Impulsivity is a clinically important symptom of mental health conditions that can be defined in various ways [1]. Previous studies have shown that there is a difference between state impulsivity and trait impulsivity [2]. State impulsivity is relatively transient and sensitive to environmental conditions: it therefore varies over time in the same individuals. Contrarily, trait impulsivity represents a long-lasting personality characteristic regarded as a reasonably stable condition [3,4]. As a trait, impulsivity is defined as ‘a predisposition toward rapid unplanned reactions to internal or external stimuli without regard to the negative consequences of these reactions to the impulsive individual or other’. Herein, we focused on trait impulsivity, since it is more closely related to general mental health [5,6] and is more likely to be associated with brain structure due to its long-term stability.

Although impulsivity is more likely to be present in individuals with specific psychiatric disorders, it can also be observed in individuals with or without mental health conditions [7-9]. High impulsivity is related to substance use disorders [10], eating disorders [11], personality disorders, particularly antisocial [12], borderline [13], and bipolar disorders [14]; conduct disorders; and attention deficit hyperactivity disorders [13]. Additionally, high impulsivity is associated with many public health issues such as domestic violence [15], risky driving behaviors [16] and high risk sexual behaviors [17]. Thus, although high impulsivity itself is not a disease, it is relevant in several ways. Discovering the neural substrates responsible for impulsivity can help understand the severity of psychiatric symptoms, personality characteristics, and the degree of social adjustment.

Several studies have attempted to identify the brain regions responsible for impulsivity and identify the related neural substrates. Some studies have examined the neural correlates mediating increased trait impulsivity [18]. Specifically, structural magnetic resonance imaging (MRI) was performed with multiple modalities to determine the correlations between gray matter (GM) characteristics and trait impulsivity. For example, a study using voxel-based morphometry suggested that the orbitofrontal cortex is a specific region that correlates with trait impulsivity [19]. In addition, a study using surface-based morphometry (SBM), which illustrated cortical folding as a marker of early brain development, proposed that changes in the fronto-temporo-parietal regions correspond to higher impulsivity in healthy individuals [20]. Although only few studies have been conducted, changes in white matter (WM) connectivity based on diffusion tensor imaging have been correlated with certain brain regions [21-24]. However, these previous investigations were inconsistent due to their heterogeneity in the characteristics of the study participants, imaging modalities, and statistical analyses. Although a previous study primarily indicated that the prefrontal cortex (PFC) is a central region associated with impulsivity control [25], other studies suggest that temporal gyrus [26], parietal cortex [27], occipital lobe [28] and insula [29] are also related with trait im-pulsivity. Hence, our aim was to identify structural correlates of impulsiveness using cortical thickness (CT) and local gyrification index (LGI) in whole-brain vertex analyses in healthy individuals.

In the current study, we purposed to investigate whether the degree of trait impulsivity affects the CT and LGI in healthy individuals using surface-based approaches. Although GM volume measures have mainly been used to assess structural brain images, CT and LGI provide additional insights into the underlying volumetric differences. The SBM, showed that GM volume represents the product of CT and surface area, which are influenced by genetically different developmental processes [30]. In addition, the degree of gyrification can be used as a marker of abnormal development [31,32]. Overall, we expect that the current approach to the two surface domains (CT and LGI) will provide new insights into the brain regions associated with trait impulsivity in healthy individuals.

The aims of this study were (i) to identify specific brain regions that have significant correlations between trait impulsivity and CT or LGI values in healthy individuals using whole-brain vertex-wise correlation analysis; and (ii) to explore the association between clinical characteristics of participants such as anxiety, resilience, and quality of life and decreased CT or LGI in impulsivity-related GM regions.

Participants

Between April 2021 and October 2022, 75 healthy individuals (31 males and 44 females) were recruited from the local community through online and print advertising at the Department of Psychiatry, CHA Bundang Medical Center (Seongnam, Korea). All participants were Korean and ranged in age from 18 to 62 years (mean, 30.21 years). The exclusion criteria were as follows: 1) any history of Axis I or Axis II psychiatric disorders, including intellectual disabilities and substance use disorders; 2) history of neurological and neurodevelopmental disorders; 3) traumatic brain injury with loss of consciousness; 4) pregnancy; 5) left-handed individuals (based on the Edinburgh Handedness Inventory) [33]; 6) any contraindications for brain MRI; and 7) first-relative family history of psychiatric disorders.

The research procedures were reviewed and approved by the Institutional Review Board (no. 2019-05-030, 2021-03-001) of the CHA Bundang Medical Center, in accordance with the latest version of the Declaration of Helsinki and the principles of Good Clinical Practice. All study participants provided written informed consent after receiving a thorough explanation of the study procedures.

Assessments

The Barratt Impulsiveness Scale, 11th edition (BIS-11) [34] was used to estimate impulsivity. We assessed impulsive personality traits using the Korean version of the BIS-11-Revised (K-BIS-11-R), which revealed acceptable internal consistency (Cronbach’s α = 0.78) and test-retest reliability [35]. The BIS-11 is a self-reporting questionnaire that contains a total of 30 items rated on a 4-point Likert scale (1 = rarely/never to 4 = almost always/always). The BIS revealed three subscales as follows: (i) ‘attentional impulsiveness’ (8 items; the tendency to not focus on the task at hand), (ii) ‘motor impulsiveness’ (11 items; the tendency to act on the spur of the moment), and (iii) ‘non-planning impulsiveness’ (11 items; the tendency to not engage in careful thinking or planning).

The Beck Anxiety Inventory (BAI) [36] was analyzed to assess anxiety symptoms at baseline. The total score ranges from 0 to 63, with higher scores indicating more severe anxiety symptoms. We used the Korean version of BAI, which has shown good internal consistency (Cronbach’s α = 0.92) [37].

The World Health Organization Quality of Life (WHOQOL-BREF) was also used to assess quality of life. It is a 26-item survey consisting of four domains: physical health, psychological health, social relationships, and environmental health, along with overall quality of life and general health items. The Korean version of the WHOQOL-BREF [38] has high internal consistency (Cronbach’s α = 0.84) and validity [39]. Although there is no cut-off point, higher scores indicate a better quality of life.

The Korean version of the Connor-Davidson Resilience Scale (K-CD-RISC), a brief self-rating questionnaire for measuring resilience, was also administered to evaluate resilience among participants and shows great internal consistency (Cronbach’s α = 0.93) [40,41]. In the general population of 577 people, the mean CD-RISC was 80.4 (standard deviation [SD] = 12.8) and the median value was 82.3 (73, 90) [42].

All study participants underwent brain MRI at baseline using a 3.0-Tesla GE Signa HDxt scanner (GE Healthcare) with an 8-channel phased array head coil. Three dimensional T1-weighted fast spoiled gradient recalled echo sequence had the following parameters: repetition time = 6.3 ms, echo time = 2.1 ms, flip angle = 12°, slice thickness = 1 mm, field of view = 256 × 256 mm², matrix = 256 × 256, and isotropic voxel size = 1 × 1 × 1 mm³.

FreeSurfer (version 7.1.1; http://surfer.nmr.mgh.harvard.edu) general linear model regression was applied to reconstruct a cortical surface model [43,44] to calculate the mean CT and LGI for whole-brain vertex-wise correlation analysis. Age, sex and total intracranial volume (ICV) adjusted multiple regression models were performed. The results were corrected for multiple comparisons with a Monte Carlo simulation using a threshold of p < 0.05.

The CT was calculated as the nearest distance between the GM-WM boundary and the GM-cerebrospinal fluid boundary at each vertex on the surface [45]. The Desikan–Killiany [46] and Destrieux [47] cortical atlases, implemented in FreeSurfer, were used to label the cerebral cortex and calculate the mean CT of each region.

The LGI is defined as a metric that quantifies the amount of cortex buried in the sulcal folds compared to the amount of visible cortex in the regions of interest [48]. LGI provides information on cortical complexity to specify the regions of abnormal CT associated with specific pathological conditions [49]. A larger LGI value indicates a more folded area, whereas a smaller value indicates a smoother area. In this study, we used FreeSurfer’s built-in function to compute this value [49,50]. We registered each subject’s surface on that of FreeSurfer’s average subject, and calculated CT and LGI. The acquired values were then smoothed using a full-width at half-maximum filter with a radius of 10 mm.

The mean CT and LGI values of specific brain regions that showed significant correlation with poor impulse control scores were extracted for further analyses. Pearson’s correlation analyses were performed to examine the relationship between BIS scores and the CT or LGI in each region acquired from correlation analysis. Further correlation analyses were conducted to investigate the relationship between the extracted CT or LGI values in each region related to impulsivity and other clinical features. To correct for multiple comparison issues, a false discovery rate (FDR) correction was used (q < 0.05).

Statistical analyses were performed using Statistical Package for the Social Sciences version 27.0 (IBM Co.). For all analyses, a pvalue < 0.05 was considered statistically significant.

Demographic and Clinical Characteristics of Study Participants

Table 1 shows demographic and clinical characteristics of the study participants. The BAI total scores of the participants averaged 3.39 ± 3.76 (mean ± SD), and the total scores of CD-RISC and WHOQOL-BREF were 65.08 ± 14.81 and 99.19 ± 15.36, respectively.

Associations between Impulsivity and GM CT in Healthy Individuals

In healthy individuals, the BIS total scores showed a significant negative correlation with CT of the lateral occipital cortex (OC) (Monte Carlo simulations correction, cluster-wise p < 0.05) (Fig. 1). After analysis of covariance for age, sex, and ICV, this association remained signifi-cant. Additionally, the motor impulsivity subscale scores were negatively correlated with CT of the middle frontal gyrus and superior parietal cortex (Monte Carlo simulation correction, cluster-wise p < 0.05). The non-planning impulsivity subscale scores showed significant negative correlations with the middle temporal gyrus (Monte Carlo simulation correction, cluster-wise p < 0.05).

Associations between Impulsivity and LGI in Healthy Individuals

There was a significant negative association between BIS total scores and the LGI of the inferior frontal gyrus (IFG) (pars opercularis) (Monte Carlo simulation correction, cluster-wise p < 0.05) (Fig. 2). These findings remained unchanged after controlling for sex, age, and ICV as covariates. In addition, the motor impulsivity subscale scores were significantly and negatively correlated with IFG (pars opercularis and pars orbitalis) (Monte Carlo simulation correction, cluster-wise p < 0.05). No significant correlations were found between LGI, attention, and non-planning impulsivity subscale scores.

Correlation between LGI, CT Values, and Other Clinical Characteristics

The results of Pearson’s correlation analyses revealed that the mean CT values extracted from impulsivity-related lateral OC were significantly and negatively correlated with BAI total scores (r = −0.366, pFDR = 0.006) (Fig. 3). In addition, the mean CT values of lateral OC exhibited a significant positive correlation with total scores of WHOQOL-BREF (r = 0.256, p = 0.045) but did not survive the FDR correction. In addition, the LGI values of IFG were significantly and positively correlated with CD-RISC total scores (r = 0.359, pFDR = 0.006).

Our results provide evidence of a significant negative relationship among impulsivity, CT of the lateral OC and LGI of the IFG (pars opercularis) in healthy individuals. Furthermore, we demonstrated that mean CT and LGI values related to impulsivity were significantly associated with anxiety, resilience, and quality of life. The clinical relevance of these findings remains unclear. While earlier studies have focused on the relationship between executive dysfunction and altered prefrontal structure, the current results suggest the importance of exploring attentional processes as a mechanism underlying impulsivity.

Interestingly, we found a significant negative association between BIS total scores and CT of the lateral OC. The lateral OC is known to play an essential role in visual awareness of objects and is located in the occipital lobe which contains the primary and association visual cortex and is mostly responsible for visual processing [51,52]. Few studies have directly linked visual cortical functions of the OC with impulsivity. Previous functional MRI studies have shown greater attentional enhancement of visual responses in ventral and lateral OC, reflecting the role of endogenous attention control [53]. Additionally, Ide et al. [28] found a significant association between BIS scores and parietal occipital areas. Based on these points, poor attentional processes resulting from decreased GM thickness in the lateral OC could be an important factor in trait impulsivity in healthy individuals.

Furthermore, our subscale analysis indicated that motor impulsivity subscale scores were negatively correlated with the CT of the middle frontal gyrus and superior parietal cortex. Motor impulsivity is conceptualized as acting without thinking [54]. This finding is supported by previous functional MRI studies that reported a significant association between motor impulsivity, superior parietal lobule and dorsolateral prefrontal cortex, including the middle frontal gyrus, which is related to the role of general attention [55]. In addition, our finding of a significant negative association between non-planning impulsivity subscale scores and middle temporal gyrus is in line with the findings of a previous voxel-based meta-analysis. The temporal pole, including the middle temporal gyrus, which plays a crucial role in the control of negative emotional expression, can affect the perception of action goals related to non-planning impulsivity [56-58]. Therefore, our subscale analysis provides evidence that in healthy individuals, decreased CT in these areas is linked to subtypes of impulsivity such as motor and non-planning impulsivity.

A key finding of this study is that BIS total scores and motor impulsivity subscale scores showed significant negative correlations with LGI values in the IFG (pars opercularis). It has been reported that the process of gyrification is associated with early development of cortical connectivity [59]. Accordingly, these findings may be interpreted in terms of deviations in neurodevelopmental processes rather than state-dependent neurodegenerative processes [60]. Our findings corroborate those of previous neuroimaging studies, which suggest that the prefrontal cortex might be involved in trait impulsivity in healthy individuals and clinical samples [25,61]. Previous functional MRI studies have shown that the ventrolateral PFC serves as the core brain region for response inhibition [62,63]. Also, ventrolateral PFC was mainly considered and replicated as regions that relate to impulsivity in a previous neuroimaging review article with clinical samples [64]. However, the significant relationship between trait impulsivity in healthy adults and LGI in the IFG (pars opercularis), which is specific part of ventrolateral PFC, has not been well replicated as much. In a previous study with clinical samples, the IFG was found to be involved mainly in the cognitive and inhibitory control of motor responses related to modulating the pre-supplementary motor area-subthalamic nucleus excitatory circuit, leading to enhanced inhibition from the subthalamic nucleus to the motor cortex [26]. Furthermore, IFG is functionally related to risk aversion. Higher blood oxygen level-dependent activity in the IFG has been related to a reduced probability of making risky choices [65]. Thus, the present findings suggest that decreased LGI in the IFG, which is related to neurodevelopmental processes, may affect impulse control through a tendency toward risk aversion and inhibitory control over motor responses.

Our study also carried out correlation analysis between CT and LGI values in GM regions significantly associated with the total scores of BIS, BAI, WHOQOL, and CD-RISC. We observed a significant relationship between higher CT values in the lateral OC and lower anxiety, and a higher quality of life. Consistent with these results, previous studies have suggested that the level of functional activation in OC, leading to motivated attention, may be related to high anxiety levels [66,67]. It has also been reported that lower GM volumes in the OC are associated with lower psychological well-being, given that cognitive capacity plays a role in this region and affects psychological well-being [68]. Moreover, we demonstrated a significant positive relationship between the LGI values in the IFG (pars opercularis) related to impulsivity and resilience. This is consistent with a previous neuroimaging study that reported larger GM volume and functional activities in the IFG related to adaptive brain compensation to support emotion regulation as an aspect of psychological resilience in clinical samples [69]. Therefore, our results suggest that impairments in the attention and cognitive domains related to impulsivity likely affect level of anxiety and psychological characteristics such as quality of life and resilience.

Our study has a few limitations. First, since the sample size was relatively small, further studies with larger sample sizes are required to investigate the association between trait impulsivity, lateral OC and IFG (pars opercularis). Second, all assessments were based on self-report, and the results may have been affected by response bias. Third, because we studied only with healthy individuals, we did not include other psychological confounding factors except age and sex. Besides, there was a lack of well-known confounding factors because almost all previous studies have observed correlations between specific psychological disorders and OC’s CT or IFG’s LGI. If the confounding variables are further studied in future studies, it would be essential to investigate them with additional consideration. Finally, the sample in our study was comprised of healthy adults without symptoms of any personality disorder clinically. Accordingly, we need to consider that the associations identified in a healthy non-clinical sample, as studied here, might not necessarily be identical with those identified in patient samples. Despite this limitation of study, our findings provide neurobio-logical evidence of impulsivity which reflects a personality trait exhibited by healthy individuals, rather than excessive impulsivity in patients with mental disorders. Therefore, this study may be meaningful for healthy individuals with trait impulsivity.

In summary, we demonstrated decreased CT of the lateral OC and LGI of the IFG (pars opercularis) in healthy individuals with higher impulsivity, and its association with psychological characteristics. Our findings could provide a further basis for understanding the neurodevelopmental aspects of trait impulsivity which represents personality trait in healthy individuals. Understanding the neural correlates associated with trait impulsivity can potentially lead to multiple ways of mitigating impulsive behaviors with negative consequences in healthy individuals with high levels of impulsivity, anxiety, or poor positive resources for psychological characteristics such as resilience or quality of life.

This research was supported by the Basic Science Research Program through the National Research Found-ation of Korea, funded by the Ministry of Science and ICT (grant numbers NRF-2019M3C7A1032262), (grant number NRF-2021M3E5D9025026). This study was also funded in part by the Healthcare AI Convergence Research & Development Program through the National IT Industry Promotion Agency of Korea (NIPA), funded by the Ministry of Science and ICT (grant number S0102-23-1008). Both sets of funding were secured by S.H. Lee.

No potential conflict of interest relevant to this article was reported.

Conceptualization: Hye-Yeon Jung, Harin Bak, Sang-Hyuk Lee, Kang Soo Lee. Formal analysis: Hye-Yeon Jung, Harin Bak, Sang-Hyuk Lee, Kang Soo Lee. Funding: Sang-Hyuk Lee. Data curation: Hye-Yeon Jung, Harin Bak, Minji Bang, Sang-Hyuk Lee. Data curation—original draft: Sang-Hyuk Lee, Kang Soo Lee. Writing—original draft: Hye-Yeon Jung, Harin Bak. Supervision: Minji Bang, Sang-Hyuk Lee, Kang Soo Lee. Validation: Minji Bang, Sang-Hyuk Lee, Kang Soo Lee.