041 |
R Middle temporal gyrus | 2.896 | 0.289 | 3.012 | 0.257 | 3.738 | 0.056 |
One-way ANCOVA including age, sex, education level, and total intracranial volume as covariates, was performed. Significant group differences after Bonferroni correction were presented in a bold face. Bonferroni correction was applied for multiple comparisons: p<0.05/76=0.000658. L: left hemisphere, R: right hemisphere
Correlation between cortical thickness and impulsivity
Cortical regions | Total |
Attention |
Motor |
Non-planning |
||||
---|---|---|---|---|---|---|---|---|
r | p-value | r | p-value | r | p-value | r | p-value | |
L Inferior occipital gyrus | -0.181 | 0.078 | -0.141 | 0.169 | -0.228 | 0.026 | -0.096 | 0.350 |
L Anterior cingulate gyrus | -0.166 | 0.105 | -0.166 | 0.107 | -0.171 | 0.096 | -0.102 | 0.321 |
L Pars opercularis | -0.277 | 0.006 | -0.294 | 0.004 | -0.286 | 0.005 | -0.156 | 0.129 |
L Parahippocampal gyrus | -0.233 | 0.022 | -0.171 | 0.095 | -0.253 | 0.013 | -0.170 | 0.098 |
L Angular gyrus | -0.175 | 0.088 | -0.112 | 0.279 | -0.162 | 0.114 | -0.164 | 0.111 |
L Supramarginal gyrus | -0.177 | 0.085 | -0.208 | 0.042 | -0.143 | 0.164 | -0.124 | 0.230 |
L Lateral superior temporal gyrus | -0.125 | 0.223 | -0.106 | 0.304 | -0.153 | 0.135 | -0.066 | 0.523 |
L Planum polare | -0.185 | 0.071 | -0.211 | 0.039 | -0.150 | 0.145 | -0.133 | 0.196 |
L Inferior temporal gyrus | -0.239 | 0.019 | -0.210 | 0.040 | -0.263 | 0.010 | -0.149 | 0.147 |
R Inferior occipital gyrus | -0.186 | 0.070 | -0.204 | 0.046 | -0.154 | 0.135 | -0.136 | 0.188 |
R Subcentral gyrus | -0.294 | 0.004 | -0.235 | 0.021 | -0.268 | 0.008 | -0.250 | 0.014 |
R Anterior cingulate gyrus | -0.236 | 0.021 | -0.203 | 0.048 | -0.240 | 0.018 | -0.167 | 0.104 |
R Orbital gyrus | -0.208 | 0.042 | -0.094 | 0.360 | -0.255 | 0.012 | -0.160 | 0.120 |
R Supramarginal gyrus | -0.159 | 0.122 | -0.191 | 0.062 | -0.119 | 0.248 | -0.118 | 0.254 |
R Straight gyrus | -0.185 | 0.071 | -0.176 | 0.085 | -0.170 | 0.097 | -0.138 | 0.181 |
R Lateral superior temporal gyrus | -0.145 | 0.158 | -0.143 | 0.165 | -0.151 | 0.143 | -0.089 | 0.391 |
R Inferior temporal gyrus | -0.122 | 0.236 | -0.186 | 0.070 | -0.153 | 0.136 | -0.008 | 0.940 |
Pearson’s partial correlation adjusted for age, sex, education level, and total intracranial cavity volume was performed. L: left hemisphere, R: right hemisphere