The influence of childhood trauma on social media-induced secondary traumatic stress among college students: the chain mediating effect of self-compassion and resilience

ABSTRACT

Background: Studies have shown that media exposure to critical public events can lead to secondary traumatic stress (STS). Personal trauma history, self-compassion and resilience are important factors influencing STS in healthy professionals. However, whether these variables are associated with social media-induced STS in college students and the underlying mechanisms remain unclear. The purpose of this study was to explore the complex relationship linking childhood trauma to social media-induced STS in a large sample of college students.

Methods: A total of 1151 Chinese college students from Chengdu, Sichuan Province of China completed a web-based cross-sectional survey, which included standard assessments of childhood trauma, self-compassion, resilience and social media-induced STS, as well as sociodemographic questionnaires. The chain mediation model was tested using the PROCESS macro programme in SPSS software.

Results: There was a moderate correlation between childhood trauma and social media-induced STS (r = 0.34, p < .001). This association was significantly mediated by self-compassion (indirect effect [95% CI] = 0.14[0.11, 0.17]) and resilience (indirect effect = 0.03[0.01, 0.04]), respectively. Further, a chained mediating effect was observed with self-compassion and resilience consecutively mediated the relationship between childhood trauma and social media-induced STS (indirect effect = 0.02[0.01, 0.03]). These results persisted after sociodemographic characteristics were included as controlling variables.

Conclusions: Early life trauma impacts STS induced by exposure to traumatic materials on social media through self-compassion and resilience among Chinese college students. Psychological interventions targeting self-compassion and resilience can be implemented to reduce the risk of STS, especially in vulnerable individuals.

HIGHLIGHTS

• Exposure to traumatic information through social media can lead to secondary traumatic stress.

• Childhood trauma is related to social media-induced secondary traumatic stress.

• Self-compassion and resilience sequentially mediate this relationship between childhood trauma and social media-induced secondary traumatic stress.

1. Introduction

According to the latest statistics, as of April 2024, there were 5.07 billion (62.6% of the global population) social media users worldwide (Statista, 2024). Social media applications such as Facebook, X/Twitter, Instagram, WhatsApp, TikTok and Wechat are flooded with posts, photo and video uploads that are shared or viewed millions of times every minute (McLean, 2022). While social media is often used for entertainment and communication, it can also be a source of violent and distressing information, such as massing shooting, war footage, natural disasters, homicide, suicide, sexual assault, etc. This has raised concerns that such media exposure may cause secondary trauma, as social media users may indirectly experience traumatic events when viewing distressing content online (Dubberley et al., 2015; Pearson, 2024; Scott et al., 2023).

Studies have shown that media exposure to critical public events (e.g. collective trauma, the COVID-19 pandemic) can lead to negative mental health problems, including posttraumatic stress disorder (PTSD) symptoms (Neria & Sullivan, 2011), acute stress (Holman et al., 2014, 2020) and anxiety (Liu & Liu, 2020). These are the main characteristics of secondary traumatic stress (STS, Figley, 2013). STS refers to the emotional distress that occurs when individuals are exposed to other people's first-hand traumatic experiences rather than being directly exposed to the trauma (Bhagwagar, 2022; Bride et al., 2004). Initially, STS was considered specific to those who help traumatized or suffering people, such as trauma workers and family members (Figley, 1995a), and its symptoms are closely similar to PTSD (Figley, 1995b). The Secondary Traumatic Stress Scale (STSS, Bride et al., 2004) is a standard and well-validated tool for assessing STS. The scale is used to measure intrusion, avoidance and arousal symptoms triggered by vicarious exposure to traumatic events during work with traumatized clients (He et al., 2022; Kitano et al., 2023; Măirean, 2016). Previous studies in helping professionals (e.g. nurses, healthcare workers, clinical social workers) have shown that STSS scores are associated with mental health conditions such as depression, suicidal ideation, burnout and anxiety (Ariapooran et al., 2022; Bock et al., 2020; Orrù et al., 2021) as well as perceived physical health (Lee et al., 2018).

Recently, STSS has been increasingly used across diverse populations (Comstock & Platania, 2017; Mancini, 2019; Zhong et al., 2021), especially as the media continues to expose various traumatic events to the public (Dubberley et al., 2015; Liu & Liu, 2020; Pearson, 2024). For example, Comstock and Platania (2017) modified the STSS by rewording the items to measure STS in the context of exposure to trauma through television or social media, and found that media-induced STS predicted laypersons’ perceptions of distress. Mancini (2019) investigated STS in social media users using revised STSS which added stems linking the construct to social media use and found that STS was related to depressive symptoms and prior traumatic experience in social media users. These results suggest that STS exists in the general population and provide preliminary insights into its contributing factors and negative effects. However, to our knowledge, no study has investigated social media-induced STS (STS-SM) and its associated factors among college students. College students constitute the majority of social media users (Kemp, 2018), which may be related to their various characteristics such as strong curiosity and thirst for knowledge and information (Han & Xu, 2024), need for social support (Zhang et al., 2023a) and fear of missing out (Milyavskaya et al., 2018). While the diverse content and rich information on social media can satisfy college students’ psychological and cognitive needs, it may also have potential adverse effects (i.e, STS-SM). In order to prevent and reduce STS-SM among college students, it is important to study the personal factors and psychological constructs associated with STS-SM and to elucidate the potential mechanisms that may affect STS-SM.

1.1. Childhood trauma and STS-SM

Childhood trauma refers to all forms of physical, sexual and emotional/psychological abuse and neglect of children aged 0–17 years by parents, caregivers and other authority figures (World Health Organization, 2022). Childhood trauma has been identified as a risk factor for STS in professional helpers (Hensel et al., 2015; Leung et al., 2023; Nelson-Gardell & Harris, 2003; but see Schauben & Frazier, 1995). A recent review showed that 17 out of 26 studies reported a positive association between personal trauma history and STS in mental health workers (Leung et al., 2023), such as clinician (Bauwens & Tosone, 2014), child protection workers (Dagan et al., 2016), social workers and psychologists (Rayner et al., 2020). In addition, the total number of adverse childhood experiences has been found to be associated with STS symptoms in graduate social work students during clinical training (Butler et al., 2018). According to the conservation of resources (COR) theory (Hobfoll, 1989), a well-established framework for understanding stress and trauma in organizational settings, early resource losses can make it difficult for individuals to overcome losses in their later lives. Early traumatic experiences of mental health workers may increase their susceptibility to STS through chronic exposure to traumatic material that may re-trigger past unprocessed trauma (Figley, 1995a).

Theoretically, childhood trauma may also be associated with STS-SM. The vulnerability – stress model posits that childhood traumatic experiences are an important personal factor that may make individuals more vulnerable to the effects of later stressors (Ingram & Luxton, 2005). Childhood trauma can lead to long-term changes in the brain, emotional regulation and stress-response systems (Begemann et al., 2023; Fan et al., 2022; Liu et al., 2024a), which may create persistent vulnerability to stress. Social media has become a pervasive source of exposure to traumatic content, and this exposure may serve as a stressor that heightens the risk of experiencing STS (Pearson, 2024). Empirical research has confirmed the association between childhood trauma and STS-SM. A cross-sectional study found that STS-SM was positively correlated with the occurence and impact of prior trauma among 141 social media users (Mancini, 2019), yet prior trauma was measured with a single item without specifying when the trauma was encountered. Another study from our group showed a positive correlation between childhood trauma and STS caused by media exposure to the COVID-19 pandemic in 115 college students (Liu et al., 2024a). However, the two previous studies had small sample sizes or used non-standardized instruments. Against this background, the first aim of this study was to link childhood trauma with STS-SM in a large sample of college students (n = 1151) using well-validated standardized questionnaires. Based on prior evidence that multiple forms of abuse and neglect examined together are more predictive of mental health outcomes than specific types of trauma (Briere et al., 2008; Cloitre et al., 2009; Haselgruber et al., 2020), we focused on cumulative childhood trauma in this study, defined as the total score of the Childhood Trauma Questionnaire (see Methods and Materials for details). Based on the above theoretical and empirical findings, this study proposed Hypothesis 1: Childhood trauma is significantly and positively associated with STS-SM.

1.2. The mediation of self-compassion

Understanding how individual differences in childhood trauma predict STS-SM is crucial to developing effective strategies to prevent STS-SM. Self-compassion is a psychological construct derived from Buddhist philosophy that involves being kind to oneself when suffering, recognizing that one’s own experiences are part of the common human experience, and maintaining mindful or a balanced perspective when considering one’s own inadequacies and failures (Neff, 2003; Neff et al., 2007). It is positively associated with beneficial mental health outcomes such as life satisfaction, happiness, optimism and personal initiative (Neff, 2003, 2011; Neff et al., 2007; Zessin et al., 2015), and negatively related to psychopathology such as anxiety, depression, perceived stress and rumination (Lathren et al., 2019; MacBeth & Gumley, 2012; Neff, 2003).

According to attachment theory, early relationships with caregivers shape individuals’ internal working model of relationships and the way children treat themselves (Bowlby, 1988; Ross et al., 2019). Individuals who experience childhood trauma may internalize their caregivers’ negative views of themselves and the world and develop maladaptive schemas or self-concepts (Benjamin, 2003; Tanaka et al., 2011). In contrast, self-compassion is the ability to be kind and nonjudgmental toward oneself (Neff, 2003). Childhood trauma can negatively impact self-compassion by undermining the development of a positive self-concept and self-kindness (Tanaka et al., 2011). Indeed, research has found a consistent correlation between self-compassion and childhood maltreatment (for a review and meta-analysis, see Zhang et al., 2023b). Child neglect and abuse has been associated with less self-compassion in both children (Dai et al., 2024) and adults (Collins et al., 2023; Miron et al., 2014).

Furthermore, self-compassion is considered an important internal resource in COR theory to protect oneself from threats and manage stress (Hobfoll, 1989). Cross-sectional studies have shown that self-compassion is negatively associated with STS in healthcare professionals (Duarte et al., 2016; Harker et al., 2016; Ruiz-Fernández et al., 2021). Regression analyses also suggest that self-compassion significantly predicts STS in mental health professionals (Yazici & Özdemi̇r, 2022) and neonatal staff (Scott et al., 2021). Recent studies have demonstrated that self-compassion-focused interventions have a protective effect on PTSD symptoms in clinical and non-clinical samples (Luo et al., 2021; Winders et al., 2020). Overall, these findings suggest that self-compassion has a facilitative role in coping with trauma-related stress (Yazici & Özdemi̇r, 2022).

Summarizing, childhood trauma may affect self-compassion and STS-SM, and self-compassion may be associated with STS-SM. While no study has examined the mediating role of self-compassion in the relation between childhood trauma and STS, there is evidence that self-compassion mediates the relationship between childhood trauma and negative outcomes in adulthood such as depression (Tao et al., 2021), aggression (Zhang et al., 2023c) and PTSD symptoms (Barlow et al., 2017; Guo et al., 2021). Based on these theoretical and empirical findings, this study proposed Hypothesis 2: Self-compassion mediates the relationship between childhood trauma and STS-SM.

1.3. The mediation of resilience

Similarly, resilience may mediate the effect of childhood trauma on STS-SM. Resilience is a dynamic process of adaptation to life changes that enables individuals to cope with and recover from significant adversity, trauma, or stressful events (Rutter, 1985). It plays a role in fostering one’s well-being and the development of psychopathology such as PTSD in response to stress (Kaye-Kauderer et al., 2021).

Childhood trauma has a remote impact on early-adulthood resilience (Shen et al., 2021). According to ecological systems theory, a framework for understanding early childhood education and human development, the interpersonal relationships experienced by a developing person with their family and immediate caregivers constitute the child’s microsystem (Bronfenbrenner, 1979). Positive relationships within the microsystem promote healthy development, whereas maltreatment of children, such as neglect or abuse in this system, may lead to developmental disabilities (Algood et al., 2011). Childhood trauma is thought to disrupt the development of emotion regulation and coping mechanisms (Burke, 2024), which are essential components of resilience. A recent meta-analysis showed that there was a significant negative correlation between childhood trauma and resilience, and resilience mediated the association between childhood trauma and depression (Watters et al., 2023). These results suggest the impact of trauma on resilience and the protective role of resilience against mental health problems.

Consistent with COR theory, resilience is an important internal personal resource that can help prevent burnout and promote personal growth (Hobfoll, 1989). Studies have shown that resilience protects healthcare workers from trauma-related stress and risk of STS during the pandemic (Heath et al., 2020; Maiorano et al., 2020). In addition, Roden-Foreman et al. (2017) found that resilience was negatively associated with STS among clinician. Therefore, researchers have developed resilience-based interventions (e.g. reprocessing trauma, using the most effective coping strategies when exposed to trauma-related materials) to protect medical interns (Smith et al., 2021) and non-mental health professionals from STS (Kerig, 2019). Moreover, recent studies of the general public have shown that resilience has a protective effect against STS symptoms in the face of critical public events (Leys et al., 2021; Liu et al., 2023). Based on these theoretical and empirical findings, combined with the evidence that resilience mediates the association between childhood trauma and negative mental health outcomes such as general distress and self-harm behaviours (for a review, see Fritz et al., 2018), this study proposed Hypothesis 3: Resilience mediates the relationship between childhood trauma and STS-SM.

1.4. The chain mediation of self-compassion and resilience

Finally, self-compassion and resilience are closely related. Self-compassion serves as a key resource for promoting resilience and helping individuals cope more effectively with life challenges (Neff & McGehee, 2010; Warren et al., 2016). Previous studies have shown a positive correlation between self-compassion and resilience in the general population (Bluth et al., 2018; Pérez-Aranda et al., 2021), in clinical samples (Alizadeh et al., 2018; Baker et al., 2019), and at work (reviewed in Lefebvre et al., 2020). Longitudinal evidence suggests that self-compassion can predict future resilience (Eryılmaz et al., 2024) and has a prospective protective effect on emotional recovery following exposure to traumatic stressors in at-risk individuals (Zeller et al., 2015). Also, self-compassion training has been shown to improve resilience in nurses (Franco & Christie, 2021) and elderly women (Sadat et al., 2021). According to the model of tripartite affect regulation (Gilbert, 2005, 2015), a framework that explains how vulnerability can be treated using three emotion processing systems, self-compassion activates the soothing and contentment system while deactivating threat and drive systems during times of stress, thereby enabling adaptive emotion regulation and facilitating resilience (Meyer, 2015). Empirical research suggests that self-compassion may result in resilience by reappraising emotional difficulties (Vigna et al., 2018) and reducing physiological arousal (Kirschner et al., 2019). Several studies have shown that resilience mediates the effects of self-compassion on anxiety and depression (Pérez-Aranda et al., 2021; Yu et al., 2023). Furthermore, given that the components of self-compassion, i.e. treating oneself with kindness rather than self-criticism, bringing present-moment awareness to the distress and appraising one’s suffering as shared human experience rather than personal failures, are viewed as adaptive emotional coping approaches, it may serve as a natural resilience mechanism to buffer the impact of stressful events (Vigna & Strauss, 2023).

Together, based on all these findings and theoretical models, i.e. the attachment theory for understanding the relationship between childhood trauma and self-compassion (Bowlby, 1988; Ross et al., 2019), the model of tripartite affect regulation explaining the relationship between self-compassion and resilience (Gilbert, 2005, 2015), and the conservation of resources theory viewing self-compassion and resilience as internal resources to prevent STS (Harker et al., 2016; Hobfoll, 1989), we proposed Hypothesis 4: Self-compassion and resilience play chain mediating roles in the relationship between childhood trauma and STS-SM.

1.5. The current study

This cross-sectional study aimed to determine the association between childhood trauma and current STS induced by social media use, with a specific focus on exploring the mediating roles of self-compassion and resilience in this relationship using well-validated questionnaires. This study extends previous research in several ways. First, previous investigations about factors associated with STS have primarily focused on professional workers, while few studies have tested them in college students. Given the increasing number of social media users worldwide and exposure to traumatic posts/news reports (e.g. mass shootings, suicide, sexual assault, and mental illness), college students, who constitute the majority of social media users (Kemp, 2018), may vicariously experience traumatic events in the most vivid ways (Mancini, 2019). Therefore, we focused on a large sample of Chinese college students (n = 1151) in this study. Second, this study is the first to explore the underlying psychological mechanism of the effects of childhood trauma on STS-SM. Although previous theories and empirical studies have suggested that self-compassion and resilience may mediate the association between childhood trauma and STS, no research has directly examined the mediating effect, especially the sequential mediating effect, of self-compassion and resilience in the relationship between childhood and STS-SM. Third, previous studies generally did not control for recent life events and socioeconomic status (SES) when investigating the relationship between childhood trauma and STS (but see Liu et al., 2024a), which may have potential confounding effects (Mancini, 2019; Mock & Arai, 2010). The present study, however, was restricted to subjects who reported no recent severe life events and treated SES as a covariate to exclude the possibility that the association between childhood trauma and STS-SM was influenced by these factors.

2. Methods and materials

2.1. Participants

A total of 1326 healthy, young Chinese university students (521 females, age range = 16–26 years) participated in this cross-sectional study through convenience sampling between December 2023 and March 2024. All participants signed an informed consent form before participation. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2013. All procedures involving human subjects were approved by the Medical Research Ethics Committee of West China Hospital, Sichuan University.

All participants were from universities in Sichuan province recruited via a class presentation to introduce the project by the authors. Inclusion criteria were being undergraduate students from Chinese universities, having access to social media and being able to complete self-reported instruments. Participants were excluded if they: (1) reported current or history of psychiatry disorders (yes/no); (2) reported regular use of psychotropic substances and any other drugs (yes/no); (3) reported recent severe life events (yes/no); (4) had Beck Depression Inventory (BDI) scores > 63 (Beck et al., 1996); (5) had invalid responding, which included those who consistently selected the same option for all items or failed to pass the bogus items (e.g. I have five fingers on my left hand). These exclusion criteria were used to ensure sample homogeneity and exclude potential confounding factors that may skew the findings, thereby facilitating interpretations. A total of 1151 participants (438 females, age = 16–26 years, mean age = 18.99 ± 1.06 years) were finally included in the data analyses. The current sample size is much larger than the estimated sample size (n = 207) to detect medium-sized effects based on a standard power analysis, taking into account the 20% of incomplete surveys (Faul et al., 2009).

2.2. Measures

2.2.1. Childhood Trauma Questionnaire-Short Form (CTQ-SF)

The CTQ-SF was designed to measure the frequency and severity of exposure to various types of maltreatment during childhood (Bernstein et al., 2003). The scale consists of 25 clinical items separated into five sub-scales with five items each: emotional abuse, physical abuse, sexual abuse, emotional neglect, and physical neglect. Each item is scored on a 5-point Likert scale, ranging from 1 (never true) to 5 (very often true). A higher total score indicated a higher level of childhood trauma. The Chinese version of the CTQ-SF has demonstrated good reliability among Chinese youths (Fu & Yao, 2005; Jiang et al., 2018; Zhang et al., 2023c). The Cronbach’s α of this scale in the current sample was 0.80, indicating adequate internal reliability. In addition, the Cronbach's α for the emotional abuse, physical abuse, sexual abuse, emotional neglect, and physical neglect sub-scales were 0.84, 0.93, 0.92, 0.75, and 0.21, respectively, aligning with previous studies reporting similar range of Cronbach's α (0.61–0.90) for the subscales (Badenes-Ribera et al., 2024).

2.2.2. Secondary Traumatic Stress Scale (STSS)

The STSS is a 17-item scale utilized to assess secondary traumatic stress in professionals who provide services to victims of trauma (Bride et al., 2004), which demonstrated good reliability and validity across diverse populations (e.g. nurses, doctors and social workers, etc., Ariapooran et al., 2022; Benuto et al., 2021; He et al., 2022; Orrù et al., 2021). In the current study, we used a modified version of the STSS, adapting wording to the context of social media, to measure STS-SM among college students. Specifically, 8 items are identical to the STSS, and in the remaining 9 items, the terminology was changed from referring to ‘work with clients’ trauma’ to ‘trauma seen on social media’. Similar to the STSS, the modified STSS consists of three sub-scales measuring intrusion (5 items), avoidance (7 items) and arousal (5 items), respectively. Respondents indicate on a 5-point Likert scale (1 = never to 5 = very often) how often they experienced each of the 17 STS symptoms during the last week. A higher total score indicates a higher level of STS symptoms. The modified STSS has shown satisfactory psychometric properties, as evidenced by Cronbach’s α of 0.90 for all 17 items, as well as Cronbach’s α of 0.78–0.81 for the three subscales in previous studies (Comstock & Platania, 2017). The Cronbach’s α of the current sample was 0.92, indicating excellent internal reliability. In addition, the Cronbach's α for the intrusion, avoidance and arousal subscale were 0.82, 0.83 and 0.78, respectively. Given that the modified STSS as a complete scale possesses better reliability and predictive ability than each of the three sub-scales (Comstock & Platania, 2017), the current study focused solely on STSS total scores throughout analyses.

2.2.3. Self-compassion scale (SCS)

The SCS is a 26-item questionnaire designed to assess overall self-compassion (Neff, 2003). It covers six dimensions: self-kindness (5 items), common humanity (4 items), mindfulness (4 items), self-judgment (5 items), isolation (4 items), and over-identification (4 items). Responses to each item use a 5-point Likert scale ranging from 1 (almost never) to 5 (almost always). A higher total score indicates a greater level of self-compassion. The Chinese version of the scale has shown good internal consistency among Chinese youths (Chen et al., 2011; Zhang et al., 2023c). In this study, the Cronbach's α of SCS was 0.90, indicating excellent internal reliability.

2.2.4. Connor – Davidson resilience scale (CD-RISC)

The CD-RISC is a 10-item scale to measure resilience (Connor & Davidson, 2003). Each item is scored on a 5-point Likert scale ranging from 1 (not true at all) to 5 (true nearly all the time). The total score is obtained by summing the scores of all 10 items and a higher score indicates greater resilience. The Chinese version of the CD-RISC has demonstrated good internal consistency and test-retest reliability in healthy, young populations (Liu et al., 2023). In this study, the Cronbach's α of CD-RISC was 0.94, indicating excellent internal reliability.

2.2.5. Other variables

Other variables included sociodemographic characteristics such as age, sex, educational level, marital status, ethnicity and socioeconomic status (SES). The participants’ age was measured as the mean number of years of life, sex was categorized as female/male, educational level was categorized as freshman/sophomore/junior/senior undergraduate, marital status was categorized as single/married/divorced, ethnicity was categorized as Han Chinese/minority and SES was assessed by the Socioeconomic Status Scale (SSS). The SSS has been shown to be associated with the study variables in previous studies (Liu et al., 2023, 2024a). It measures subjective SES using a graphical representation of a ladder with 10 rungs (1 = the lowest rank; 10 = the highest rank) and participants were instructed to rank their parents’ income, education and occupational prestige levels both locally and nationally (Adler et al., 2000). The higher the ladder, the better the socioeconomic status. The participants’ sociodemographic characteristics are summarized in Table 1.

Table 1.

Sociodemographic characteristics of the participants.

Variable	Mean ± SD (N%/range)
Sex
Male	713 (61.9%)
Female	438 (38.1%)
Age (years)	18.99 ± 1.06 (16–26)
Educational level
Freshman for undergraduate	1045 (90.8%)
Sophomore for undergraduate	6 (0.5%)
Junior for undergraduate	37 (3.2%)
Senior for undergraduate	63 (5.5%)
Marital status
Single	1141 (99.1%)
Married	7 (0.6%)
Divorced	3 (0.3%)
Ethnicity
Han Chinese	1028 (89.3%)
Minority	123 (10.7%)
Subjective socioeconomic status	10.07 ± 3.03 (2–20)

Note: N, number; SD, standard deviation.

2.3. Data analysis

All behavioural data were analyzed using the statistical software SPSS 29.0 (SPSS Inc., Chicago, IL, USA). Due to the setup of the web-based survey, there was no missing data. We used Harman’s single-factor test for all items of the scales before data processing to assess the common method variance (Liu et al., 2024b; Podsakoff et al., 2003). Next, we performed descriptive statistical analyses of the demographic characteristics (such as sex, age, etc.). Furthermore, Pearson correlation analyses were conducted to investigate the bivariate relationships between these key variables (i.e. childhood trauma, self-compassion, resilience and STS-SM), providing a foundation for the subsequent chain mediation analyses.

To examine the mediating roles of self-compassion and resilience in the relation between childhood trauma and STS-SM, we conducted chain mediation analyses employing Model 6 of Hayes’ PROCESS (Version 3.5) macro in SPSS (Hayes, 2013). According to our hypothesized chain mediation model, there are four possible pathways linking childhood trauma to STS-SM: the first indirect pathway is through self-compassion (M1), the second is through resilience (M2), the third is through the sequential mediation of self-compassion (M1) and resilience (M2), and the last pathway is the direct link from childhood trauma (X) to STS-SM (Y). Notably, all variables were standardized. To control for potential confounding effects, all demographic variables including age, sex, marital status, educational level, ethnicity and SES were treated as covariates. Bootstrap confidence intervals (CIs) for the indirect effects were calculated using 5,000 random samples. The effect was considered significant if an empirical 95% CIs did not contain zero (Preacher & Hayes, 2008).

3. Results

3.1. Common method bias

A total of eleven eigenvalues greater than 1 were identified with the first factor accounting for 20.79% of the total explained variance, which was below the recommended threshold 40%, indicating that this study did not suffer from serious methodology bias (Podsakoff et al., 2003).

3.2. Descriptive statistics and bivariate correlations

Table 2 presents the mean, standard deviation (SD), range and Pearson correlations for all variables. Among the 1151 participants (438 females, 18.99 ± 1.06 years), we observed significant intercorrelations between all key variables. Specifically, childhood trauma were positively associated with STS-SM (r = 0.34, p < .001) and negatively correlated with self-compassion (r = −0.40, p < .001) and resilience (r = −0.35, p < .001). Self-compassion, resilience and STS-SM were also intercorrelated (ps < .001,Table 2).

Table 2.

Descriptive statistics and bivariate correlations of study measures.

Measure	Mean ± SD	Range	1	2	3	4	5	6	7	8	9	10	11	12
1. Sexa	–	–	–
2. Age	18.99 ± 1.06	16–26	−0.19**	–
3. SES	10.07 ± 3.03	1–10	−0.02	−0.09**	–
4. Childhood trauma	47.39 ± 15.36	28–104	−0.10**	0.07*	−0.06*	–
5. CT-ea	7.25 ± 3.39	5–25	−0.07*	0.03	−0.02	0.86**	–
6. CT-pa	6.45 ± 3.23	5–25	−0.13**	0.04	−0.00	0.83**	0.82**	–
7. CT-sa	6.34 ± 3.15	5–25	−0.16**	0.06*	0.04	0.77**	0.76**	0.80**	–
8. CT-en	11.17 ± 4.34	5–25	−0.01	0.06*	−0.10**	0.70**	0.40**	0.33**	0.27**	–
9. CT-pn	9.79 ± 3.38	5–20	−0.14**	0.12**	−0.10**	0.76**	0.56**	0.51**	0.47**	0.49**	–
10. Self-compassion	85.17 ± 11.75	48–129	0.07*	−0.07*	0.11**	−0.40***	−0.31**	−0.23**	−0.22**	−0.34**	−0.41**	–
11. Resilience	35.71 ± 7.64	10–50	−0.07*	−0.02	0.13**	−0.35***	−0.28**	−0.25**	−0.24**	−0.28**	−0.31**	0.49**	–
12. STS-SM	37.84 ± 10.99	17–73	0.07*	0.00	−0.11**	0.34***	0.33**	0.23**	0.21**	0.22**	0.30**	−0.48**	−0.38**	–

Abbreviations: CT-ea, Childhood trauma-emotional abuse; CT-pa, Childhood trauma-physical abuse; CT-sa, Childhood trauma-sexual abuse; CT-en, Childhood trauma-emotional neglect; CT-pn, Childhood trauma-physical neglect; SD, standard deviation; SES, socioeconomic status; STS-SM, social media-induced secondary traumatic stress. ^a Male, 0; Female, 1. *** p < .001; ** p < .01; * p < .05. N = 1151.

To control for potential confounding effects of sociodemographic characteristics on the links between these variables, the partial correlation analyses adjusting for age, sex, marital status, educational level, ethnicity and SES showed that the associations between the key variables remained significant (ps < .001, see Supplementary Table S1).

3.3. Chain mediating analysis

To test whether self-compassion and resilience serve as serial mediators in the relationship between childhood trauma and STS-SM, a chain mediation analysis was conducted utilizing Model 6 in the SPSS macro programme PROCESS. After sociodemographic characteristics were controlled, the total effect (β = 0.34, p < .001) and the direct effect (β = 0.15, p < .001) of childhood trauma on STS-SM were both significant (Table 3 and Figure 1). To test for chain mediating effects, a Bootstrap 95% CI was calculated. As shown in Table 3, the total indirect effect was statistically significant (indirect effect = 0.19, SE = 0.02, 95% CI = [0.16, 0.22]) accounting for 55.9% of the total variance. The indirect effect was derived from three paths: (1) childhood trauma influenced STS-SM through the mediating effect of self-compassion (indirect effect = 0.14, 95% CI = [0.11, 0.17]); (2) childhood trauma affected STS-SM through the mediating effect of resilience (indirect effect = 0.03, 95% CI = [0.01, 0.04]); (3) childhood trauma influenced STS-SM through the chain mediating effects of self-compassion and resilience (indirect effect = 0.02, 95% CI = [0.01, 0.03]). The three paths were all significant, indicating that self-compassion and resilience not only has a separate mediating effect on the relationship between childhood trauma and STS-SM but also play a chain mediating role. The Chain mediating model is presented in Figure 1.

Table 3.

Total, direct, and indirect effects of the chain mediation model.

Effect	Effect size	Bootstrap SE	Bootstrap 95% CI
Total effect (Childhood trauma → STS-SM)	0.34	0.03	[0.29, 0.40]
Direct effect (Childhood trauma → STS-SM)	0.15	0.03	[0.10, 0.21]
Indirect effect	0.19	0.02	[0.16, 0.22]
Childhood trauma → self-compassion → STS-SM	0.14	0.01	[0.11, 0.17]
Childhood trauma → resilience → STS-SM	0.03	0.01	[0.01, 0.04]
Childhood trauma → self-compassion → resilience → STS-SM	0.02	0.01	[0.01, 0.03]

Abbreviations: SE, standard error; CI, confidence interval; STS-SM, social media-induced secondary traumatic stress. Age, sex, educational level, marital status, ethnicity and socioeconomic status were treated as covariates in the model.

Figure 1.

Model of the mediating role of self-compassion and resilience in the relationship between childhood trauma and social media-induced secondary traumatic stress (STS-SM). Standardized regression coefficients were displayed in the path diagram. c, total effect; c’, direct effect; ***p < .001, **p < .01. Age, sex, educational level, marital status, ethnicity and socioeconomic status were treated as covariates in the model.

4. Discussion

To our knowledge, this is the first study demonstrating that individual differences in childhood trauma directly and indirectly affect social media-induced secondary traumatic stress in later life, with self-compassion and resilience as intervening variables. Specifically, childhood trauma affected STS-SM through the chain mediation of self-compassion and resilience, and these findings were independent of recent stressful events and sociodemographic characteristics. This study provides insights into the psychological mechanism by which early trauma affects secondary traumatic stress induced by social media among college students.

The negative association of cumulative childhood trauma with STS-SM was consistent with previous reports that trauma history (especially in childhood trauma) is an indicator of posttraumatic stress after a worker has indirectly experienced traumatic events (Nelson-Gardell & Harris, 2003; Williams et al., 2012; for reviews see, Lerias & Byrne, 2003; Leung et al., 2023). Butler et al. (2018) conducted a survey of social work students and found that more than three quarters of the sample had experienced one or more adverse childhood experiences before age 18. Further, the sum of adverse childhood events influenced the development of STS symptoms through the experience of retraumatization (i.e. reactivation of emotions/memories associated with past negative life events) (Butler et al., 2018) . It has been suggested that a combination of multiple types of childhood trauma confers the greatest risk for susceptibility to STS and that emotional abuse and neglect are the strongest predictors of STS (Bride et al., 2007), which aligns with the present findings (see Table 2). Our study extends previous literature by suggesting that cumulative childhood trauma has an impact on STS-SM in college students, possibly because childhood trauma makes individuals more vulnerable to trauma-related material on social media based on the vulnerability – stress model.

Consistent with Hypothesis 2, we found that the effect of childhood trauma on STS-SM could be partially explained by self-compassion. This is in line with previous research showing that self-compassion mediates the relationship between childhood trauma and negative outcomes in adulthood (Reffi et al., 2019), especially PTSD symptoms (Barlow et al., 2017; Guo et al., 2021). High levels of childhood trauma are associated with lower levels of self-compassion, possibly because negative childhood experiences may undermine an individual’s positive perception of self-worth, leading to a lack of self-kindness and care when facing difficulties (Neff & McGehee, 2010). For instance, college students who have experienced more trauma tend to blame themselves more and feel more resentful when they encounter personal failures and adversities (Baydemır et al., 2014). Moreover, individuals with low self-compassion are more likely to perceive stress (Hufnagle et al., 2018) and show higher stress levels when facing traumatic events directly (Winders et al., 2020) and indirectly (Ondrejková & Halamová, 2022). The present findings support attachment theory (Bowlby, 1988) and Gilbert’s social mentality theory (Gilbert, 2005), which states that childhood maltreatment can lead to self-criticism by internalizing the way the caregiver treated them. This may further produce stress responses and result in greater susceptibility to psychopathology according to the vulnerability – stress model (Ingram & Luxton, 2005). These findings have important implications for fostering self-compassion in individuals with childhood trauma to prevent STS-SM.

Additionally, the results of this study support Hypothesis 3 that resilience partially mediated the relationship between childhood trauma and STS-SM. This is consistent with previous studies showing that resilience mediates the association between childhood trauma and negative mental health outcomes (Ding et al., 2017; Fritz et al., 2018; Hu et al., 2015). For example, Vieira et al. (2020) found that all sub-types of trauma were associated with bipolar disorder and major depressive disorder, and resilience partly mediated the associations. This accords with the model of ‘three resilience system mechanism’ which postulates that resilience fosters mental health through harm reduction, protection, and promotion (Davydov et al., 2010). It is advocated to use trauma-informed approaches, that recognizes that everyone may have a trauma history, in a variety of settings to use this information to develop and implement strategies to enhance resilience and address the impact of trauma-related exposures (Allen et al., 2023; Butler et al., 2018; Scott et al., 2023). Prior neuroimaging evidence from our group suggested that childhood trauma influenced COVID-related STS through functional connectivity between frontoparietal and default-mode regions (Liu et al., 2024a) and resilience mediated the effect of functional connectivity in the default-mode network on STS (Liu et al., 2023). This provides a potential neural explanation for the mediating role of resilience in the relationship between childhood trauma and STS. Together with the present findings, it is practical to target resilience for psychological intervention and neuromodulation to ameliorate the potential deleterious effects of childhood trauma on STS-SM.

An important contribution of this study is the elucidation of a chain mediating mechanism in the relationship between childhood trauma and STS-SM, that is, the sequential indirect effect of self-compassion and resilience. This is consistent with previous findings that self-compassion and resilience mediate the association between childhood exposure to domestic violence and PTSD during COVID-19 pandemic (Chi et al., 2021). Higher levels of childhood trauma are associated with lower levels of self-compassion, which contributed to lower levels of resilience and ultimately lead to higher levels of secondary traumatic stress due to social media use, highlighting the potential for enhancing self-compassion and resilience to prevent STS-SM. According to the COR theory, people strive to maintain their current resources and pursue new resources (Hobfoll, 1989). Self-compassion and resilience are perceived as important personal resources and play protective roles against burnout and STS (Harker et al., 2016; Ruiz-Fernández et al., 2021). Previous studies have demonstrated that people with self-compassion tend to develop positive coping strategies (Barnard & Curry, 2011) and emotional intelligence (Neff et al., 2005), and implement health-promoting behaviours (Sirois et al., 2015), all of which are elements of building resilience. Specifically, the balanced perspective and lack of harsh self-judgement of a self-compassionate person may facilitate ‘bouncing back’ from life’s difficulties, just as a resilient person may recover from challenging circumstances (Warren et al., 2016), consistent with the model of tripartite affect regulation (Gilbert, 2005). Thus, higher early life stress exposure may prevent people from developing a warm and nonjudgmental cognitive schema of self-compassion (Tanaka et al., 2011), leading to a depletion of resilience resource and rendering them vulnerable to negative impact of environmental factors later in life, such as frequent exposure to trauma via social media, which can induce secondary traumatic stress.

5. Conclusion

In conclusion, our study found that childhood trauma not only directly affects social media-induce secondary traumatic stress but also has indirect effects through self-compassion, resilience and self-compassion-resilience among college students. These findings may have implications to reduce secondary traumatic stress and its negative outcomes such as depression (Mancini, 2019) and anxiety (Comstock & Platania, 2017). Psychological interventions aiming at strengthening self-compassion and resilience are suggested to be developed and adopted to manage individuals’ response to stress. For example, self-compassion training (e.g. compassion-focused therapy) can be implemented as an intervention to reduce self-criticism and rumination and enhance resilience in people who have experienced early life trauma, thereby buffering the effects of traumatic information and stressors (Bluth et al., 2018; Hofmann et al., 2011).

6. Limitations of the study

Our study has several limitations. First, the cross-sectional nature of our data limits the ability to draw causal inferences, so the results need to be interpreted with caution, and the implications of the current findings warrant further investigation. Second, although this study provides preliminary support for a potential chain mediating mechanism between childhood trauma and STS-SM, further longitudinal research is needed to test the proposed theoretical model and other alternative models before any strong conclusions can be articulated on the role of self-compassion and resilience. Another limitation is the use of convenience sampling method and strict exclusion criteria which only includes healthy Chinese college students without recent trauma experiences. It is important to take into account the nature of the sample when considering the generalizability of the current findings, and future research using samples from different demographics (e.g. age, culture, education etc.) and clinical backgrounds (e.g. psychiatric condition, substance use etc.) are welcome. In addition, all data are self-reported, which may be biased by social expectations, memory issues, or emotional states (Kormos & Gifford, 2014). Future studies should consider employing more objective evaluations (e.g. others report) to precisely measure these variables. Finally, previous neuroimaging studies have elucidated the neural mechanisms relating STS to childhood trauma (Liu et al., 2024a) and resilience (Liu et al., 2023) separately, yet how these variables are intertwined in the brain remains unknown. Future neuroimaging studies are needed to uncover the neural mechanisms of the associations among childhood trauma, self-compassion, resilience and STS-SM.

Funding Statement

This work was supported by China National Postdoctoral Program for Innovative Talents (Grant No. BX20240238), China Postdoctoral Science Foundation (Grant No.2024M762230), Natural Science Foundation of Sichuan Province (Grant No. 2024NSFSC1772), Postdoctoral Research and Development Fund of West China Hospital of Sichuan University (Grant No. 2024HXBH089), Sichuan University Interdisciplinary Innovation Fund, and Philosophy and Social Science Foundation of Sichuan Province (Grant No. SCJJ24ND204). The funding sources had no involvement in the study design, data collection and analysis, results interpretation, writing or decision to publish of the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data related to this study can be obtained from OSF (https://osf.io/xqbsa/).