Skip to main content
NCBI home page
Social Cognitive and Affective Neuroscience logoLink to Social Cognitive and Affective Neuroscience
. 2025 Jul 30;20(1):nsaf077. doi: 10.1093/scan/nsaf077

Social support and depressive symptoms: the role of childhood maltreatment and brain activity in a moderated mediation model

Qianyi Luo 1,2, Qing Xu 3, Liwen Zhu 4, Jiyun Liao 5, Xiaoyu Liang 6, Jinrou Xia 7, Xiaohui Lin 8, Hongjun Peng 9,10,✉
PMCID: PMC12413245  PMID: 40737483

Abstract

Low social support, characterised by insufficient emotional, instrumental, or informational aid, increases the risk of major depressive disorder (MDD), yet its impact on brain function remains unclear. To examine this, we recruited 32 participants with MDD and low social support, 52 with high social support, and 54 healthy controls. We collected data including functional magnetic resonance imaging measures, sociodemographic information, social support dimensions, depressive symptom severity, and childhood maltreatment to investigate how social support relates to neural activity and depression severity, with childhood maltreatment as a potential moderating factor. Compared to the high social support group, participants with low social support showed decreased fractional amplitude of low-frequency fluctuation (fALFF) in the right superior temporal pole and increased fALFF in the left postcentral gyrus, along with reduced functional connectivity between the right superior temporal pole and right middle occipital gyrus. fALFF in the left postcentral gyrus correlated with physical abuse and inversely with social support. The moderated mediation model indicated that FC between the right superior temporal pole and right middle occipital gyrus mediated the relationship between perceived support and depression severity, with physical abuse moderating this effect. These findings underscore the relationship between social support, brain function, and depression severity.

Keywords: Major depressive disorder, Low Social support, Childhood maltreatment, fractional amplitude of low-frequency fluctuation, Functional connectivity

Introduction

Major depressive disorder (MDD) is a leading cause of global disability, affecting about 280 million people worldwide (WHO 2019). Depressive symptoms are generally categorised into emotional, cognitive, and behavioural components (Collaborators 2018). Emotional symptoms include anhedonia, low motivation, and guilt; cognitive symptoms involve negative thinking patterns, such as cognitive biases; and behavioural symptoms include fatigue, weight changes, appetite loss, and sleep disturbances (Collaborators 2018). A World Mental Health Survey across 18 countries found lifetime depression rates ranging from 11.1% to 14.6%, with 12-month rates between 5.5% and 5.9% (Bromet et al. 2011). In 2018, the World Health Organization ranked MDD third in disease burden, with an expected rise to first by 2030 (Malhi and Mann 2018). Despite the complex aetiology of MDD, research consistently identifies low social support as an important risk factor for its onset and severity (George et al. 1989, Ibarra-Rovillard and Kuiper 2011, Scardera et al. 2020). The lack of emotional support, practical assistance, and psychological reinforcement increases vulnerability to life’s challenges, exacerbating depressive symptoms (MaloneBeach and Zarit 1995, Ibarra-Rovillard and Kuiper 2011, Marroquín 2011).

Social support, encompassing emotional, instrumental, and informational aid, is vital for mental well-being (House et al. 1988, Slavin and Rainer 1990). It is often categorised as actual social support, referring to tangible help like financial or emotional assistance; perceived social support, reflecting an individual’s belief in their social network’s reliability; and utilised social support, measuring how effectively individuals access and employ available support resources when needed. Studies show that perceived support often plays a more considerable role in emotional well-being (Son et al. 2008, Ibarra-Rovillard and Kuiper 2011  Gariépy et al. 2016, Rueger et al. 2016). For example, Zhu and colleagues found that perceived social support mediates the relationship between attachment styles and depressive symptoms, buffering the effects of attachment anxiety and avoidance on depression in college students (Zhu et al. 2016). Similarly, Kang and colleagues found that perceived social support weakens the impact of negative parenting on depression (Kang et al. 2024). Furthermore, Son and colleagues found that perceived social support has a more significant protective effect on mental health than actual social support, suggesting that individuals’ perception of available support may be more effective in alleviating depression than the actual support received (Son et al. 2008). To provide a more comprehensive assessment, this study examines the dimensions of social support—perceived, objective, and utilised—and their relationship with depressive symptoms, focusing on the link between low social support and depression severity.

Existing research has consistently demonstrated that low social support is associated with more severe depressive symptoms and abnormal brain function in individuals with MDD. For example, Luo et al. (2024) found that MDD patients with low overall social support (assessed using a composite measure of objective, subjective and utilised support dimensions) exhibited decreased degree centrality in the superior temporal pole and medial geniculate nucleus, along with increased functional connectivity (FC) between the superior temporal pole and both the inferior temporal gyrus and the supramarginal gyrus, compared to those with high social support (Luo et al. 2024). These findings implicate regions central to social-emotional processing, with the superior temporal pole playing key roles in social cognition and emotional processing, while the supramarginal gyrus contributes to empathy and interpersonal perception—suggesting that low social support may disrupt neural networks essential for adaptive social functioning. Similarly, Jiang and colleagues reported a negative correlation between social support and cingulate cortex among individuals with depression, a region involved in emotional regulation and cognitive control (Jiang et al. 2023). Furthermore, Cheng and colleagues observed that cerebellar posterior lobe FC was negatively correlated with depressive symptoms and that social support mediated the relationship between cerebellar FC and depressive severity among women with postpartum depression (Cheng et al. 2022). While these findings suggest associations between low social support and altered neural networks supporting emotional regulation and social cognition in depression, the cross-sectional nature of these studies limits causal inferences, as neural function may also influence how individuals perceive and utilise social support in their environment.

However, studies suggest that the relationship between social support and depression varies (Gariépy et al. 2016), with factors like childhood maltreatment potentially influencing it. Childhood maltreatment refers to adverse experiences or events in early development, encompassing childhood neglect and abuse (Braithwaite et al. 2017, Nelson et al. 2017, Teicher et al. 2022,). Previous research has shown a close relationship between childhood maltreatment and social support (Struck et al. 2020, Su et al. 2022, Winter et al., 2023). Victims of childhood maltreatment who receive adequate social support experience a significantly reduced risk of mental health issues compared to maltreated individuals with poor social support networks (Su et al. 2022, Winter et al., 2023). Conversely, lacking sufficient social support may exacerbate the adverse effects of childhood maltreatment (Su et al. 2020). Moreover, childhood maltreatment is closely linked to MDD (Nelson et al. 2017). Previous studies suggest that childhood maltreatment may increase the risk of MDD by shaping individuals’ perceptions of social relationships and their ability to form and maintain supportive connections. Specifically, maltreatment may foster mistrust, fear of rejection, and social withdrawal, leading to decreased access to and utilisation of social support. Additionally, it can heighten individuals’ need for support due to increased emotional vulnerability while simultaneously impairing their capacity to seek and accept help (Struck et al. 2020, Grave et al. 2021, Su et al. 2022). Previous research has examined the relationships among social support, childhood maltreatment, and MDD. For instance, O’Shields et al. (2024) found that childhood maltreatment was linked to increased depressive symptoms and inflammation, while social support alleviated these symptoms. Struck and colleagues (2020) identified emotional abuse and neglect as significant factors in MDD severity, with attachment insecurity and social support serving as sequential mediators (Struck et al. 2020). Yu and Zhang (2025) highlighted the protective role of social support against MDD, despite the negative impact of childhood maltreatment. Su et al. (2022) demonstrated that exposure to all childhood maltreatment subtypes increased MDD risk, with social support mediating this relationship alongside mastery. Collectively, these studies support the view that social support plays an important role in mitigating the impact of childhood maltreatment on MDD (Su et al. 2022).

In this study, we explored the regional and inter-regional brain activity of participants with MDD and varying levels of social support (considering all three dimensions: actual, perceived, and utilised support) to investigate its relationship with brain function. Based on previous research, we hypothesised that: (i) Participants with high and low social support (derived from all three dimensions) would exhibit differences in brain activity. (ii) Abnormal brain activity mediates the relationship between low social support and depression severity, as perceived support has shown stronger associations with psychological outcomes in previous literature. (iii) Childhood maltreatment moderates the mediating role of abnormal brain activity between social support and depression severity. Given the temporal gap between childhood maltreatment and adult depression onset, we treated childhood maltreatment as a moderator rather than as a direct predictor. As a moderator, childhood maltreatment can change the strength or direction of the relationship between social support and depression or between brain function and depression, reflecting its role in creating vulnerability rather than directly causing depression symptoms years later. To test these hypotheses, we calculated indices such as the amplitude of low-frequency fluctuation (ALFF), fractional ALFF (fALFF), degree centrality (DC), and regional homogeneity (ReHo) to comprehensively assess regional brain activity. Regions showing group differences in these analyses were used as regions of interest for voxel-based FC analysis. By examining the associations between social support, brain function, and depression severity, this study aims to provide insights into the role of low social support in MDD.

Methods

Participants

The data for this study were derived from the baseline dataset of a registered clinical trial (ChiCTR2300078193). A total of 84 drug-naïve, first-episode participants with MDD (≥18 years old) and 54 healthy controls (HCs) were recruited. The diagnosis of MDD was made by two psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5). Any disagreements were resolved through consultation with a third psychiatrist. The clinician-administered Hamilton Depression Rating Scale (HAMD) was used to assess the severity of depressive symptoms among participants with MDD but not for diagnostic purposes. Drug-naïve, first-episode participants with MDD were recruited from the Affiliated Brain Hospital of Guangzhou Medical University, while age-, gender-, and education-matched HCs were recruited through community advertisements. HCs were required to have no psychiatric diagnoses, confirmed through psychiatric evaluations. “Drug-naïve” was defined as participants in both groups who had not taken any psychotropic medications, including antidepressants or antipsychotics. First-episode cases were included to ensure sample homogeneity.

To explore brain function differences based on social support, participants with MDD were divided into high and low social support groups (see Section Social Support Rating Scale for details). The study included 32 participants with MDD and low social support, 52 participants with MDD and high social support, and 54 HCs. Of note, HCs were not stratified by social support levels because our primary focus was to investigate the relationship between social support, brain function, and depressive symptoms in participants with MDD. HCs served as a reference group to establish baseline brain function. Moreover, stratifying HCs would have reduced subgroup sample sizes and statistical power, and the clustering of high social support levels among HCs further limited meaningful subgroup analysis. Exclusion criteria included (a) other mental illnesses (e.g. bipolar disorder, anxiety disorders, PTSD); (b) neurological disorders (e.g. stroke, brain tumours) or chronic illnesses (e.g. hypertension, diabetes, thyroid issues); (c) current or prior medication, psychotherapy, or physical therapy; and (d) contraindications for MRI scans. Participants with MDD meeting any exclusion criteria, regardless of treatment history, were excluded to minimise confounding effects. The study was approved by the Ethics Committee of the Affiliated Brain Hospital of Guangzhou Medical University, and all participants provided written informed consent.

Measures

Participant background information

We gathered sociodemographic data (age, sex at birth, education level) and clinical data (diagnosis, first onset) from the participants.

Hamilton Depression Rating Scale

The clinician-administered Hamilton Depression Rating Scale (HAMD) includes 24 items used to assess the severity of depressive symptoms and guide treatment (Hamilton 1960, Williams 1988, Bagby et al. 2004).

Social Support Rating Scale

Social support was assessed using the Social Support Rating Scale and semi-structured interviews. The Social Support Rating Scale provided a quantitative measure of overall social support, while the interviews evaluated pre-existing low social support through fixed and open-ended questions. Example questions included: “Was someone available to help you when needed?,” “How do you typically receive emotional support?,” and “Did low social support begin before or after your diagnosis of MDD or first depressive episode?” Participants self-reported the timing of their low social support, and only those with low support prior to their diagnosis or first episode were included. This criterion was implemented to establish temporal precedence, allowing us to examine low social support as a potential risk factor for MDD rather than a consequence of the disorder, thereby focusing on how pre-existing social support deficits might contribute to depression pathophysiology and neural mechanisms. Approximately 20% of potential participants were excluded based on this criterion.

The Social Support Rating Scale, designed by Xiao (1994), measures social support levels with 10 items across three dimensions: objective support, perceived support, and support utilisation. Objective support has 3 items (e.g. item 2: What is your living environment like); perceived support has 4 items (e.g. item 1: How many close friends can provide you with support). The total score ranges from 12 to 66, with higher scores indicating better support across all three dimensions. We used the total score for group classification and dimension-specific subscores for targeted analyses. In this study, the Social Support Rating Scale demonstrated good internal consistency, with a Cronbach’s alpha coefficient of 0.88. Social support was categorised into low and high groups based on Social Support Rating Scale total scores. Following previous research, we used a cut-off score of 22 points, with participants scoring 22 or below classified as having low social support and those scoring above 22 classified as having high social support (Xiao 1994).

Childhood Trauma Questionnaire

The Childhood Trauma Questionnaire evaluates childhood maltreatment (Bernstein et al. 1997, Bernstein et al. 2003). Participants rate statements for emotional, physical, and sexual abuse, as well as emotional and physical neglect, from 0 (never) to 4 (very often). Total scores range from 0 to 20 per subscale, with higher scores indicating more severe maltreatment. Scores ≥8 for sexual abuse, ≥13 for emotional abuse, ≥10 for physical abuse, ≥10 for physical neglect, and ≥15 for emotional neglect signify considerable maltreatment experiences. In our sample, the Childhood Trauma Questionnaire demonstrated good internal consistency, with Cronbach’s alpha coefficients of 0.84 for the total scale.

MRI data acquisition

MRI data were collected at the Radiology Department of the Affiliated Brain Hospital of Guangzhou Medical University using a Siemens Magnetom Prisma 3.0 T MRI scanner with a 64-channel head coil. Data were collected for 6 min and 10 s, during which participants were instructed to remain still, keep their eyes closed, and avoid falling asleep. Functional echo-planar images covered the entire brain with parameters including repetition time = 800 ms, echo time = 30 ms, flip angle = 56°, slice thickness = 2 mm, number of slices = 72, and field of view = 208 × 208 mm. Anatomical images were acquired using a sagittal 3D gradient-echo T1-weighted sequence with parameters repetition time = 2000 ms, echo time = 2.32 ms, inversion time = 900 ms, flip angle = 8°, number of slices = 208, slice thickness = 0.9 mm, and matrix = 256 × 256.

MRI data preprocessing

Functional images were preprocessed using the DPABI toolbox (version 6.1) in MATLAB (version R2019b) (Yan et al. 2016).

Preprocessing for ALFF and fALFF calculations

To stabilise longitudinal magnetisation, the first 25 functional volumes were discarded, and the remaining 425 volumes underwent slice timing correction. Subsequently, head motion correction was performed using Friston’s 24 motion parameters, and the average frame-wise displacement (FD) was calculated using the Jenkinson model (Jenkframe-wise displacementinson Jenkinson et al. 2002). FD is a measure of head motion calculated as the sum of the absolute values of relative displacements in translation and rotation between consecutive frames. All participants had an FD < 0.2 mm, with no significant group differences in FD (F = 0.131, P = .877). T1-weighted images were then co-registered to the mean functional image and segmented into grey matter, white matter, and cerebrospinal fluid. Following this, images were smoothed using an isotropic Gaussian kernel with a full width at half maximum (FWHM) of 4 mm. Finally, nuisance regression was conducted to remove signals from white matter, cerebrospinal fluid, and the Friston 24 motion parameters.

Of note, global signal regression was not performed to avoid centralising the FC distribution to zero, which can affect between-group differences and result in the loss of significant data regarding psychiatric conditions (Yang et al. 2014, Murphy and Fox 2017). This approach is supported by studies such as Cole et al. (2016) and Stoodley et al. (2017).

Preprocessing for DC and ReHo calculations

The steps included slice timing correction, realignment, normalisation, nuisance regression, and smoothing.

Preprocessing for FC calculations

The steps included slice timing correction, realignment, normalisation, smoothing, nuisance regression, and filtering.

Calculation of resting-state functional magnetic resonance imaging indices

Resting-state functional magnetic resonance imaging (fMRI) (rs-fMRI) is a neuroimaging technique that measures spontaneous brain activity during rest to study regional brain activity and FC. To characterise regional brain activity (neural activity within specific brain regions, often measured by neuroimaging techniques like fMRI to assess localised brain function), we extensively calculated ALFF, fALFF, DC, and ReHo. By integrating these metrics, we can comprehensively understand brain function changes from multiple aspects, including spontaneous neural activity, node importance, local region synchronisation, and overall network connectivity. This multi-faceted approach helps address our research questions by identifying both localised and network-level abnormalities, providing a comprehensive understanding of brain function in MDD. Regions showing significant differences in these measures among participants with MDD and low social support, participants with MDD and high social support, and HCs were used as regions of interest for FC analysis to explore interregional activity.

  1. ALFF measures the power of BOLD (blood-oxygen-level-dependent) signals within the low-frequency range (typically 0.01–0.08 Hz), reflecting the intensity of spontaneous neural activity in a brain region. Higher ALFF values indicate stronger spontaneous activity, potentially linked to the region’s function. ALFF is widely used to study various mental and neurological disorders, revealing disease-related brain function abnormalities by comparing ALFF differences between groups (Yu-Feng et al. 2007).

  2. fALFF is the normalised version of ALFF. It reduces the impact of physiological differences between individuals by calculating the ratio of low-frequency BOLD signal power (typically 0.01–0.08 Hz) to the power across the entire frequency spectrum (e.g. 0–0.25 Hz). fALFF reflects the relative intensity of spontaneous neural activity within the low-frequency range in each brain region. Compared to ALFF, fALFF more effectively reduces high-frequency noise, making cross-group comparisons more accurate and reliable (Zou et al. 2008).

  3. DC refers to the number of direct connections a brain region has with other regions, quantifying its connectivity. It reflects the importance of the region within the brain network. Regions with high DC are typically seen as hub nodes with strong information integration and dissemination capabilities (Zuo et al. 2012). Abnormal DC may indicate impaired connectivity or integration within the brain’s functional network.

  4. ReHo measures local brain region synchronisation by calculating the temporal consistency of a voxel with its neighbouring voxels. It reflects the strength and synchronisation of local FC. High ReHo values indicate highly synchronised neural activity within the region, suggesting tighter functional connections and cooperation. ReHo is commonly used to study the functional organisation of the brain in the resting state and to explore local synchronisation changes in various mental and neurological disorders (Zang et al. 2004).

  5. FC measures the functional cooperation or co-participation of different brain regions in specific brain functions by calculating the correlation of time series between them. High FC indicates strong coordinated activity between two brain regions, possibly forming a functional network. FC analysis is widely used to study the brain’s network structure of various mental and neurological disorders (Biswal et al. 1995).

Statistical analysis

Model assumptions were checked for all analyses. For ANOVAs, normality was tested using Shapiro-Wilk tests, and homogeneity of variances was assessed using Levene’s test. For the moderated mediation model, linearity, homoscedasticity, and multicollinearity were evaluated. No significant violations were detected; minor deviations were addressed using robust statistical methods. Gender, education level, and age were included as covariates to control for variability introduced by subgrouping MDD participants based on social support levels, as these factors can independently influence brain function and depression severity.

  1. We utilised one-way ANOVA, χ2 tests, and two-sample t-tests to examine intergroup differences in demographic data, clinical characteristics, and psychological assessments.

  2. One-way ANOVA was employed to investigate differences in brain function among participants with MDD and low social support, participants with MDD and high social support, and HCs. Social support was categorised into high or low groups using summed scores of Social Support Rating Scale. False discovery rate (FDR) correction was applied with a significance threshold of P < .05, followed by multiple comparisons using Bonferroni correction. Gender, education level, and age were considered covariates.

  3. Partial correlation analysis with FDR correction was conducted to explore the relationships between social support, childhood maltreatment, and abnormal brain activity, while controlling for gender, education level, and age. For these analyses, we examined both summed scores and subscales of social support and childhood maltreatment. Summed scores were used to assess overall associations, while subscales allowed for a detailed examination of specific dimensions and their unique contributions to brain activity.

  4. Using the PROCESS macro in SPSS (http://processmacro.org/index.html), we conducted moderated mediation analysis with bias-corrected 95% confidence intervals (CI) derived from 5000 bootstrap resamples. Our model hypothesised that abnormal brain function mediates the relationship between social support and MDD severity, with childhood maltreatment acting as a moderator. We tested each Childhood Trauma Questionnaire subscale (emotional abuse, physical abuse, sexual abuse, emotional neglect, and physical neglect) as potential moderators, as well as the summed childhood maltreatment score. Social support was treated as the independent variable (using summed scores for overall analysis and subscales for specific dimensions when necessary), MDD severity as the dependent variable, abnormal brain function as the mediator, and childhood maltreatment as the moderator. Childhood maltreatment was conceptualised as a moderator to capture its impact on the interaction between social support, brain function, and depressive severity, given the significant temporal gap between childhood maltreatment and the onset of MDD in adulthood. A moderated mediation effect was deemed significant if the 95% CI for the interaction term did not include 0. All analyses controlled for covariates such as age, gender, and education level, and significance was defined at P < .05.

Results

Participant demographics

Table 1 summarises the demographic information, HAMD scores, and levels of social support and childhood maltreatment for all participants. Significant age and educational level differences were observed between participants with MDD (both with low and high social support) and HCs. To control for the confounding effects of demographic factors such as age and educational level, these variables were treated as covariates in the subsequent analyses. In addition, significant variations were found in both total and subscale scores of childhood maltreatment across groups. Significant variations were found in childhood maltreatment scores across groups, with MDD participants with low social support reporting the highest levels (132.56 ± 31.34), followed by those with high social support (118.31 ± 34.08), and HCs reporting the lowest (72.93 ± 32.66). Emotional neglect (MDD-low support: 16.63 ± 5.13; MDD-high support: 14.12 ± 4.71; healthy controls: 9.44 ± 3.38) and physical neglect (MDD-low support: 10.25 ± 3.88; MDD-high support: 8.63 ± 3.37; healthy controls: 6.52 ± 1.79) differed significantly between all three groups (P < .05), with MDD patients with low social support reporting the highest levels, followed by those with high social support, and healthy controls reporting the lowest levels. For abuse subscales, both MDD groups reported significantly higher scores than controls without differing significantly from each other.

Table 1.

Demographic and clinical characteristics.

MDD with low social support (i) MDD with high social support (j) Healthy controls (k) F/t/χ2 P Post hoc tests
Age (years), mean ± SD 20.38 ± 4.006 21.52 ± 6.446 26.59 ± 4.402 18.96 < .001 (i < k)*, (j < k)*
Gender (male/female, %) 12/20 (37.5% male, 62.5% female) 14/38 (26.9% male, 73.1% female) 19/35 (35.2% male, 64.8% female) 1.276 .528 –
Educational level (years), mean ± SD 12.13 ± 3.643 13.1 ± 3.593 16.63 ± 3.625 19.84 < .001 (i < k)*, (j < k)*
Illness (weeks), mean ± SD 14.22 ± 13.246 17.65 ± 15.943 – −1.021 .310 –
HAMD score 21.47 ± 8.636 22.35 ± 7.101 – −0.506 .614 –
Social support 20.91 ± 1.838 30.6 ± 4.16 – −14.636 < .001 –
Objective support 4.62 ± 1.621 6.96 ± 2.048 – −5.479 < .001 –
Perceived support 12.16 ± 1.505 17 ± 3.419 – −8.910 < .001 –
Support utilization 5.09 ± 1.329 6.63 ± 1.597 – −4.569 < .001 –
Childhood maltreatment 132.56 ± 31.341 118.31 ± 34.084 72.93 ± 32.66 41.030 < .001 (i > k)*, (j > k)*
Emotional abuse 12.47 ± 5.809 11.54 ± 4.742 7.72 ± 2.736 15.349 < .001 (i > k)*, (j > k)*
Physical abuse 7.09 ± 3.809 6.73 ± 2.904 6.3 ± 2.098 3.041 .021 (i > k)*, (j > k)*
Sexual abuse 5.78 ± 1.93 6.21 ± 1.796 5.28 ± 0.787 2.857 .004 (i > k)*, (j > k)*
Emotional neglect 16.63 ± 5.129 14.12 ± 4.71 9.44 ± 3.38 30.826 < .001 (i > j)*, (j > k)*, (i > k)*
Physical neglect 10.25 ± 3.877 8.63 ± 3.372 6.52 ± 1.788 16.489 < .001 (i > j)*, (j > k)*, (i > k)*
Mean FD (mm) 0.071 ± 0.031 0.073 ± 0.026 0.070 ± 0.029 0.131 .877 -
Open in a new tab

FD: frame-wise displacement; HAMD: Hamilton Depression Rating Scale; SD: standard deviation; MDD: major depressive disorder.

*

P < .05.

Social support is associated with regional brain activity in participants with MDD in participants with MDD

We employed ALFF, fALFF, DC, and ReHo analyses to identify abnormal brain activity in participants with MDD and low social support. Given the calculation of five brain activity indices, a significance threshold of padj < 0.01 was applied following FDR correction. Among these indices, fALFF demonstrated significant abnormalities. Significant group differences in fALFF were found in the right superior temporal pole and left postcentral gyrus among the three groups (Table 2 and Fig. 1).

Table 2.

Brain regions showing significant fractional amplitude of low-frequency fluctuation differences between MDD with low social support, MDD with high social support, and healthy control groups.

Brain region MNI
 Cluster size F
x y z
Right superior temporal pole 40.5 17.5 −26.5 124 14.2909
Left postcentral gyrus −43.5 −26.5 45.5 83 14.5775
Open in a new tab

Figure 1.

Figure 1.

Open in a new tab

Effects of social support on intraregional brain activity in participants with MDD. Compared to participants with MDD and high social support, participants with MDD and low social support showed decreased fALFF in the right superior temporal pole and increased fALFF in the left postcentral gyrus. MDD: major depressive disorder; HCs: healthy controls; fALFF: fractional amplitude of low-frequency fluctuation.

Post hoc comparisons with FDR correction revealed that participants with MDD and low social support had reduced fALFF in the right superior temporal pole and increased fALFF in the left postcentral gyrus compared to those with high social support and HCs (padj < 0.025). In contrast, participants with MDD and high social support showed higher fALFF in the right superior temporal pole and lower fALFF in the left postcentral gyrus compared to HCs (Fig. 1 and Table 3).

Table 3.

Multiple comparisons of fractional amplitude of low-frequency fluctuation.

Brain region (I) (J) Mean difference (I-J) padj 95% confidence interval
Lower bound Upper bound
Right superior temporal pole MDD with low social support MDD with high social support −0.3323875* 0.003 −0.469326 −0.195449
MDD with low social support Health controls −0.1529168* 0.0336 −0.288885 −0.016948
MDD with high social support Health controls 0.1794707* 0.0045 0.061053 0.297888
Left postcentral gyrus MDD with low social support MDD with high social support 0.2330911* 0.003 0.147232 0.31895
MDD with low social support Health controls 0.1538330* 0.003 0.068582 0.239084
MDD with high social support Health controls −0.0792580* 0.037 −0.153505 −0.005011
Open in a new tab
*

padj < 0.05.

MDD: major depressive disorder.

Social support influences interregional brain connectivity in participants with MDD

Regions showing significant fALFF differences, specifically the right superior temporal pole and the left postcentral gyrus, were selected as regions of interest for the FC analysis. Significant group differences in FC were identified between the right superior temporal pole and the right middle occipital gyrus (Table 4). However, no significant group differences in FC were found when the left postcentral gyrus was considered as the regions of interest.

Table 4.

Brain regions showing significant functional connectivity differences between MDD with low social support, MDD with high social support, and healthy control groups.

Region of interest Brain region MNI
 Cluster size F
x y z
Right superior temporal pole Right middle occipital gyrus 48.5 −66.5 23.5 30 9.4688
Open in a new tab

Post hoc comparisons with FDR correction showed that participants with MDD, both with low and high social support, had reduced FC between the right superior temporal pole and the right middle occipital gyrus compared to HCs. However, there was no significant difference in FC between the two MDD groups (Fig. 2, Table 5).

Figure 2.

Figure 2.

Open in a new tab

Effects of social support on interregional brain activity in participants with MDD. Compared to healthy controls, participants with MDD and low social support showed decreased FC between the right superior temporal pole and the right middle occipital gyrus. Similarly, participants with MDD and high social support also showed decreased FC in this connection compared to healthy controls. MDD: major depressive disorder; HCs: healthy controls; FC: functional connectivity.

Table 5.

Multiple comparisons of functional connectivity.

Region of interest Brain region (I) (J) Mean difference (I-J) padj 95% confidence interval
Lower bound Upper bound
Right superior temporal pole Right middle occipital gyrus MDD with low social support MDD with high social support −0.0541143 0.1 −0.11877 0.010542
MDD with low social support health controls −0.1375609* 0.003 −0.201759 −0.073363
MDD with high social support health controls −0.0834466* 0.006 −0.139358 −0.027535
Open in a new tab
*

padj < 0.05.

MDD: major depressive disorder.

Partial correlation analysis between psychological characteristics and abnormal brain function

Partial correlation analyses, with FDR correction applied, were conducted to explore the relationship between abnormal brain function and social support, as well as childhood maltreatment. Among all participants, the fALFF in the left postcentral gyrus exhibited a positive correlation with the physical abuse subscale (r = 0.233, padj = 0.036), and negative correlations with the objective support subscale (r = −0.199, padj = 0.038) and support utilisation subscale (r = −0.225, padj = 0.032). Additionally, the FC between the right superior temporal pole and right middle occipital gyrus demonstrated positive correlations with total social support scores (r = 0.226, padj = 0.032), the objective support subscale (r = 0.179, padj = 0.0467), and the perceived support subscale (r = 0.196, padj = 0.042) (Fig. 3, Table 6).

Figure 3.

Figure 3.

Open in a new tab

Correlation between social support and brain dysfunction in participants with MDD. (a) The fALFF of the left postcentral gyrus correlated positively with physical abuse. The fALFF of the left postcentral gyrus correlated negatively with (b) objective support and (c) support utilisation. FC between the right superior temporal pole and right middle occipital gyrus correlated positively with (d) total social support, (e) objective support, and (f) subjective support. MDD: major depressive disorder; fALFF: fractional amplitude of low-frequency fluctuation; FC: functional connectivity.

Table 6.

Partial correlation analysis (n = 138).

Childhood maltreatment Emotional abuse Physical abuse Sexual abuse Emotional neglect Physical neglect Social support Objective support Perceived support Support utilization
fALFF of right superior temporal pole −0.091 −0.034 −0.148 0.078 0.048 0.011 0.063 0.086 −0.01 0.063
fALFF of left postcentral gyrus 0.182 0.126 0.233* 0.06 0.074 0.109 −0.170 −0.199* −0.048 −0.225*
FC between right superior temporal pole and right middle occipital gyrus 0.14 −0.155 −0.133 −0.118 −0.178 −0.199 0.226* 0.179* 0.196* 0.141
Open in a new tab
*

padj < 0.05.

fALFF: fractional amplitude of low-frequency fluctuation; FC: functional connectivity.

Abnormal brain activity mediates the relationship between social support and MDD severity, with childhood maltreatment moderating this effect

According to our hypothesis, childhood maltreatment (W) may moderate the indirect effects between social support (X) and MDD severity (Y) through abnormal brain function (M). Specifically, W can moderate Path a (X→M: social support→abnormal brain function) and Path b (M→Y: abnormal brain function→MDD severity). We tested moderated mediation models using each Childhood Trauma Questionnaire subscale as well as the total score. Only physical abuse showed a significant moderating effect. Significant moderated mediation effects were found in the model where social support (X) affects MDD severity (Y) through the FC between the right superior temporal pole and the right middle occipital gyrus (M), with physical abuse (W) as the moderator. As shown in Table 7 and Fig. 4, W moderates the indirect effects of X on Y through M (X→M: B = 0.0072, 95% CI 0.0016, 0.0129; M→Y: B = −14.9324, 95% CI −26.9276, −2.9371). This suggests that physical abuse moderates the indirect effects of social support on MDD severity through abnormal FC.

Table 7.

Moderated mediation analysis (n = 138).

Outcome: FC between right superior temporal pole and right middle occipital gyrus  
Variables B SE t LLCI ULCI
Perceived support 0.0072 0.0029 2.5223 0.0016 0.0129
Physical abuse 0.003 0.0053 0.5774 −0.0074 0.0134
Perceived support × physical abuse 0.0022 0.0011 2.0779 0.0001 0.0043
Outcome: HAMD  
Variables B SE t LLCI ULCI
Perceived support −1.0458 0.1951 −5.36 −1.4316 −0.6599
FC between right superior temporal pole and right middle occipital gyrus −14.9324 6.0653 −2.462 −26.9276 −2.9371
Open in a new tab

HAMD: Hamilton Depression Rating Scale; FC: functional Connectivity; SE: standard error.

Figure 4.

Figure 4.

Open in a new tab

Moderated mediation model. The FC between the right superior temporal pole and the right middle occipital gyrus mediated the relationship between perceived support and depression severity, with physical abuse moderating this mediation effect. FC: functional connectivity; CI: confidence interval; MOG: middle occipital gyrus.

Further analysis of the indirect effects of X on Y at different levels of W supported the significant moderated mediation model. As shown in Table 8, physical abuse was categorised into low (mean −1 SD), moderate (mean), and high (mean + 1 SD) levels. When physical abuse (W) was high in participants with MDD (B = −0.1777, 95% CI: −0.36, −0.0251), the FC between the right superior temporal pole and the right middle occipital gyrus (M) significantly mediated the relationship between social support (X) and MDD severity (Y). Models using other childhood maltreatment subscales and the total score did not show significant moderated mediation effects.

Table 8.

Conditional indirect effects of social support on depression severity at values of physical abuse.

Different levels of physical abuse B SE BootLLCI BootULCI
−1.8768 (mean—SD) −0.0459 0.0519 −0.1566 0.0552
−1.8768(mean) −0.0459 0.0519 −0.1566 0.0552
2.1232(mean + SD) −0.1777 0.0865 −0.36 −0.0251
Open in a new tab

SD: standard deviation; SE: standard error.

Discussion

In this study, we examined abnormal brain activity in participants with MDD and low social support. Our findings revealed that (1) compared to participants with MDD and high social support and HCs, those with low social support exhibited decreased fALFF in the right superior temporal pole, increased fALFF in the left postcentral gyrus, and reduced FC between the right superior temporal pole and the right middle occipital gyrus. (2) The fALFF in the left postcentral gyrus was positively correlated with physical abuse and negatively correlated with objective support and support utilisation. The FC between the right superior temporal pole and the right middle occipital gyrus was positively correlated with social support, objective support, and perceived support. (3) In the moderated mediation model, the FC between the right superior temporal pole and the right middle occipital gyrus mediated the relationship between perceived support and MDD severity, with physical abuse moderating this mediation. Even though demographic factors such as age, gender, and educational level were included as covariates, the potential impact of confounding factors cannot be entirely dismissed. It is crucial to consider these factors when interpreting the differences in brain function among the groups. Our findings highlight the complex interplay between low social support, abnormal brain function, depressive severity, and childhood maltreatment in participants with MDD. These results offer valuable insights for developing targeted interventions to address these interconnected factors.

We have the following understanding of the decreased local activity in the superior temporal pole among participants with MDD and low social support: the superior temporal pole is crucial for processing and integrating social information (Deen et al. 2015, Deen et al. 2020). Low social support correlates with abnormal activity in this region, impairing the ability of participants with MDD to effectively process social and emotional information (Pan et al. 2015, Yang et al. 2016), exacerbating feelings of isolation and helplessness (Zacková et al. 2021). Additionally, this region is integral to the theory of mind (Glasel et al. 2011, Beauchamp 2015), the ability to understand others’ emotions, beliefs, and intentions (Roumazeilles et al. 2021). Consequently, participants with low social support may struggle with theory of mind, impeding their ability to form social connections (Che et al. 2014). Overall, the reduced activity in this region likely stems from the complex impact of low social support on the brain function of participants with MDD, reflecting impairments in social cognition, emotional processing, and theory of mind.

Previous studies have highlighted the crucial role of the superior temporal pole in social support and functioning. For instance, O’Connell and colleagues discovered that stroke patients with impaired emotion recognition, linked to cortical regions such as the superior temporal pole, exhibited reduced engagement in social activities and reported lower levels of social support (O’Connell et al. 2022). In a large-scale study, Liu and colleagues demonstrated associations between altruistic behaviours and increased grey matter volume in the superior temporal pole (Liu et al. 2022). Similarly, Winter and colleagues identified specific white matter tract networks, particularly involving the superior temporal pole, that were linked to childhood maltreatment and levels of social support (Winter et al. 2023). Importantly, their findings revealed a negative correlation between childhood maltreatment and social support levels, emphasising the crucial role of social support in the functioning of the superior temporal pole and its implications for social interactions (Winter et al. 2023). Overall, these studies underscore the importance of social support in the superior temporal pole’s functioning. Our study contributes further empirical evidence to understand the neurobiological underpinnings of low social support in MDD.

The left postcentral gyrus, part of the somatosensory cortex, processes bodily sensory information (Iwamura 1998). We observed increased fALFF in this region among participants with MDD and low social support. Some studies suggest that increased activity in the somatosensory cortex may reflect greater focus on bodily sensations, helping individuals manage emotional distress (Min et al. 2022, Gao et al. 2023). For those with low social support, this could mean relying more on internal coping strategies, with the postcentral gyrus playing a role in regulating emotions and physical states. In contrast, other research indicates that MDD is associated with diminished attention to bodily sensations (Eggart et al. 2019). Thus, increased activity in the postcentral gyrus may not reflect enhanced bodily awareness but rather compensatory or altered neural processing, potentially reflecting disrupted engagement with emotional and physiological states. Given the limitations of rs-fMRI, our interpretations regarding the role of the postcentral gyrus in emotional regulation and interoception are speculative. Future task-based studies are needed to clarify whether heightened activity in this region represents compensatory mechanisms, maladaptive processes, or other aspects of neural function in MDD.

Consistent with our findings, current research supports the close relationship between abnormal activity in the postcentral gyrus, social support, and MDD. In patients with post-traumatic stress disorder, Chen and colleagues found increased FC between the dorsal anterior cingulate cortex and bilateral postcentral gyrus, with trauma symptom scores and social support positively correlated with this connectivity (Chen et al. 2019). Additionally, Chen and colleagues observed decreased activity in the bilateral postcentral gyrus in patients with MDD, who also experienced more severe emotional neglect and poorer social support (Chen et al. 2023). Emotional neglect indirectly affected MDD severity through abnormal postcentral gyrus activity and social support (Chen et al. 2023). Luo and colleagues investigated brain function in participants with MDD with childhood maltreatment using fALFF and FC analyses, discovering that MDD and childhood maltreatment affected activity in the postcentral gyrus and observed reduced FC between the left postcentral gyrus and several other brain regions (Luo et al. 2022). These studies highlight the role of postcentral gyrus dysfunction in somatosensory sensitivity and emotional regulation in patients with MDD, providing additional insights into the impact of social support on MDD.

We observed a reduction in FC between the right superior temporal pole and the right middle occipital gyrus in participants with MDD and low social support. The superior temporal pole is crucial for social cognition, emotion processing, and language comprehension (Pehrs et al. 2017), while the middle occipital gyrus primarily handles visual processing and spatial perception (Renier et al. 2010). Low social support may impair patients’ ability to understand and cope with social situations, affecting both their recognition of others’ emotions and social cues and their processing of visual information in complex environments (Cohen and Wills 1985, Gariépy et al. 2016, Hornstein et al. 2022). This reduced connectivity between the brain regions limits participants’ perception and understanding of their surroundings, exacerbating their social difficulties and emotional distress.

Previous studies have confirmed functional abnormalities in the superior temporal pole and middle occipital gyrus of patients with MDD, impacting emotion processing and facial emotion recognition. For instance, Guo and colleagues found reduced FC between the insula and prefrontal-limbic systems, emotional processing circuits, and visual areas, including connections with the left superior temporal pole and right middle occipital gyrus in drug-naive patients with MDD (Guo et al. 2015). Similarly, Hu and colleagues reported consistent findings in adolescents with MDD (Hu et al. 2019). Wang and colleagues observed reduced cerebral blood flow in the bilateral superior temporal pole and middle occipital gyrus of patients with MDD compared to HCs (Wang and Yang 2022). Additionally, Reker and colleagues demonstrated that higher levels of alexithymia were associated with lower activation levels in response to happy faces in these regions. However, the wide confidence intervals observed for the indirect effect of FC between the right superior temporal pole and the right middle occipital gyrus on depression severity (via perceived support) in the moderated mediation model may reflect inherent variability in brain function and its relationship with MDD severity. This variability could be influenced by individual differences in clinical and demographic factors, as well as the interaction effects introduced by the moderated mediation model. Additionally, the use of bootstrapping, while robust, tends to produce wider confidence intervals when the indirect effects are small or when the sample size within subgroups is limited. Overall, these studies support the presence of functional abnormalities in the superior temporal pole and middle occipital gyrus of patients with MDD, primarily affecting emotion processing and facial emotion recognition (Reker et al. 2010).

Limitations

This study has several limitations. Firstly, while our cross-sectional design precludes drawing causal inferences, we believe it is important to address our theoretical causal framework. Based on previous longitudinal research showing that social support often precedes and predicts depression outcomes (Gariépy et al. 2016), our primary hypothesis is that low social support contributes to altered brain function, which in turn influences depression severity. However, we acknowledge that bidirectional relationships are plausible—neural differences might also influence one’s ability to form and maintain social relationships. Future research should examine whether healthy controls with low social support show brain activity patterns similar to those in MDD patients, which would help determine if these neural changes precede depression and specifically relate to social support deficits rather than depression itself. If we could establish causality, we would seek to determine: (i) whether improving social support directly leads to normalisation of the observed brain activity patterns; (ii) whether these neural changes mediate improvements in depressive symptoms; and (iii) the extent to which childhood maltreatment moderates these causal pathways. Future longitudinal and intervention studies that manipulate social support levels could help disentangle these complex relationships and establish temporal precedence and causality. Secondly, there were significant age and education level differences between HCs and participants with MDD. Although we included age and education level as covariates, the potential impact of confounding factors cannot be fully dismissed. It is crucial to account for these confounders when interpreting brain function differences between the groups. Lastly, an important limitation is the potential memory bias in our assessment of social support and childhood maltreatment. Individuals with MDD often perceive past experiences more negatively (Everaert et al. 2022), potentially leading to underestimated social support or overestimated maltreatment in our retrospective self-reports.

Conclusion

In conclusion, this study examined the associations between social support and brain activity in participants with MDD. Our cross-sectional findings revealed differences in regional brain activity in the right superior temporal pole and left postcentral gyrus between MDD participants with high versus low social support. We observed significant correlations between brain activity patterns and both social support and childhood maltreatment measures. Furthermore, our analyses suggest that the FC between the right superior temporal pole and right middle occipital gyrus statistically mediates the relationship between perceived support and MDD severity, with physical abuse moderating this relationship.

Contributor Information

Qianyi Luo, Department of Clinical Psychology, The Affiliated Brain Hospital, Guangzhou Medical University, Guangzhou 510370, China; Key Laboratory of Neurogenetics and Channelopathies of Guangdong Province and the Ministry of Education of China, Guangzhou Medical University, Guangzhou 510370, China.

Qing Xu, Department of Clinical Psychology, The Third Hospital of Longyan, Longyan 364000, China.

Liwen Zhu, Department of Clinical Psychology, The Affiliated Brain Hospital, Guangzhou Medical University, Guangzhou 510370, China.

Jiyun Liao, Department of Clinical Psychology, The Affiliated Brain Hospital, Guangzhou Medical University, Guangzhou 510370, China.

Xiaoyu Liang, Department of Clinical Psychology, The Affiliated Brain Hospital, Guangzhou Medical University, Guangzhou 510370, China.

Jinrou Xia, Department of Clinical Psychology, The Affiliated Brain Hospital, Guangzhou Medical University, Guangzhou 510370, China.

Xiaohui Lin, Department of Clinical Psychology, The Affiliated Brain Hospital, Guangzhou Medical University, Guangzhou 510370, China.

Hongjun Peng, Department of Clinical Psychology, The Affiliated Brain Hospital, Guangzhou Medical University, Guangzhou 510370, China; Key Laboratory of Neurogenetics and Channelopathies of Guangdong Province and the Ministry of Education of China, Guangzhou Medical University, Guangzhou 510370, China.

Author contributions

Qianyi Luo (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Funding acquisition [equal], Methodology [equal], Software [lead], Visualisation [equal], Writing—original draft [lead]), Qing Xu (Investigation [equal]), Liwen Zhu (Data curation [equal], Investigation [equal]), Jiyun Liao (Formal analysis [equal], Methodology [equal]), Xiaoyu Liang (Formal analysis [equal], Project administration [equal], Resources [equal]), Jinrou Xia (Data curation [lead], Software [equal]), and Xiaohui Lin (Data curation [equal], Formal analysis [equal]), Hongjun Peng (Funding acquisition [lead], Supervision [lead], Validation [lead], Visualisation [equal]).

Conflict of interest

The authors declare that they have no conflict of interest.

Funding

This work was supported by the Guangdong Natural Science Foundation, China (2015A030313800 to H.P.), the Guangzhou Medical University Research Capacity Enhancement Program, China (2024SRP203 to Q.L.), the Guangzhou Health Science and Technology General Guidance Project (20241A011053 to Q.L.), and Guangzhou Research-oriented Hospital Foundation. The funding source had no role in the study design, analysis, or interpretation of data or in the preparation of the report or decision to publish.

Data availability

The data included in this study are available from the corresponding author upon reasonable request.

Author statement

The manuscript has been revised and approved by all authors. The authors guarantee that the manuscript is an original work, which has not been previously published and is not currently under consideration for publication elsewhere.

References

  1. Bagby RM, Ryder AG, Schuller DR  et al. The Hamilton depression rating scale: has the gold standard become a lead weight? Am J Psychiatry  2004;161:2163–77. [DOI] [PubMed] [Google Scholar]
  2. Beauchamp MS.  The social mysteries of the superior temporal sulcus. Trends Cogn Sci  2015;19:489–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bernstein DP, Ahluvalia T, Pogge D  et al. Validity of the Childhood Trauma Questionnaire in an adolescent psychiatric population. J Am Acad Child Adolesc Psychiatry  1997;36:340–8. [DOI] [PubMed] [Google Scholar]
  4. Bernstein DP, Stein JA, Newcomb MD  et al. Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Abuse Negl  2003;27:169–90. [DOI] [PubMed] [Google Scholar]
  5. Biswal B, Yetkin FZ, Haughton VM  et al. Functional connectivity in the motor cortex of resting human brain using echo‐planar MRI. Magn Reson Med  1995;34:537–41. [DOI] [PubMed] [Google Scholar]
  6. Braithwaite EC, O’Connor RM, Degli-Esposti M  et al. Modifiable predictors of depression following childhood maltreatment: a systematic review and meta-analysis. Transl Psychiatry  2017;7:e1162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bromet E, Andrade LH, Hwang I  et al. Cross-national epidemiology of DSM-IV major depressive episode. BMC Med  2011;9:90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Che X, Zhang Q, Zhao J  et al. Synchronous activation within the default mode network correlates with perceived social support. Neuropsychologia  2014;63:26–33. [DOI] [PubMed] [Google Scholar]
  9. Chen HJ, Zhang L, Ke J  et al. Altered resting-state dorsal anterior cingulate cortex functional connectivity in patients with post-traumatic stress disorder. Aust N Z J Psychiatry  2019;53:68–79. [DOI] [PubMed] [Google Scholar]
  10. Chen S, Yin Y, Zhang Y  et al. Abnormal spontaneous activity of regions related to mood regulation mediates the effect of childhood emotional neglect on major depressive disorder. Psychiatry Res Neuroimaging  2023;336:111729. [DOI] [PubMed] [Google Scholar]
  11. Cheng B, Roberts N, Zhou Y  et al. Social support mediates the influence of cerebellum functional connectivity strength on postpartum depression and postpartum depression with anxiety. Transl Psychiatry  2022;12:54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cohen S, Wills TA.  Stress, social support, and the buffering hypothesis. Psychol Bull  1985;98:310–57. [PubMed] [Google Scholar]
  13. Cole MW, Ito T, Bassett DS  et al. Activity flow over resting-state networks shapes cognitive task activations. Nat Neurosci  2016;19:1718–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Collaborators GDAI.  Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet  2018;392:1789–858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Deen B, Saxe R, Kanwisher N.  Processing communicative facial and vocal cues in the superior temporal sulcus. Neuroimage  2020;221:117191. [DOI] [PubMed] [Google Scholar]
  16. Deen B, Koldewyn K, Kanwisher N  et al. Functional organization of social perception and cognition in the superior temporal sulcus. Cereb Cortex  2015;25:4596–609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Eggart M, Lange A, Binser MJ  et al. Major depressive disorder is associated with impaired interoceptive accuracy: a systematic review. Brain Sci, 2019;9: [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Everaert J, Vrijsen JN, Martin-Willett R  et al. A meta-analytic review of the relationship between explicit memory bias and depression: depression features an explicit memory bias that persists beyond a depressive episode. Psychol Bull  2022;148:435–63. [Google Scholar]
  19. Gao Z, Lu K, Hao N.  Transcranial direct current stimulation (tDCS) targeting the postcentral gyrus reduces malevolent creative ideation. Soc Cogn Affect Neurosci  2023;18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gariépy G, Honkaniemi H, Quesnel-Vallée A.  Social support and protection from depression: systematic review of current findings in Western countries. Br J Psychiatry  2016;209:284–93. [DOI] [PubMed] [Google Scholar]
  21. George LK, Blazer DG, Hughes DC  et al. Social support and the outcome of major depression. Br J Psychiatry  1989;154:478–85. [DOI] [PubMed] [Google Scholar]
  22. Glasel H, Leroy F, Dubois J  et al. A robust cerebral asymmetry in the infant brain: the rightward superior temporal sulcus. Neuroimage  2011;58:716–23. [DOI] [PubMed] [Google Scholar]
  23. Grave U, Glanert S, Borchfeld K  et al. Differential effect of childhood emotional abuse on present social support in borderline disorder and depression: a cross-sectional study. Eur J Psychotraumatol  2021;12:1968612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Guo W, Liu F, Xiao C  et al. Decreased insular connectivity in drug-naive major depressive disorder at rest. J Affect Disord  2015;179:31–7. [DOI] [PubMed] [Google Scholar]
  25. Hamilton M.  A rating scale for depression. J Neurol Neurosurg Psychiatry  1960;23:56–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hornstein EA, Craske MG, Fanselow MS  et al. Reclassifying the unique inhibitory properties of social support figures: a roadmap for exploring prepared fear suppression. Biol Psychiatry  2022;91:778–85. [DOI] [PubMed] [Google Scholar]
  27. House JS, Umberson D, Landis KR.  Structures and processes of social support. Annu Rev Sociol  1988;14:293–318. [Google Scholar]
  28. Hu L, Xiao M, Ai M  et al. Disruption of resting-state functional connectivity of right posterior insula in adolescents and young adults with major depressive disorder. J Affect Disord  2019;257:23–30. [DOI] [PubMed] [Google Scholar]
  29. Ibarra-Rovillard MS, Kuiper NA.  Social support and social negativity findings in depression: perceived responsiveness to basic psychological needs. Clin Psychol Rev  2011;31:342–52. [DOI] [PubMed] [Google Scholar]
  30. Iwamura Y.  Hierarchical somatosensory processing. Curr Opin Neurobiol  1998;8:522–8. [DOI] [PubMed] [Google Scholar]
  31. Jenkinson M, Bannister P, Brady M  et al. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage  2002;17:825–41. [DOI] [PubMed] [Google Scholar]
  32. Jiang C, Jiang W, Chen G  et al. Childhood trauma and social support affect symptom profiles through cortical thickness abnormalities in major depressive disorder: a structural equation modeling analysis. Asian J Psychiatr  2023;88:103744. [DOI] [PubMed] [Google Scholar]
  33. Kang B, Li Y, Zhao X  et al. Negative parenting style and depression in adolescents: a moderated mediation of self-esteem and perceived social support. J Affect Disord  2024;345:149–56. [DOI] [PubMed] [Google Scholar]
  34. Liu Y, Zhang Y, Thyreau B  et al. Altruistic social activity, depressive symptoms, and brain regional gray matter volume: voxel-based morphometry analysis from 8,695 old adults. J Gerontol A Biol Sci Med Sci  2022;77:1789–97. [DOI] [PubMed] [Google Scholar]
  35. Luo Q, Chen J, Li Y  et al. Altered regional brain activity and functional connectivity patterns in major depressive disorder: a function of childhood trauma or diagnosis? J Psychiatr Res  2022;147:237–47. [DOI] [PubMed] [Google Scholar]
  36. Luo Q, Xu Q, Zhu L  et al. Major depressive disorder and perceived social support: moderated mediation model of security and brain dysfunction. J Psychiatr Res  2024;177:392–402. [DOI] [PubMed] [Google Scholar]
  37. Malhi GS, Mann JJ.  Depression. Lancet  2018;392:2299–312. [DOI] [PubMed] [Google Scholar]
  38. MaloneBeach EE, Zarit SH.  Dimensions of social support and social conflict as predictors of caregiver depression. Int Psychogeriatr  1995;7:25–38. [DOI] [PubMed] [Google Scholar]
  39. Marroquín B.  Interpersonal emotion regulation as a mechanism of social support in depression. Clin Psychol Rev  2011;31:1276–90. [DOI] [PubMed] [Google Scholar]
  40. Min J, Nashiro K, Yoo HJ  et al. Emotion downregulation targets interoceptive brain regions while emotion upregulation targets other affective brain regions. J Neurosci  2022;42:2973–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Murphy K, Fox MD.  Towards a consensus regarding global signal regression for resting state functional connectivity MRI. Neuroimage  2017;154:169–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Nelson J, Klumparendt A, Doebler P  et al. Childhood maltreatment and characteristics of adult depression: meta-analysis. Br J Psychiatry  2017;210:96–104. [DOI] [PubMed] [Google Scholar]
  43. O’Connell K, Marsh AA, Edwards DF  et al. Emotion recognition impairments and social well-being following right-hemisphere stroke. Neuropsychol Rehab  2022;32:1337–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. O’Shields J, Mowbray O, Cooper Z.  The effects of childhood maltreatment on social support, inflammation, and depressive symptoms in adulthood. Soc Sci Med  2024;340:116481. [DOI] [PubMed] [Google Scholar]
  45. Pan LA, Ramos L, Segreti A  et al. Right superior temporal gyrus volume in adolescents with a history of suicide attempt. Br J Psychiatry  2015;206:339–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Pehrs C, Zaki J, Schlochtermeier LH  et al. The temporal pole top-down modulates the ventral visual stream during social cognition. Cereb Cortex  2017;27:777–92. [DOI] [PubMed] [Google Scholar]
  47. Reker M, Ohrmann P, Rauch AV  et al. Individual differences in alexithymia and brain response to masked emotion faces. Cortex  2010;46:658–67. [DOI] [PubMed] [Google Scholar]
  48. Renier LA, Anurova I, De Volder AG  et al. Preserved functional specialization for spatial processing in the middle occipital gyrus of the early blind. Neuron  2010;68:138–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Roumazeilles L, Schurz M, Lojkiewiez M  et al. Social prediction modulates activity of macaque superior temporal cortex. Sci Adv  2021;7:eabh2392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Rueger SY, Malecki CK, Pyun Y  et al. A meta-analytic review of the association between perceived social support and depression in childhood and adolescence. Psychol Bull  2016;142:1017–67. [DOI] [PubMed] [Google Scholar]
  51. Scardera S, Perret LC, Ouellet-Morin I  et al. Association of social support during adolescence with depression, anxiety, and suicidal ideation in young adults. JAMA Netw Open  2020;3:e2027491. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Slavin LA, Rainer KL.  Gender differences in emotional support and depressive symptoms among adolescents: a prospective analysis. Am J Community Psychol  1990;18:407–21. [DOI] [PubMed] [Google Scholar]
  53. Son J, Lin N, George LK.  Cross-national comparison of social support structures between Taiwan and the United States. J Health Soc Behav  2008;49:104–18. [DOI] [PubMed] [Google Scholar]
  54. Stoodley CJ, D’Mello AM, Ellegood J  et al. Altered cerebellar connectivity in autism and cerebellar-mediated rescue of autism-related behaviors in mice. Nat Neurosci  2017;20:1744–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Struck N, Krug A, Feldmann M  et al. Attachment and social support mediate the association between childhood maltreatment and depressive symptoms. J Affect Disord  2020;273:310–7. [DOI] [PubMed] [Google Scholar]
  56. Su YY, Li M, D’Arcy C  et al. To what extent do social support and mastery mediate the association between childhood maltreatment and depression? A sequential causal mediation analysis. Epidemiol Psychiatr Sci  2022;31:e77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Su Y, D’Arcy C, Meng X.  Social support and positive coping skills as mediators buffering the impact of childhood maltreatment on psychological distress and positive mental health in adulthood: analysis of a national Population-Based sample. Am J Epidemiol  2020;189:394–402. [DOI] [PubMed] [Google Scholar]
  58. Teicher MH, Gordon JB, Nemeroff CB.  Recognizing the importance of childhood maltreatment as a critical factor in psychiatric diagnoses, treatment, research, prevention, and education. Mol Psychiatry  2022;27:1331–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Wang YM, Yang ZY.  Aberrant pattern of cerebral blood flow in patients with major depressive disorder: a meta-analysis of arterial spin labelling studies. Psychiatry Res Neuroimaging  2022;321:111458. [DOI] [PubMed] [Google Scholar]
  60. WHO 2019. Depression. https://www.who.int/health-topics/depression#tab=tab_1).(9 August 2025, date last accessed).
  61. Williams JB.  A structured interview guide for the Hamilton depression rating scale. Arch Gen Psychiatry  1988;45:742–7. [DOI] [PubMed] [Google Scholar]
  62. Winter A, Gruber M, Thiel K  et al. Shared and distinct structural brain networks related to childhood maltreatment and social support: connectome-based predictive modeling. Mol Psychiatry  2023;28:4613–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Xiao S.  Theoretical basis and research application of social support rating scale. J Clin Psychiatry  1994;4:98. [Google Scholar]
  64. Yan CG, Wang XD, Zuo XN  et al. DPABI: Data processing & analysis for (resting-state) brain imaging. Neuroinformatics  2016;14:339–51. [DOI] [PubMed] [Google Scholar]
  65. Yang GJ, Murray JD, Repovs G  et al. Altered global brain signal in schizophrenia. Proc Natl Acad Sci U S A  2014;111:7438–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Yang X-h, Huang J, Lan Y  et al. Diminished caudate and superior temporal gyrus responses to effort-based decision making in patients with first-episode major depressive disorder. Prog Neuropsychopharmacol Biol Psychiatry  2016;64:52–9. [DOI] [PubMed] [Google Scholar]
  67. Yu JJ, Zhang Z.  Long-term impact of adverse childhood experiences and perceived social support on depression trajectories. J Affect Disord  2025;369:255–64. [DOI] [PubMed] [Google Scholar]
  68. Yu-Feng Z, Yong H, Chao-Zhe Z  et al. Altered baseline brain activity in children with ADHD revealed by resting-state functional MRI. Brain Dev  2007;29:83–91. [DOI] [PubMed] [Google Scholar]
  69. Zacková L, Jáni M, Brázdil M  et al. Cognitive impairment and depression: meta-analysis of structural magnetic resonance imaging studies. Neuroimage Clin  2021;32:102830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Zang Y, Jiang T, Lu Y  et al. Regional homogeneity approach to fMRI data analysis. Neuroimage  2004;22:394–400. [DOI] [PubMed] [Google Scholar]
  71. Zhu W, Wang CD, Chong CC.  Adult attachment, perceived social support, cultural orientation, and depressive symptoms: a moderated mediation model. J Couns Psychol  2016;63:645–55. [DOI] [PubMed] [Google Scholar]
  72. Zou Q-H, Zhu C-Z, Yang Y  et al. An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: fractional ALFF. J Neurosci Methods  2008;172:137–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Zuo X-N, Ehmke R, Mennes M  et al. Network centrality in the human functional connectome. Cereb Cortex  2012;22:1862–75. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Data Availability Statement

The data included in this study are available from the corresponding author upon reasonable request.

Articles from Social Cognitive and Affective Neuroscience are provided here courtesy of Oxford University Press

RESOURCES