Pre-coronavirus disease 2019 brain structure might be associated with social anxiety alterations during the pandemic : Chinese Medical Journal

Secondary Logo

Journal Logo


Pre-coronavirus disease 2019 brain structure might be associated with social anxiety alterations during the pandemic

Zhang, Xun1,2,3; Li, Qingyuan1,2,3,4; Yang, Xun5; Pan, Nanfang1,2,3; Suo, Xueling1,2,3; He, Min1,2,3; Wang, Song1,2,3; Kemp, Graham J.6; Gong, Qiyong1,7

Editor(s): Pan, Xiangxiang; Wei, Peifang

Author Information
Chinese Medical Journal ():10.1097/CM9.0000000000002363, April 26, 2023. | DOI: 10.1097/CM9.0000000000002363

To the Editor: As the novel coronavirus disease 2019 (COVID-19) pandemic rapidly spreads around the world, it has resulted in unprecedented economic loss, social turmoil, and detrimental effects on public mental health. In particular, the implementation of disease control measures such as nationwide lockdowns, social distancing, school closures, and home quarantine has given rise to a wide range of social anxiety (SA) around the world, from nonclinical manifestations to clinical social anxiety disorder (SAD).[1] Given the potentially severe impairments from SA and the limited mental health care resources, it is vital to identify whether there are vulnerability/risk neuropsychological factors contributing to the progression of SA during the pandemic. In addition, obsessive-compulsive symptoms/disorder (OCS/OCD), another important mental health issue during the pandemic, is conceptually and empirically related to SA and shares some neuroanatomical substrates with SA;[2] however, this potential link in the context of the COVID-19 pandemic has been less investigated. The current study therefore aims to explore the differences of neuroanatomical structures for discrepant changes of SA in normal young people, and the potential mediating roles of pandemic-specific OCS in the linking of brain structure with SA alterations during the COVID-19 pandemic.

This study was approved by the Medical Research Ethics Committee of West China Hospital at Sichuan University (No. 2021-1216). Written informed consent was obtained from all subjects at each experimental stage. The overall plan of the study is summarized in Supplementary Figure 1,

Before the outbreak of COVID-19 in Wuhan, from October 2019 to January 2020 (T1), 151 right-handed general university students with no history of neuropsychiatric disorders were recruited for structural magnetic resonance imaging (MRI) scanning and psychological measures (including the Liebowitz Social Anxiety Scale [LSAS], the Trait Anxiety Inventory [TAI], the Self-Rating Life Events Checklist [SRLEC], and the Socioeconomic Status Scale [SSS]). Subsequently, all participants were re-contacted for taking COVID-related behavioral measurements (Obsession with COVID-19 Scale) and were followed up for psychological evaluations (LSAS) during the COVID-19 outbreak in Wuhan (February 2020 to April 2020; T2). More details about brain image acquisition and behavioral measures are provided in Supplementary File, Finally, 115 subjects were identified as eligible for this study. As proven by polymerase chain reaction testing, no participants or their close relatives had been infected by COVID-19. To investigate whether pre-pandemic neuroanatomical differences are relevant to susceptibility to SA exacerbation during the COVID-19 pandemic, 63 participants who scored higher total score of the Liebowitz Social Anxiety Scale (LSAST) at T2 than T1 were designated as the higher-SA (HSA) group (36 females, aged 22.3 ± 2.0 years), and 52 participants who received lower scores at T2, compared with T1, were assigned to the lower social anxiety (LSA) group (30 females, aged 22.5 ± 2.1 years).

An optimized and standardized voxel-based morphometry analysis was conducted to measure the brain gray matter volume (GMV) using Statistical Parametric Mapping software (SPM12; Welcome Department of Cognitive Neurology, London, UK; (for details of image pre-processing and quality control, see Supplementary File, Then, whole-brain voxel-wise pre-pandemic GMV was compared between the HSA and LSA groups using independent-sample t-test with age, sex, and total GMV as covariates in SPM12. The false discovery rate (FDR) correction approach was used for multiple comparisons with a significance threshold of voxel-wise P < 0.001 and FDR-corrected P < 0.05 at the cluster level. In addition, to explore the correlations between GMV differences and SA dimensional alterations, the average GMV in the clusters with group differences was extracted, and a partial correlation analysis was performed between LSAS scores and pre-pandemic in the identified clusters with sex, age, and total GMV as covariates, using IBM SPSS Statistics 22.0 (IBM Corp, Released 2013. IBM SPSS Statistics for Windows, Armonk, New York, USA). Furthermore, to investigate the potential roles of pandemic-specific OCS levels in the relationship between the identified pre-pandemic GMV variations and SA alterations, a mediating effect analysis was conducted with the SPSS (IBM Corp. Released 2013) macro PROCESS by bootstrapping approach. In this analysis, the GMV of the brain region with group differences was the independent variable, OCS-19 scores were the mediator variable, categorized or dimensional alterations in SA were the dependent variable, and age, sex, and total GMV were covariates. For details of the statistical analyses, see Supplementary File,

There were no significant differences between HSA and LSA groups in terms of sex composition (male: 42.9% [27/63] vs. 42.3% [22/52]; χ2 = 0.004; P = 0.953), mean age (22.3 ± 2.0 years vs. 22.5 ± 2.1 years; t = −0.549; P = 0.584), LSAST (40.4 ± 18.6 vs. 42.4 ± 20.0; t= −0.561; P = 0.576), fear score of the Liebowitz Social Anxiety Scale (LSASF) (22.3 ± 10.3 vs. 22.3 ± 10.9; t = −0.017; P = 0.986), avoidance factor scores of the Liebowitz Social Anxiety Scale (LSASA) (18.1 ± 10.0 vs. 20.1 ± 10.3; t = −1.050; P = 0.296), TAI (42.4 ± 8.3 vs. 41.2 ± 7.9; t = 0.771; P = 0.442), frequency subscale of Self-Rating Life Events Checklist (12.3 ± 5.8 vs. 12.2 ± 5.8; t = 0.121; P = 0.904), impact subscale of Self-Rating Life Events Checklist (28.6 ± 16.5 vs. 27.5 ± 17.0; t = 0.335; P = 0.739), and SSS scores (5.0 ± 1.6 vs. 4.9 ± 1.3; t = 0.537; P = 0.592) at T1. Compared with the LSA group, the HSA group had significantly higher LSAST (57.8 ± 22.8 vs. 32.1 ± 17.5; t = 6.655; P < 0.001), LSASF (30.8 ± 12.7 vs. 18.8 ± 11.0; t = 5.330; P < 0.001), LSASA (27.0 ± 12.1 vs. 13.2 ± 8.0; t = 7.022; P < 0.001), and higher pandemic-specific OCS (6.0 [5.0, 8.0] vs. 5.0 [4.0, 6.0]; U = −3.081; P = 0.002) at T2 [Supplementary Table 1,].

The HSA group, compared to LSA, had larger GMV in the right supramarginal gyrus (SMG) (peak coordinate: X = 48, Y = −31.5, Z = 13.5; peak t = 5.239; cluster size = 560 voxels) [Figure 1A and Supplementary Table 2,]. At an FDR adjusted Pcorrected < 0.05, the pre-pandemic GMV of the right SMG was significantly and positively correlated with SA alterations (r = 0.237, P = 0.012; Figure 1B) and OCS (r = 0.256, P = 0.007) after controlling for age, sex, and total GMV. Furthermore, mediating effect analysis showed that there was a mediating role for pandemic-specific OCS between right SMG volume and categorical SA alterations (indirect effect = 4.056; 95% confidence interval [CI]: [0.257, 13.025]; P < 0.05; Figure 1C], and between right SMG volume and dimensional SA alterations (indirect effect = 26.734; 95% CI: [1.138, 62.766]; P < 0.05; Figure 1D). Crucially, all these results were robust and not affected by the potential interference effects of pre-pandemic TAI, SRLEC, and SSS scores (for details, see Supplementary File,

Figure 1:
Brain structures relating to discrepant alterations of SA during the pandemic, and the mediating roles of pandemic-specific OCS in the relationship between R_SMG volume and SA alterations. (A) Brain regions with significant differences in GMV between HSA and LSA participants after adjusting for age, sex, and total GMV. Red and orange colors (positive values) represent larger GMV, while blue and green colors (negative values) represent smaller GMV in HSA compared to LSA participants. (B) The correlations of GMV in R_SMG with SA alterations during the COVID-19 pandemic (r = 0.237, P = 0.012). The X-axis shows the standardized residuals of GMV in the R_SMG after adjusting for age, sex, and total GMV. The Y-axis shows the SA alterations (i.e., total LSAS score changes); n = 115. (C) Mediating roles of pandemic-specific OCS between the R_SMG volume and categorical SA alterations. (D) Mediating roles of pandemic-specific OCS between the R_SMG volume and dimensional SA alterations. Unstandardized regression coefficients are shown next to the arrows ( P < 0.05, P < 0.01, P < 0.001). Age, sex, and total GMV were adjusted in the model. CI: Confidence interval; COVID: Coronavirus disease; GMV: Gray matter volume; HSA: Higher social anxiety group; LSA: Lower social anxiety group; LSAS: Liebowitz Social Anxiety Scale; OCS: Obsessive-compulsive symptoms; R_SMG: Right supramarginal gyrus; SA: Social anxiety.

As a key area in the neurobiology of SA, the SMG is important in extended neurobiological model of SAD.[3] Generally, SMG is considered an important component of the mirror neuron system which is involved in imitative and empathic social behavior and as the neural basis of overcoming emotional egocentricity bias in social judgments.[4] Self-focused attention bias and negative emotional experiences, which impair empathy, are core components of SAD, and empathy dysfunction may account for many of characteristic social impairments.[3] In this sense, we tentatively suggested that larger GMV of the SMG may be a neurostructural correlate of empathy dysfunction in those who become excessively socially anxious during the pandemic.

Furthermore, we found that pandemic-specific OCS levels mediate the effects of pre-pandemic SMG volume on SA alterations during the pandemic. The prevalence of OCS and SA were high in both clinical and general populations during the pandemic,[1,2] and OCS can predict pandemic-related anxiety. The reasons might be that individuals with severe pandemic-specific OCS feared to violate quarantine polices, lead to viral transmission and incur public criticism, so they avoided to take part in social interactions (i.e., deteriorating SA).[2] From the point of neurology, the mediating effects of pandemic-specific OCS seem consistent with known links between SMG and OCS. Structural MRI showed larger GMV and gyrification of the right SMG in OCD patients, while diffusion tensor imaging in OCD found a structural abnormalities in the SMG and a correlation between axial/radial diffusivity in the SMG and the extent of obsession.[5] The SMG is also associated with self-regulatory functioning, such as goal setting, action planning, set shifting, and response inhibition; the repetitive and uncontrollable thoughts or behaviors of OCS are considered to reflect a failure of such self-regulation.[4] Thus, it can be assumed that larger GMV of the SMG may be related to the obvious abnormal behavior in OCS.

In conclusion, this study extends our neuroanatomical understanding by showing that a larger pre-pandemic SMG volume could prospectively predict the SA aggravation during the COVID-19 pandemic. Furthermore, it showed a neuropsychological pathway of the changes of SA during the COVID-19 pandemic, and evidenced that COVID-19-related obsession can mediate the effects of pre-pandemic SMG volume on SA alterations. Clinically, this may point to potential neuropsychological risk factors for the development of SA, which may help to identify candidate neuroanatomical biomarkers and adopt timely interventions, such as neurofeedback procedures and cognitive behavioral therapy, for individuals who tend to suffer SA aggravation during major epidemics and similar societal stress.


This study was supported by the National Natural Science Foundation of China (Nos. 81621003, 81761128023, 81820108018, 82027808, and 31800963).

Conflicts of interest



1. Loades ME, Chatburn E, Higson-Sweeney N, Reynolds S, Shafran R, Brigden A, et al. Rapid systematic review: the impact of social isolation and loneliness on the mental health of children and adolescents in the context of COVID-19. J Am Acad Child Adolesc Psychiatry 2020;59:1218–1239.e3. doi: 10.1016/j.jaac.2020.05.009.
2. Shafran R, Coughtrey A, Whittal M. Recognising and addressing the impact of COVID-19 on obsessive-compulsive disorder. Lancet Psychiatry 2020;7:570–572. doi: 10.1016/S2215-0366(20)30222-4.
3. Bruhl AB, Delsignore A, Komossa K, Weidt S. Neuroimaging in social anxiety disorder-a meta-analytic review resulting in a new neurofunctional model. Neurosci Biobehav Rev 2014;47:260–280. doi: 10.1016/j.neubiorev.2014.08.003.
4. Silani G, Lamm C, Ruff CC, Singer T. Right supramarginal gyrus is crucial to overcome emotional egocentricity bias in social judgments. J Neurosci 2013;33:15466–15476. doi: 10.1523/JNEUROSCI.1488-13.2013.
5. Fan Q, Palaniyappan L, Tan L, Wang J, Wang X, Li C, et al. Surface anatomical profile of the cerebral cortex in obsessive-compulsive disorder: a study of cortical thickness, folding and surface area. Psychol Med 2013;43:1081–1091. doi: 10.1017/S0033291712001845.

Supplemental Digital Content

Copyright © 2023 The Chinese Medical Association, produced by Wolters Kluwer, Inc. under the CC-BY-NC-ND license.