Childhood trauma and real-world social experiences in psychosis

Abstract

Background:

Childhood trauma is associated with a variety of negative outcomes in psychosis, but it is unclear clear if childhood trauma affects day-to-day social experiences. We aimed to examine the association between childhood trauma and functional and structural characteristics of real-world social relationships in psychosis.

Methods:

Participants with psychotic disorders or affective disorders with psychosis completed ecological momentary assessments (EMAs) over ten days (N = 209). Childhood trauma was assessed retrospectively using the Childhood Trauma Questionnaire. Associations between childhood trauma and EMA-assessed social behavior and perceptions were examined using linear mixed models. Analyses were adjusted for sociodemographic characteristics and psychotic and depressive symptom severity.

Results:

Higher levels of childhood trauma were associated with more perceived threat (B = −0.19, 95 % CI [−0.33, −0.04]) and negative self-perception (B = −0.18, 95 % CI [−0.34, −0.01]) during recent social interactions, as well as reduced social motivation (B = −0.29, 95 % CI [−0.47, −0.10]), higher desire for social avoidance (B = 0.34, 95 % CI [0.14, 0.55]), and lower sense of belongingness (B = −0.24, 95 % CI [−0.42, −0.06]). These negative social perceptions were mainly linked with emotional abuse and emotional neglect. In addition, paranoia was more strongly associated with negative social perceptions in individuals with high versus low levels of trauma. Childhood trauma was not associated with frequency (i.e., time spent alone) or type of social interactions.

Conclusion:

Childhood trauma – particularly emotional abuse and neglect – is associated with negative social perceptions but not frequency of real-world social interactions. Our findings suggest that childhood trauma may affect day-to-day social experiences beyond its association with psychosis.

1. Introduction

Psychotic disorders are characterized by impairments in social functioning (Harvey and Strassnig, 2012). Compared to healthy controls, individuals with psychosis tend to spend more time alone (Granholm et al., 2020; Oorschot et al., 2012), report less pleasure from social interactions (Barkus and Badcock, 2019), and experience more loneliness (Michalska da Rocha et al., 2018) and social stress (Mote and Fulford, 2020). Furthermore, social relationships are closely related to well-being and quality of life (Leigh-Hunt et al., 2017; Yanos et al., 2001), and may play a central role in the recovery from severe mental illness (Schön et al., 2009; Tew et al., 2012). Although psychotic symptoms, depression, and cognitive deficits partly account for poor social functioning in psychosis (Bowie et al., 2006), a substantial amount of the variation in social functioning remains unexplained.

A potential factor that may contribute to poor social functioning in psychosis is childhood trauma — a robust risk factor for psychosis (Alameda et al., 2021; Bendall et al., 2008; Varese et al., 2012). Exposure to childhood trauma may interfere with social development as it can involve disturbances in attachment, interpersonal violence, and deprivation of social interactions (Cyr et al., 2010; Doyle and Cicchetti, 2017). Childhood trauma may also disrupt a child’s feeling of safety and lead to negative perceptions of self and others (Pilkington et al., 2021), which in turn may negatively affect social relationships.

Although a few studies reported that childhood trauma is associated with poor social functioning in psychosis (Alameda et al., 2017; Hjelseng et al., 2022; Turner et al., 2020; van Nierop et al., 2016), it is unclear whether childhood trauma affects both structural and functional aspects of social relationships. Structural aspects refer to quantitative characteristics (e.g., frequency and type of interactions), whereas functional aspects refer to qualitative characteristics (e.g., appraisal of interactions) (Valtorta et al., 2016). In addition, it is largely unknown whether specific forms of childhood trauma are differentially related to social outcomes in psychosis. This is important given the increasing recognition of using dimensional approaches to study the effects of childhood trauma in addition to examining global effects (McLaughlin and Sheridan, 2016). Furthermore, it is unclear whether childhood trauma may affect the association between psychotic symptoms and social experiences, which could have clinical implications.

Prior studies mainly relied on retrospective reports of social functioning. Such measures may suffer from recall bias and limit insight into context-specific social experiences on a moment-to-moment basis. Ecological momentary assessment (EMA) is a repeated-measures technique that allows for the investigation of social experiences across naturalistic settings (Hektner et al., 2007). Rather than relying on recall and global judgments over a past period, EMA provides insight into social experiences as they unfold in everyday life, thereby enhancing ecological validity (Granholm et al., 2020; Myin-Germeys et al., 2018).

The goal of this EMA study is to investigate the association of childhood trauma with structural and functional characteristics of real-world social relationships in psychosis. We hypothesized that childhood trauma would be associated with a low frequency of social interactions and negative social perceptions, such as negative appraisals of interactions, social demotivation, and desire to avoid others. Given the limited evidence to date, we did not have a-priori hypotheses about whether associations would be specific to types of childhood trauma or driven by a more global effect. Furthermore, we expected that childhood trauma would moderate the association between psychotic symptoms and negative social experiences.

2. Material and methods

2.1. Participants

This observational study is embedded in an ongoing longitudinal study of the relationship between social cognition and suicide in severe mental illness. The study is conducted across three sites in the United States (University of California, San Diego, University of Miami, and University of Texas-Dallas) (Depp et al., 2021; Parrish et al., 2021). The study was approved by each site’s Institutional Review Board, and participants completed an assessment of capacity to consent to research before written informed consent.

The participants included in the sample all had a Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) diagnosis of schizophrenia, schizoaffective disorder, bipolar disorder, or major depressive disorder and presented with lifetime or current psychotic symptoms, as defined by the Structured Clinical Interview for the DSM-5 (SCID 5; First et al., 2015) and Mini International Neuropsychiatric Interview (MINI; Sheehan et al., 1998). Overall, the study sample expressed a low severity of mania (Depp et al., 2021). Other eligibility criteria were: 1) proficiency in English, 2) aged 18–65 years, 3) in outpatient care, and 4) availability of an informant with regular contact. Exclusion criteria were: 1) hospitalizations or dose changes within the last 6 weeks, 2) history of head trauma or a neurological/neurodegenerative disorder, 3) DSM-5 substance use disorder in the past three months (except tobacco and cannabis), as determined by the SCID-5 (First et al., 2015), and 4) IQ level below 70 as estimated by the Wide Range Achievement Test-4 (WRAT-4; Wilkinson and Robertson, 2006).

This study used baseline data from a subsample of the target population enrolled between October 2019 and January 2022. Of the 238 participants who were enrolled, 222 participants had data available on childhood trauma and participated in EMA. Consistent with prior studies (Badal et al., 2021; Granholm et al., 2020), we excluded participants who completed fewer than 33 % of surveys (n = 13; 6 %). This resulted in a final sample of N = 209 participants.

2.2. Measures

2.2.1. Childhood trauma

The Childhood Trauma Questionnaire (CTQ) is a well-validated 28-item self-report questionnaire that retrospectively assesses childhood trauma (Bernstein et al., 1998). The CTQ includes five subscales: emotional, physical, and sexual abuse, and emotional and physical neglect. Each subscale includes five items, which are scored on a five-point Likert scale ranging from (1) “never true” to (5) “very often true.” Besides using the total score as a continuous variable (range 25–125), we dichotomized the variable into “none to low-to-moderate” (score ≤ 51) and “moderate-to-severe to extreme” (score > 51) levels of childhood trauma (Bernstein et al., 1998). This dichotomous variable was used only in stratified analyses to help interpret any interaction effects.

2.2.2. Ecological momentary assessment (EMA)

Participants were assessed three times a day for ten days. Text notifications were sent to participants’ (own or lab-provided) smartphones containing a link to complete the surveys. Participants could indicate a preferred one-hour time slot during the morning, afternoon, and evening, with a minimum duration of 2 h in between each survey. The link stayed active for 1 h after being sent. For each survey completed, participants received $1.66 with a maximum of $50. Participants were contacted by research assistants on the first day or if they missed more than three consecutive surveys to resolve potential concerns and maintain survey adherence.

2.2.2.1. Structural aspects of social relationships.

First, participants were asked who they were with at the time of the survey, which was categorized into “alone” versus “with people” (Granholm et al., 2020). The type of social interaction was further categorized into “with friends/family” (i.e., friends, spouse or partner, and other family members) and “with others” (i.e., strangers, roommates, co-workers, healthcare providers, and others). Second, participants indicated with whom they had interacted since the last survey. Of note, these responses could therefore extend beyond physical interactions (e.g., telephone contact). This question was categorized in the same way as the previous question.

2.2.2.2. Functional aspects of social relationships.

Participants who interacted since the past survey rated the 1) pleasure felt during these interactions, which was assessed on a seven-point Likert scale ranging from (1) “not at all” to (7) “very much,” 2) perceived level of threat during these interactions, which was assessed on a seven-point Likert scale ranging from (1) “on guard or threatened” to (7) “trusting or warm,” and 3) self-perception during these interactions, which was assessed on a seven-point Likert scale ranging from (1) “unlikeable or inferior” to (7) “likable or capable” (Parrish et al., 2022). Participants who did not interact since the past survey rated their social motivation/interest during this period. Finally, all participants rated their social motivation/interest and desire to avoid others throughout the rest of the day, as well as their feelings of burdensomeness and belongingness. These questions were rated on a seven-point Likert scale ranging from (1) “not at all” to (7) “very much.”

2.2.2.3. Psychotic symptoms.

Participants rated the extent to which they had been bothered by 1) voices and 2) suspicious thoughts since the past survey, which were both rated on a seven-point Likert scale ranging from (1) “not at all” to (7) “very much.”

2.2.3. Covariates

The following sociodemographic covariates were included: sex, age, race (White/Black or African American/Other), ethnicity (Hispanic/non-Hispanic), and years of education. Clinically-rated psychotic symptom severity was assessed by the Positive and Negative Syndrome Scale (PANSS; Kay et al., 1987) and depression severity was assessed by the Montgomery–Åsberg Depression Rating Scale (MADRS; Montgomery and Åsberg, 1979).

2.3. Statistical analysis

We first assessed the association of total childhood trauma scores with structural and functional aspects of social relationships using linear mixed models with random intercepts for subjects. Only if a statistically significant association was observed, analyses were repeated using trauma subscales. In a secondary step, we repeated the main analysis in subgroups of participants with non-affective psychosis (i.e., schizophrenia and schizoaffective disorder; n = 128) and affective psychosis (i. e., bipolar disorder and major depressive disorder with psychotic features; n = 81). Second, using linear mixed models, we investigated the statistical interaction between childhood trauma and psychotic symptoms in relation to social experiences. To elaborate on significant interactions, we conducted stratified analyses in participants with higher versus lower levels of childhood trauma. Given that the independent and dependent variables were both measured using EMA in these latter analyses, we were able to evaluate within-person associations, i.e., momentary deviations from person-averaged mean levels.

Analyses were adjusted for covariates in a stepwise manner. Model 1 was adjusted for sex, age, race, ethnicity, and years of education. Model 2 was additionally adjusted for positive and negative symptom severity and depression severity. Of note, positive symptom severity was not included as a covariate in the interaction analyses because it assesses the same underlying constructs as the EMA-assessed psychotic symptoms. Childhood trauma scores and psychotic symptoms were converted into Z-scores for ease of interpretation. A false discovery rate (FDR) correction was applied to control for type I errors for each set of analyses separately (Benjamini and Hochberg, 1995). The level of statistical significance was set at p = .05. All analyses were performed using SPSS Statistics v.28.

3. Results

3.1. Sample characteristics

The characteristics of the study sample (N = 209) are shown in Table 1. The majority of participants had a history of childhood trauma (80.9 %). Overall, 81.3 % of all EMA surveys were completed, with an average of 24.5 surveys per participant. Since we did not exclude incomplete EMA surveys, there was a slightly different number of EMA observations for each study outcome (ranging from 5095 to 5125 EMA observations per analysis). Descriptive characteristics of all EMA variables can be found in Table S1.

Table 1.

Characteristics of the study sample.

Characteristic	Study sample (N = 209)
	n
Sex, % female	136	65.1
Age (years), mean (SD)	209	41.7 (11.6)
Race, %	209
White	80	38.3
Black or African American	83	39.7
Other	46	22.0
Ethnicity, % Hispanic	47	22.5
Education in years, mean (SD)	209	13.6 (2.4)
Primary diagnosis, %	209
Schizophrenia	59	28.2
Schizoaffective disorder	69	33.0
Bipolar disorder with psychotic features	76	36.4
Major depressive disorder with psychotic features	5	2.4
Childhood trauma (CTQ), mean (SD), range
Total score	209	58.8 (22.5), 25–121
Emotional abuse	209	13.8 (6.1), 5–25
Physical abuse	209	10.8 (5.8), 5–25
Sexual abuse	209	11.5 (7.3), 5–25
Emotional neglect	209	13.1 (5.3), 5–25
Physical neglect	209	9.6 (4.2), 5–22
Symptom severity, mean (SD)
Positive symptoms (PANSS)	209	16.3 (5.6)
Negative symptoms (PANSS)	209	11.9 (4.1)
Depression severity (MADRS)	209	14.1 (11.3)

CTQ = Childhood Trauma Questionnaire; PANSS = Positive and Negative Syndrome Scale; MADRS = Montgomery–Åsberg Depression Rating Scale.

3.2. Structural aspects of social relationships

Childhood trauma was not associated with the structural aspects of social relationships, including time spent alone and time spent with friends/family versus other people (Table 2). This was consistently observed both for social interactions during the survey and since the past survey. We similarly detected no associations between childhood trauma and social behavior in subgroups of participants with non-affective and affective psychosis.

Table 2.

The association between childhood trauma and structural aspects of social relationships.

Exposure	Outcome	Model 1			Model 2
		OR	95 % CI	p	OR	95 % CI	p
Childhood trauma (total CTQ score)a	Alone (vs. interacted with people)b
	During survey	1.04	0.84–1.30	.72	1.03	0.82–1.29	.77
	Since previous survey	1.00	0.79–1.27	.98	0.98	0.77–1.24	.86
	With others (vs. with friends/family)c
	During survey	0.82	0.58–1.17	.28	0.85	0.59–1.22	.38
	Since previous survey	1.18	0.82–1.69	.37	1.24	0.87–1.78	.23

Model 1 is adjusted for sex, age, race, ethnicity, and years of education. Model 2 is additionally adjusted for positive and negative symptom severity (PANSS) and depression severity (MADRS). Model estimates are based on linear mixed models with random intercepts for subjects.

^aChildhood trauma scores are standardized (M = 0, SD = 1).

^bInteracted with people = reference category.

^cInteracted with friends/family = reference category.

We observed an interaction between childhood trauma and paranoia in relation to being alone during the survey (Table S2). Stratified analyses revealed that, in individuals with higher levels of childhood trauma, momentary within-person increases in paranoia were associated with a higher likelihood of being alone (OR = 1.13, 95 % CI [1.07–1.21]). In contrast, this association was not present in individuals with lower levels of childhood trauma (OR = 1.01, 95 % CI [0.93–1.10]). However, this interaction was not statistically significant when considering multiple testing.

3.3. Functional aspects of social relationships

Participants with higher levels of childhood trauma appraised their recent social interactions more negatively (Table 3). More specifically, a higher level of childhood trauma was associated with a lower level of pleasure (B = −0.18, 95 % CI [−0.35, −0.02]), higher level of perceived threat (B = −0.24, 95 % CI [−0.39, −0.09]), and more negative self-perception (B = −0.22, 95 % CI [−0.38, −0.05]) during recent social interactions. For individuals who had not interacted since the past survey, higher levels of childhood trauma were associated with less social motivation during this period (B = −0.35, 95 % CI [−0.57, −0.13]). Furthermore, individuals with higher levels of childhood trauma were less motivated to socially interact (B = −0.34, 95 % CI [−0.54, −0.15]) and expressed a greater desire to avoid people (B = −0.35, 95 % CI [−0.57, −0.13]) throughout later that day. Finally, higher levels of childhood trauma were associated with increased feelings of burdensomeness (B = 0.28, 95 % CI [0.06, 0.49]) and decreased feelings of belongingness (B = −0.35, 95 % CI [−0.57, −0.13]). Importantly, these associations remained after adjustment for psychotic and depression symptom severity, except for the associations of childhood trauma with pleasure and burdensomeness (Table 3).

Table 3.

The association between childhood trauma and functional aspects of social relationships.

Exposure	Outcome	Model 1			Model 2
		B	95 % CI	p	B	95 % CI	p
Childhood trauma (total CTQ score)a	Appraisal of recent interactions
	Pleasure	−0.18	−0.35, −0.02	.032*	−0.12	−0.28, 0.03	.125
	Trust/warmth	−0.24	−0.39, −0.09	.002*	−0.19	−0.33, −0.04	.021*
	Positive self-perception	−0.22	−0.38, −0.05	.010*	−0.18	−0.34, −0.01	.033*
	Social motivation and avoidance
	Motivation since prior survey	−0.35	−0.57, −0.13	.002*	−0.28	−0.50, −0.05	.015*
	Motivation future interactions	−0.34	−0.54, −0.15	<.001*	−0.29	−0.47, −0.10	.003*
	Avoidance future interactions	0.39	0.18, 0.60	<.001*	0.34	0.14, 0.55	.001*
	Burdensomeness and belongingness
	Burden	0.28	0.06, 0.49	.012*	0.20	0.00, 0.41	.051
	Belonging	−0.31	−0.50, −0.12	.001*	−0.24	−0.42, −0.06	.009*

Model 1 is adjusted for sex, age, race, ethnicity, and years of education. Model 2 is additionally adjusted for positive and negative symptom severity (PANSS) and depression severity (MADRS). Model estimates are based on linear mixed models with random intercepts for subjects.

^aChildhood trauma scores are standardized (M = 0, SD = 1).

^*Statistically significant after FDR correction.

We repeated the analyses in subgroups of non-affective and affective psychosis (Tables S3 and S4). While several associations were slightly more pronounced in participants with non-affective psychosis (i.e., for social motivation, avoidance, and belongingness), the associations were not statistically significantly different from those in participants with affective psychosis.

3.3.1. Types of childhood trauma in relation to social perceptions

In secondary analyses, we examined the associations between types of childhood trauma and functional aspects of social relationships (Table S5). The strongest associations were observed for emotional abuse and emotional neglect. More specifically, after adjusting for multiple testing and symptom severity, emotional abuse was associated with all negative social perceptions, and emotional neglect was associated with a higher level of perceived threat, more negative self-perception, lower social motivation, greater desire to avoid others, and lower sense of belongingness. In contrast, physical abuse was only significantly associated with a greater desire to avoid others, whereas physical neglect was only significantly associated with a higher level of perceived threat. We found no evidence for associations between sexual abuse and any social perceptions.

3.3.2. Interaction between childhood trauma and psychotic symptoms in relation to social perceptions

We observed statistical interactions between childhood trauma and paranoia in relation to pleasure, perceived level of threat, negative self-perception, and belongingness (Table S6). Stratified analyses revealed that the association of momentary, within-person increases in paranoia with negative social perceptions was greater in individuals with higher levels of childhood trauma (e.g., pleasure: B = −0.25, 95 % CI [−0.29, −0.21]) than in individuals with lower levels of childhood trauma (e.g., pleasure: B = −0.14, 95 % CI [−0.19, −0.10]; Table 4). Although we also observed statistical interactions between childhood trauma and voice-hearing in relation to pleasure and perceived level of threat, these did not survive adjustment for multiple testing (Table S7).

Table 4.

The within-person (momentary) associations between paranoia and functional aspects of social relationships in participants with lower and higher levels of trauma.

Exposure	Outcome	None to low-moderate childhood trauma (n = 90)			Moderate-severe to extreme childhood trauma (n = 119)
		B	95 % CI	p	B	95 % CI	p
Momentary paranoia	Pleasure	−0.14	−0.19, −0.10	<.001	−0.25	−0.29, −0.21	<.001
	Trust/warmth	−0.15	−0.19, −0.11	<.001	−0.26	−0.30, −0.22	<.001
	Positive self-perception	−0.13	−0.17, −0.09	<.001	−0.21	−0.24, −0.17	<.001
	Belonging	−0.19	−0.23, −0.14	<.001	−0.22	−0.25, −0.18	<.001

Analyses are adjusted for person-averaged levels of paranoia, sex, age, race, ethnicity, years of education, negative symptom severity (PANSS), and depression severity (MADRS). Model estimates are based on linear mixed models with random intercepts for subjects. Momentary paranoia is the unstandardized deviation in paranoia from the person-averaged level of paranoia during each EMA assessment.

4. Discussion

This study found that childhood trauma is associated with negative social experiences in the everyday lives of adults with psychosis. Several key findings emerged. First, childhood trauma was associated with functional aspects of social relationships, including negative appraisals of recent social interactions, low social motivation, desire for social avoidance, and low sense of belongingness — even after accounting for psychotic and depressive symptom severity. These negative social perceptions were mainly linked to emotional trauma rather than physical or sexual trauma. Second, childhood trauma moderated the association between paranoia and negative social perceptions. Specifically, paranoia was more strongly associated with negative appraisals of recent social interactions and a low sense of belongingness in individuals with high versus low levels of trauma. Third, we found no evidence that childhood trauma was associated with structural aspects of social relationships, including the frequency and type of social interactions.

Most prior studies on childhood trauma and social dysfunction in psychosis used global measures of social functioning (Alameda et al., 2017; Hjelseng et al., 2022; Turner et al., 2020; van Nierop et al., 2016). While such composite scores indicate overall impairment, they provide limited insight into subdomains of social functioning. Our findings suggest that, in the context of psychosis, childhood trauma may be primarily related to functional (i.e., qualitative and subjective) rather than structural (i.e., quantitative) aspects of social relationships. This aligns with previous findings of the association between childhood trauma and loneliness (Steenkamp et al., 2022), social avoidance (Boyette et al., 2014), and social functioning dissatisfaction (Stain et al., 2014) in individuals with psychosis. Consistent with our observation, Stain et al. (2014) did not find a lower frequency of social interactions in individuals with childhood trauma. Our study extends these prior findings by using momentary data collected in a naturalistic setting, thereby reducing recall bias and improving ecological validity in assessing social experiences (Granholm et al., 2020; Myin-Germeys et al., 2009). Given the limited research to date, and particularly the lack of EMA studies, future work is required to increase our understanding of the relationship between childhood trauma and various domains of social functioning in psychosis. For instance, the blended use of active and passive EMA, such as GPS tracking (Depp et al., 2019), may allow for a more continuous and fine-grained assessment of daily social experiences and activities. In addition, it will be relevant to study whether the observed associations are specific to the social domain, i.e., whether individuals with childhood trauma specifically appraise their social interactions more negatively or their environment in general (e.g., other daily-life activities).

Our findings suggested that emotional abuse and neglect were associated with a range of negative social perceptions, whereas physical abuse and neglect were associated with few negative social perceptions. Surprisingly, sexual abuse was associated with none of the studied social perceptions. These findings indicate that there may be some degree of specificity of individual trauma types in predicting later negative social perceptions. Our results align with a recent study in schizophrenia and bipolar spectrum disorders showing that emotional abuse and neglect had the strongest association with global impairment in social functioning (Hjelseng et al., 2022). Moreover, a recent meta-analysis reported that individuals with a history of emotional abuse and/or neglect were more vulnerable to loneliness than individuals with a history of physical or sexual trauma (de Heer et al., 2022). Nevertheless, other studies reported mixed findings with regard to the impact of different subtypes of trauma on social functioning (Lysaker et al., 2001; Trotta et al., 2016; Turner et al., 2020), and it may be that the accumulation of trauma across subtypes is particularly relevant for negative life outcomes (Copeland et al., 2018; Dunn et al., 2018), including poor social functioning in psychosis (Hjelseng et al., 2022).

There are a variety of mechanisms that could explain the relationship between childhood trauma and negative social perceptions in psychosis. Childhood trauma is a predictor of more severe psychotic and depressive symptoms (Aas et al., 2016; Read et al., 2005; Stanton et al., 2020), which can contribute to poor functional outcomes (Alameda et al., 2017; Bowie et al., 2006). In the current study, we found that childhood trauma moderated the association between paranoia and negative social perceptions, including negative appraisals of recent social interactions and a low sense of belongingness. These results extend previous findings of the association between paranoia and poor social functioning (Hajduk et al., 2019; Pinkham et al., 2016) by indicating that this association is stronger in patients with a history of childhood trauma and present on a momentary, within-person level, i.e., negative social perceptions when people start to feel more paranoid than their average level of paranoia. Consistently, an increasing body of evidence suggests that childhood trauma is associated with sensitized psychotic reactivity to social stressors in daily life (Myin-Germeys et al., 2018; Rauschenberg et al., 2017; Reininghaus et al., 2016). Future studies may seek to explore the potential impact of childhood trauma on dynamic links between paranoia and social perceptions over time, such as by using dynamic network analysis of EMA data (Badal et al., 2021).

Another mechanism could be that childhood trauma affects the formation of secure parental attachment (Cyr et al., 2010; Doyle and Cicchetti, 2017). Insecure attachment is relatively stable throughout life and can affect the quality of interpersonal relationships (Candel and Turliuc, 2019; Shaver and Mikulincer, 2002). Similarly, parental bonding and parenting styles have been related to the qualitative (but not quantitative) aspects of daily-life social interactions (Achterhof et al., 2021; Achterhof et al., 2022). Furthermore, exposure to childhood trauma may interfere with cognitive development, including impaired neurocognition (Rodriguez et al., 2021) and social cognition (Rokita et al., 2018). Emerging evidence suggests that childhood trauma is associated with social cognitive deficits in psychosis (Chalker et al., 2022; Dauvermann and Donohoe, 2019) and that these deficits are related to impaired social functioning (Fett et al., 2011; Parrish et al., 2022). The relationship between trauma and negative social perceptions may further be explained through mechanisms of stress sensitivity (Lardinois et al., 2011), emotion dysregulation (Lavi et al., 2019), negative cognitive schemas (Pilkington et al., 2021), and hypervigilance to threat (de Heer et al., 2022; Reininghaus et al., 2016). Other mediating factors may include living and employment situations, substance use, and illness-related characteristics (e.g., medication use). Alternatively, while we adjusted for a range of covariates, there might be unmeasured confounding (e.g., socio-environmental or genetic vulnerabilities) that could partly explain the association between childhood trauma and social experiences.

The current findings might have several clinical implications. The social demotivation and avoidance observed in psychosis may partly be attributed to the long-reaching consequences of adverse experiences in childhood. Childhood trauma likely affects the quality of social relationships as it may shape how individuals appraise their social interactions and whether they feel connected to other people. Given that a low sense of belongingness is a significant risk factor for suicidality (Chu et al., 2017; Van Orden et al., 2010), these findings further highlight the role of childhood trauma in suicide risk (Angelakis et al., 2019), and may point towards potential mediating pathways that may be subject to intervention. Taking these findings together, it might be important to routinely assess childhood trauma in individuals with psychosis, also considering that childhood trauma may exacerbate the interrelationship between paranoia and social dysfunction. Of note, trauma-focused therapies are rarely offered to individuals with psychosis, because of fear of symptom exacerbation and adverse events (Becker et al., 2004). Encouragingly, however, emerging evidence suggests that trauma-focused therapies are effective and safe in individuals with psychotic disorders and comorbid posttraumatic stress disorder (PTSD; Mueser et al., 2008; van den Berg et al., 2015, 2016). Further research is needed to determine if these outcomes extend to more positive social perceptions. In addition, future research on childhood trauma and social functioning in psychosis would benefit from examining to what extent social functioning impairments (e.g., avoidance) might be accounted for by the presence of PTSD.

This study is strengthened by the large sample of adults with psychotic disorders and affective disorders with psychotic features, and the investigation of real-time social experiences using EMA. However, several limitations should be discussed. First, while the CTQ is a widely used and well-validated scale to assess childhood trauma (Saini et al., 2019), it is not without limitations. The CTQ does not assess the timing or duration of events and its retrospective nature may be subject to recall bias (Williams, 1994). Although we adjusted for symptom severity to minimize bias by current mental health status, other illness-related factors may have affected trauma reports, and future studies using prospective designs are needed to replicate the current findings. Second, other domains of childhood adversity could have been relevant to consider, including bullying, parental loss (McLaughlin et al., 2012), and systems-level factors (e.g., crime exposure and social fragmentation) (Vargas et al., 2020), as well as traumatic events that happened after childhood. Third, by design of the parent trial, this study had an overrepresentation of individuals who experienced suicidality, which has potentially led to a higher prevalence and severity of childhood trauma and may reduce the generalizability of our findings. Indeed, roughly 80 % of participants had a history of childhood trauma, which is slightly higher than other reported prevalence rates in psychosis (Vila-Badia et al., 2021). This also limited the availability of a comparison group of participants without childhood trauma. Fourth, the PANSS may not have captured the full range of negative symptoms (e.g., social anhedonia). Future studies may benefit from using more comprehensive scales such as the Clinical Assessment Interview for Negative Symptoms (Kring et al., 2013). Fifth, the observational and cross-sectional nature of this study precludes any causal inferences. Finally, given that our data was largely collected during the COVID-19 pandemic, the restrictions in socialization may have reduced the generalizability of our findings. However, prior studies in this sample found little differences in EMA reports before or during the COVID-19 pandemic (Pinkham et al., 2020), such as time spent alone (Parrish et al., 2022).

5. Conclusions

The findings of this study suggest that childhood trauma may affect the perceived quality but not the quantity of real-world social relationships in psychosis. The association of childhood trauma with social demotivation, social avoidance, and negative appraisals of social interactions underlines the role of adverse childhood experiences in social functioning deficits of individuals with psychosis. While research on emotional trauma has generally received less attention than research on physical or sexual trauma, this study suggests that emotional abuse and neglect may have particularly harmful social consequences in the context of psychosis.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.schres.2022.12.039.