Explaining the Association Between Childhood Maltreatment and Psychological Adjustment in Youth in a Romantic Relationship

Experiencing cumulative forms of maltreatment during childhood, such as abuse and neglect, is associated with reduced intrapersonal and interpersonal functioning and a lower psychological adjustment in youth. However, little is known about the role of interpersonal and intrapersonal functioning of youth in a romantic relationship on the association between cumulative maltreatment and psychological adjustment. This study aims to document how cumulative forms of childhood maltreatment are associated with subsequent lower resilience or elevated symptoms of psychological distress through social cognition (mentalization, emotion regulation), romantic attachment style (anxious, avoidant), perceived social support, and relationship satisfaction. A sample of 1387 young adults in a romantic relationship aged 18 to 25 completed online questionnaires. A series of multiple linear regressions revealed that : (1) higher avoidant attachment and less perceived social support and emotion regulation explained why a higher number of forms of childhood maltreatment were associated with less resilience, while (2) higher attachment anxiety, less relationship satisfaction, perceived social support, and emotion regulation explained why a higher number of forms of childhood maltreatment were associated with higher psychological symptoms. Further investigations are needed to determine whether interventions to enhance resilience or interpersonal and intrapersonal functioning in young adults who experienced childhood maltreatment should focus on mentalization. They should, however, concentrate on romantic attachment, relationship satisfaction, the perception of social support, and emotion regulation abilities.

Keywords

Childhood maltreatment; resilience; psychological symptoms; psychological adjustment; social cognition; romantic attachment; social support; relationship satisfaction.

Conceptual Framework

More than one person out of four is the victim of emotional, physical, or sexual abuse or neglect before 18 in high-income countries such as Canada or the United States (Finkelhor, 2020; Stoltenborgh et al., 2015). Experiencing cumulative forms of maltreatment during childhood, such as abuse and neglect, is associated with reduced intrapersonal and interpersonal functioning and results in lower levels of resilience and more psychological symptoms in youth (Copeland et al., 2018; Daigneault, Hébert, & Tourigny, 2007; Le et al., 2018; Topitzes et al., 2013). Studies on psychological adjustment following childhood maltreatment have been focused primarily on intrapersonal functioning, such as one’s own cognitive and emotional impaired processes (Copeland et al., 2018; Henschel et al., 2019). However, some studies have recently focused on interpersonal functioning (Muller, Thornback, & Bedi, 2012; Sharma-Patel et al., 2014), including studies about youth romantic relationships (Dion, Gervais, Bigras, Blackburn, & Godbout, 2019; Simpson et al., 2011; Simpson et al., 2007). A few studies have focused on romantic attachment (i.e. processes by which we create an emotional bond with significant others) to determine whether perceived social support and intrapersonal functioning acted as intervening variables between childhood maltreatment and psychological adjustment, but only for older adults and not specifically in the context of romantic relationships (Hankin, 2005; Ortigo et al., 2013; Struck et al., 2020; Zhu et al., 2016). Youth are more at risk to report more conflicts and mood changes if they are in a romantic relationship. The common relational turbulence characterized by youth’s romantic relationships have short and long-term effects on resilience and psychological symptoms, particularly if youth were exposed to childhood maltreatment (Brisini & Solomon, 2019; Laursen, 1995; Simpson et al., 2011; Verhallen et al., 2021). Romantic partners, family, and friends may be important sources of support for youth (Campbell, Simpson, Boldry, & Kashy, 2005; Don and Hammond, 2017; Jensen et al., 2019). However, little is known about the role of interpersonal and intrapersonal functioning, such as social cognition, of youth in a romantic relationship on the association between cumulative maltreatment and psychological adjustment.

Childhood Maltreatment and Youth’s Subsequent Interpersonal and Intrapersonal Functioning

How does childhood maltreatment affect interpersonal functioning in youth? It has been suggested that the internalization of relational experiences within the family and with peers during childhood frames the capacity to have a differentiated self while creating intimate connections with other humans as adults (W. A. Collins, Hennighausen, Schmit, & Sroufe, 1997; Sroufe et al., 1999). Loss and traumatic experiences internalized during early childhood may induce deeply ingrained dysfunctional beliefs toward oneself and others that orient relational expectations and the choice of a romantic partner during adulthood (Bowlby, 1982; Collins & Read, 1994; Roisman et al., 2005). Repeated traumatic events during childhood, such as maltreatment and neglect, have been associated with insecure attachment patterns (i.e. inability to create a secure bond with the caregiver) that result in anxious or avoidant attachment styles in adulthood, which are characterized respectively by a fear of abandonment and discomfort with intimacy (Bartholomew, Cobb, & Poole, 1997; Hazan and Shaver, 1987; MacIntosh and Ménard, 2021). Hence, maltreated youth are more likely to display insecure attachment styles throughout adulthood than their non-maltreated counterparts.

The social support system is associated with the attachment system. It serves an adaptive function and is activated when the intimate partner shows signs of distress (Shaver, Hazan, & Bradshaw, 1988). Studies have shown that romantic relationship dyads in which one partner or the other presents an anxious or avoidant attachment style are more likely to offer inadequate conjugal support to one another and to interpret more negatively the support each other receives (Collins & Feeney, 2004; Mallinckrodt & Wei, 2005; McLeod et al., 2020). Youth maltreated during childhood also perceived less social support in their romantic relationships than those who weren’t (Fitzgerald & Gallus, 2020).

Adults’ attachment styles have commonly been studied in association with relationship satisfaction in university students, dating and newly married couples (Birnie-Porter & Hunt, 2015; Feeney, 2008; Kachadourian et al., 2004). A secure attachment style (i.e. the ability to form a secure relationship with a loved one) was found to be associated with more availability, intimacy, and support to the romantic partner, which have in turn been associated with higher relationship satisfaction (Collins & Read, 1990; Li & Chan, 2012; Mikulincer et al., 2002). Conversely, insecure attachment is correlated with more negative emotions, lower support, and relationship satisfaction (Candel & Turliuc, 2019; Crangle & Hart, 2017; Shaver et al., 2005; Simpson, 1990; Simpson et al., 2002). Emotional abuse during childhood has been associated with higher attachment anxiety and avoidance in the couple during young adulthood, which was associated with less relationship satisfaction (Riggs, Cusimano, & Benson, 2011).

Romantic relationships happen within partners’ complex social networks of friends, family, and relatives, and research has neglected the impact of youth’s in a romantic relationships social network (Sprecher, Felmlee, Orbuch, & Willetts, 2002). Study results have shown that peers and family play an important supportive role that fosters romantic relationships’ stability, quality, and satisfaction in youth (Jensen et al., 2019; Sprecher & Felmlee, 1992). However, youth are significantly more at risk to perceive less social support from their peers and family in young adulthood if they experienced childhood maltreatment (Kobak & Sceery, 1988; McLaughlin et al., 2020; Pepin & Banyard, 2006). Thus, youth who were exposed to childhood maltreatment are more at risk to report difficulties, less satisfaction, and dissolution within their romantic relationships than their non-maltreated peers (Agnew, Loving, & Drigotas, 2001; Jensen and Rauer, 2014).

Childhood maltreatment has also been shown to alter the development of necessary social-cognitive competencies, namely for the acquisition of high-level cognitive processes that allow for sensical interpretation of relational experiences in young adulthood, such as mentalization and emotion regulation (Choudhury, Blakemore, & Charman, 2006; Fiske and Taylor, 2008; Kay and Green, 2016; Kilford et al., 2016). Youth that were maltreated during childhood often exhibit lower levels of social cognitive skills, such as poor emotion regulation (Bradbury & Shaffer, 2012; Henschel et al., 2019), and poor ability to mentalize their inner mental states and those of others (i.e. the ability to perceive and interpret human behaviors in terms of intentional mental states such as feelings, wishes, goals or desires; Berthelot et al., 2019; Fonagy, Gergely, Jurist, & Target, 2002; van Schie et al., 2017), particularly when compared to their non-maltreated counterparts.

Youth’s Interpersonal and Intrapersonal Functioning Associated with Psychological Adjustment Following Childhood Maltreatment

Past studies have highlighted interactions between maltreatment and attachment, relationship satisfaction, social support, psychological functioning, and resilience, but results to date do not seem to meet consensus. For instance, attachment anxiety, but not attachment avoidance, has been associated with relationship dissatisfaction in youth in a romantic relationship, which has been associated with more depressive symptoms (Rivera, Cruz, & Muñoz, 2011; Scott and Cordova, 2002; Simpson and Rholes, 2008). Another study has found that attachment anxiety and avoidance were both associated with relationship dissatisfaction and perceived support and with more depressive symptoms (Chi Kuan Mak, Bond, Simpson, & Rholes, 2010). Yet another study found that attachment avoidance, but not attachment anxiety, was associated with less perceived social support, which was, in turn, associated with less resilience, but not with more depressive symptoms (Dodd, Driver, Warren, Riggs, & Clark, 2015). Others found that attachment avoidance, but not attachment anxiety, was associated with less social support, which was in turn associated with more depressive symptoms (Zhu et al., 2016). In contrast, some describe insecure attachment style as associated with less social cognition and more symptoms of post-traumatic stress (Ortigo et al., 2013). One possible explanation of this discrepancy between results could be that the attachment measures vary between studies. Another plausible explanation would be that attachment tend to vary according to gender, such that women are more likely to show more attachment anxiety, whereas men are more likely to show more attachment avoidance in general (Cioccia et al., 2020).

Some studies have shown that romantic attachment style, social support, relationship satisfaction, or cognitive-affective functioning explain the association between childhood maltreatment and psychological functioning in adulthood. Indeed, cumulative forms of childhood maltreatment have been associated with more attachment anxiety and avoidance (Dion et al., 2019; Muller et al., 2012), relationship dissatisfaction (Cao, Zhou, Leerkes, & Su, 2021; Fitzgerald, 2021; River et al., 2020), less perceived social support (Lagdon et al., 2021; Pepin & Banyard, 2006), and poor mentalization and emotion regulation abilities (Belvederi Murri et al., 2017; Berthelot et al., 2014; Burns et al., 2010), which were in turn, associated with more psychological symptoms.

Social cognition during childhood has partly explained the relationship between cumulative forms of childhood maltreatment and later resilience in young adulthood (Topitzes et al., 2013). Mentalization and emotion regulation appear to explain the link between cumulative maltreatment and psychological adjustment and resilience following a breakup in youth in a romantic relationship aged between eighteen and twenty-five years old (Francoeur et al., 2020). In this previous study, the more youth experienced different forms of maltreatment during childhood, the lower were their levels of social cognition and resilience following relationship dissolution and the more psychological symptoms they presented.

Despite our knowledge about these interactions between romantic attachment, social support, social cognition, psychological adjustment, and couples’ distress following childhood maltreatment, no study has investigated how these mechanisms could concurrently interact to explain the association between cumulative forms of maltreatment, resilience and psychological symptoms in youth in romantic relationships. It appears crucial to bridge this gap in the literature to determine which factors are associated with the most resilience and fewer symptoms to orient future interventions for youth who report childhood maltreatment in romantic relationships.

The Present Study

This study aims to document how cumulative forms of childhood maltreatment are associated with later resilience or symptoms of psychological distress while considering the intrapersonal and interpersonal functioning of youth in romantic relationships. We wish to determine if youth’s social cognition, romantic attachment style, perceived social support, and relationship satisfaction explain the link between childhood maltreatment and later psychological adjustment, assessed through resilience and psychological symptoms. Based on previous studies, it is expected that the more youth will report cumulative forms of childhood maltreatment, the more they will exhibit anxious and avoidant attachment styles, and the lower their levels of relationship satisfaction, perceived social support, and social cognition (emotion regulation, mentalization) will be. All these factors will, in turn, be associated with less resilience and more psychological symptoms.

Methodology

This study is part of a more extensive study on youth’s romantic relationships and adjustment following a romantic breakup using a longitudinal design approved by the University of Montreal and Centre intégré universitaire de santé et services sociaux de l’Est-de-L’Île-de-Montréal ethical committees. The current study uses cross-sectional data collected at baseline when all participants reported being in a romantic relationship.

Procedures and Participants

Potential participants were recruited through university social media (Facebook, Instagram, Twitter). The recruitment ad presented the following information: participants had to consider themselves in a romantic relationship and be aged between 18 and 25 years. The ad also indicated the length of the anticipated engagement in the study (six to eighteen months), the online nature of the participation, and the financial compensation offered (participation in a 150$ lottery for an Amazon Gift card). To be eligible, the participants had to have access to a smartphone, a computer, or a tablet to complete the questionnaires on Qualtrics, an online secured survey platform, and be fluent in French. Those who accepted to participate completed all survey questionnaires, which lasted approximately ninety minutes. Eligible participants who completed the consent form and the study questionnaires form the final convenience sample of 2120 youth aged 18 to 25 years old in a romantic relationship.

Measures

All measures that were used were translated from English to French. The Early Trauma Inventory short-form of 7 items was used to assess seven forms of childhood maltreatment, including physical, emotional, sexual abuse, and witnessed physical or psychological violence (ETI-SF; Bremner et al., 2000; Cyr et al., 2007). Participants indicated whether they experienced each form of maltreatment, Yes (1) or No (0), and the total score ranges between 0 and 7. The original English questionnaire was validated among maltreated and non-maltreated adult samples and among clinical and non-clinical adult samples (Bremner et al., 2000). The means, standard deviation scores, and alphas for each measure in the current study are presented in Table 1.

Table 1.

Descriptive statistics for all study variables (mean, standard deviation, range, alpha) and correlations among variables

	M	SD	Range	Alpha	1	2	3	4	5	6	7	8
1. Cumulative maltreatment	1.51	1.71	0.00–7.00	0.72
2. Attachment anxiety	4.29	1.39	1.00–7.00	0.87	0.18*
3. Attachment avoidance	2.05	0.89	1.00–7.00	0.82	0.12*	0.11*
4. Relationship satisfaction	16.85	2.63	5.00–21.00	0.77	− 0.15*	− 0.14*	− 0.52*
5. Social support	3.33	0.38	1.63-4.00	0.89	− 0.20*	− 0.27*	− 0.39*	0.37*
6. Emotion regulation	3.28	0.47	1.75–4.75	0.85	− 0.12*	− 0.32*	− 0.15*	0.14*	0.31*
7. Mentalization	3.74	0.40	2.43–4.86	0.83	0.01	− 0.11*	− 0.25*	0.13*	0.33*	0.27*
8. Resilience	4.00	0.54	2.08-5.00	0.82	− 0.36*	− 0.27*	− 0.33*	0.31*	0.61*	0.36*	0.32**
9. Psychological Symptoms	1.65	0.58	1.00-4.11	0.90	0.28*	0.39*	0.13*	− 0.20*	− 0.28*	− 0.30*	− 0.12*	− 0.34*

*p < .001.

The abridged 12-item Experiences in Close Relationship scale was used to assess anxious (6 items) and avoidant (6 items) attachment (ECRS; Lafontaine et al., 2016). The Likert-type scale varies between 1 (disagree strongly) and 7 (agree strongly), and the averaged total score ranges between 1 and 7 for each dimension. The short version was validated in French and English among students and adults from the general and clinical populations (Lafontaine et al., 2016).

The abridged 4-item French version of the Dyadic Adjustment Scale (DAS; Sabourin et al., 2005; Spanier, 1976) measured relationship satisfaction. The Likert-type scale for the three first items that assess various aspects of the relationship ranges between 1 (always) and 6 (never). The Likert-type scale for the fourth item, which measures the degree of happiness in the relationship, varies between 1 (extremely unhappy) and 7 (perfectly happy). The total score ranges between 0 and 21. The scale was developed and validated for married and non-married couples (Sabourin et al., 2005; Spanier, 1976).

The French version of the Social Provision Scale of 24 items was used as a measure of perceived social support in six domains (attachment, social integration, reassurance of worth, reliable alliance, guidance, opportunity for nurturance; SPS; Caron, 1996; Cutrona and Russell, 1983). The Likert-type scale varies between 1 (strongly disagree) and 4 (strongly agree), and the averaged total score used ranges between 1 and 4. Both French and English versions were validated among youth and clinical samples (Caron, 1996; Cutrona & Russell, 1983).

The Cognitive Emotion Regulation Questionnaire of 36 items was used to assess social-emotional regulation (CERQ; d’Acremont and Van der Linden, 2007; Garnefski and Kraaij, 2007). The CERQ measures five appropriate (acceptance, positive refocusing, refocus on planning, positive reappraisal, putting into perspective) and four inappropriate (self-blame, rumination, catastrophizing, blaming others) cognitive emotion regulation strategies in response to negative events. The Likert-type scale varies between 1 (almost never) and 5 (almost always), and the averaged total score used ranges between 1 and 5. The French version was validated among adolescents and presented adequate convergent validity (d’Acremont & Van der Linden, 2007). The English version was validated among adults from the general population and showed adequate convergent validity (Garnefski & Kraaij, 2007).

The Mentalization Scale of 28 items assesses three domains of mentalization, namely self-related mentalization, others-related mentalization, and motivation to mentalize (Dimitrijević, Hanak, Altaras Dimitrijević, & Jolić Marjanović, 2018; Francoeur et al., 2020). The Likert-type scale varies between 1 (completely inexact) and 5 (completely exact), and the averaged total score used ranges between 1 and 5. The questionnaire was validated among youth and clinical samples (Dimitrijević et al., 2018).

The Child and Youth Resilience Questionnaire-short form of 12 items assesses resilience according to three dimensions (individual/social, family, community/spiritual) and was developed and validated for use with 12 to 24 years old (Liebenberg, Ungar, & LeBlanc, 2013). The Likert-type scale varies between 1 (not at all) and 5 (very much), and the averaged total score used ranges between 1 and 5. The extended French version was validated among adolescents with a history of childhood maltreatment (Daigneault, Dion, Hébert, McDuff, & Collin-Vézina, 2013).

The Brief symptoms inventory of 18 items was used to assess three psychological symptoms (Somatization, Depression, Anxiety; BSI; Derogatis, 2000; Gosselin & Bergeron, 1993). The Likert-type scale varies between 1 (not at all) and 5 (extremely), and the averaged total score ranges between 1 and 5. The English questionnaire was validated among youth and clinical samples (Derogatis & Spencer, 1993). The French version was validated among adults and adolescents from the general population, and clinical samples (Gosselin & Bergeron, 1993).

Sample Size and Power Calculation

An a priori statistical power analysis was computed to determine the number of participants required for the study considering the percentage of romantic breakup and the attrition rate over 18 months. A final sample of 200 participants who experienced a romantic breakup was necessary to detect a medium effect with a statistical power of 0,80 using the piecewise linear trajectory model (Diallo & Morin, 2015). Accounting for attrition and breakup rates over time, an initial sample size of 2000 participants in a romantic relationship was thus targeted to observe small effect sizes if they existed, enough to detect similar effect sizes in the current study using the whole initial sample of youth in a romantic relationship.

Data Analyses

The principal analyses are a series of multiple linear regressions using PROCESS for IBM SPSS 27, model 4 (Hayes, 2018). These analyses allow studying each intermediate variable’s contribution simultaneously in the overall models. The potential intervening variables were entered simultaneously in two analyses with each outcome (Y). Their respective contributions to explaining the relationship between childhood maltreatment and each outcome were compared. The analyses for indirect effects (paths ab) were based on 10 000 resampling biased-corrected bootstrap confidence intervals, using a 95% confidence interval. Figure 1 illustrates the hypothesized interactions among intermediate variables in the model.

Fig. 1.

Statistical diagram of the intermediate variables’ analyses adapted from model 4 in process templates (Hayes, 2018). Note: indirect effect of Childhood Maltreatment on Resilience and Psychological Symptoms through Romantic Attachment Style, Component of Social Cognition and Social Support (path ab); direct effect of Childhood Maltreatment on Resilience and Psychological Symptoms when considering Romantic Attachment Style, Component of Social Cognition and Social Support (path c′); total effect of Childhood Maltreatment on Resilience and Psychological Symptoms (not shown: path c = c′ + ab)

Results

Descriptive Statistics and Correlation Coefficients

The final sample consists of 1387 participants who consented to participate and completed all the questionnaires. Participants were on average 21.72 years (SD = 2.09), primarily women (85%), French speakers (95%), Canadians (91%), and university students (66%). A majority (61%) reported experiencing at least one of the seven documented forms of childhood maltreatment. Correlations among study variables are reported in Table 1 and reveal that they were all below 0,75, allowing for the inclusion of all potential intermediate variables in our subsequent analysis (Tabachnick & Fidell, 2013). Furthermore, no multicollinearity problems were detected in the two regressions presented below (Variance inflation factor < 5).

Intermediate Variables’ Analyses

All associations between childhood maltreatment and intermediate variables were significant except one (paths a_1 − 6 from Fig. 1). According to Cohen’s conventions (1988), cumulative childhood maltreatment had a small positive association with attachment anxiety (ß = 0.15, p < .001, 95% CI [0.11; 0.19], R² = 0.03), and avoidance (ß = 0.06, p < .0001, CI95% [0.03; 0.09], R² = 0.01), a small negative association with relationship satisfaction (ß = -0.23,, p < .001, 95% CI [-0.31; -0.15], R² = 0.02), social support (ß = -0.05, p < .001, 95% CI [-0.06; -0.03], R² = 0.04), and emotion regulation (ß = -0.03, p < .001, 95% CI [-0.05; -0.02], R² = 0.01). The association with mentalization was non-significant (ß = 0.00, p = .656, 95% CI [-0.01; 0.02]).

As shown in Table 2, considering romantic attachment, social support, and social cognition, cumulative childhood maltreatment was negatively associated with resilience and positively associated with psychological symptoms (paths c’). Hence, exposure to more childhood maltreatment was directly associated with less resilience and more symptoms, indicating more psychological adjustment difficulties overall. The direct effects of the intervening variables on resilience and symptoms are also shown in Table 2 (paths b_1 − 6), and reveal that the variables in the models accounted for 48% of the variance of resilience (F_{(7, 1379)} = 184.42, p < .0001, R² = 0.48), and 24% of the variance of psychological symptoms (F_(7,1379) = 63.85, p < .0001, R² = 0.24).

Table 2.

Direct effects of maltreatment and all intermediate variables on resilience and psychological symptoms (path c’ and b1 − 6)

	Resilience							Psychological Symptoms
Variables (paths)	ß	SE	p	95%LLCI	ULCI	ß	SE		p	95%LLCI	ULCI
Constant	0.92	0.18	< 0.000	0.57	1.27	2.69	0.23		< 0.000	2.24	3.15
Maltreatment (c’)	-0.07	0.01	< 0.000	-0.09	-0.06	0.06	0.01		< 0.000	0.05	0.08
Anxiety (b1)	-0.01	0.01	0.163	-0.03	0.00	0.11	0.01		< 0.000	0.09	0.13
Avoidance (b2)	-0.04	0.01	0.006	-0.07	-0.01	-0.02	0.02		0.279	-0.06	0.02
Relationship Satisfaction (b3)	0.01	0.00	0.129	-0.00	0.02	-0.02	0.01		0.003	-0.03	-0.01
Social Support (b4)	0.60	0.03	< 0.000	0.54	0.67	-0.15	0.04		0.001	-0.23	-0.06
Emotion regulation (b5)	0.18	0.02	< 0.000	0.13	0.23	-0.19	0.03		< 0.000	-0.25	-0.13
Mentalization (b6)	0.16	0.03	< 0.000	0.11	0.22	-0.04	0.04		0.276	-0.11	0.03

An indirect effect is significant when the bootstrap confidence interval excludes 0 (corrected for biased resampling error). As shown in Table 3, in both models, the bootstrap confidence interval excludes 0 for the total indirect effect of maltreatment on resilience and psychological symptoms through all intervening variables combined (path c). This indicates that at least one of the six intermediate variables explained the association between maltreatment, and resilience and psychological symptoms. Specific indirect effects through intervening variables on resilience and symptoms are presented in Table 3 (paths ab_1 − 6). The association between cumulative childhood maltreatment and resilience was partly explained by three of the six intermediate variables: attachment avoidance, social support, and emotion regulation. This indicates that the more participating youth reported being exposed to different forms of childhood maltreatment, the more attachment avoidance they exhibited and the less social support and emotion regulation they reported, which explained why they reported less resilience. The association between childhood maltreatment and symptoms was partly explained by four of the six intervening variables: attachment anxiety, relationship satisfaction, social support, and emotion regulation. This indicates that the more youth reported different forms of childhood maltreatment, the more attachment anxiety they exhibited and the less relationship satisfaction, social support and emotion regulation they reported, which explained why they reported more psychological symptoms.

Table 3.

Total maltreatment effect and indirect effects through intermediate variables on resilience and psychological symptoms (n = 1387)

	Resilience					Psychological Symptoms
Variables (paths)	ß	SE	LLCI	ULCI	ES	ß	SE	LLCI	ULCI	ES
Total (c)	-0.04	0.01	-0.05	-0.03		0.03	0.00	0.02	0.04
Anxiety (ab1)	-0.00	0.00	-0.00	0.00	0.04	0.02	0.00	0.01	0.02	0.51
Avoidance (ab2)	-0.00	0.00	-0.00	-0.00	0.06	-0.00	0.00	-0.00	0.00	0.04
Relationship Satisfaction (ab3)	-0.00	0.00	-0.00	0.00	0.04	0.00	0.00	0.00	0.01	0.13
Social Support (ab4)	-0.03	0.00	-0.04	-0.02	0.72	0.01	0.00	0.00	0.01	0.21
Emotion Regulation (ab5)	-0.01	0.00	-0.01	-0.00	0.15	0.01	0.00	0.00	0.01	0.19
Mentalization (ab6)	0.00	0.00	-0.00	0.00	0.00	-0.00	0.00	-0.00	0.00	0.00
Note. Total and indirect effects and 95% confidence interval based on 10,000 bootstrap samples. ES = Effect size based on the ratio of the intermediate variables’ effect coefficient on the total effect coefficient.

Discussion

The goal of the current study was to document how cumulative forms of childhood maltreatment were associated with later resilience and symptoms when considering intrapersonal and interpersonal functioning for youth in romantic relationships. This is the first paper to address the romantic attachment style, relationship satisfaction, perceived social support, and cognitive vulnerabilities as concurrent intermediate variables in the association between childhood maltreatment, resilience and psychological symptoms in youth in a romantic relationship. Our hypotheses were partially confirmed as most but not all intervening variables explained the relationship between maltreatment and resilience and symptoms. Two of the six hypothesized intervening variables significantly explained the link between childhood maltreatment, resilience and symptoms: social support and emotion regulation. The association between child maltreatment and resilience was significantly explained by attachment avoidance, social support, and emotion regulation, while attachment anxiety, relationship satisfaction, social support, and emotion regulation explained cumulative maltreatments’ relationship with psychological symptoms. Attachment anxiety and avoidance and relationship satisfaction were intermediate variables in only one regression model each. Mentalization was the only hypothesized intervening variable that never significantly explained part of the cumulative maltreatment association with resilience and psychological symptoms.

Childhood Maltreatment

Childhood maltreatment was still significantly associated with resilience and symptoms when all intermediate variables were included in the models. These results are consistent with previous findings indicating that childhood maltreatment was associated with psychological adjustment even when considering attachment insecurity, relationship satisfaction, perceived social support, emotion regulation, and mentalization as individual predictors (Belvederi Murri et al., 2017; Berthelot et al., 2019; Burns et al., 2010; Cao et al., 2021; Dion et al., 2019; Fitzgerald, 2021; Lagdon et al., 2021; Liu et al., 2020; Muller et al., 2012; Pepin & Banyard, 2006; River et al., 2020). Our results add to the literature by considering all these intervening variables simultaneously. Even though our model was more comprehensive than prior ones, the fact that maltreatment remains associated with both outcomes indicates that other mechanisms need to be considered and included in future models.

Social Support

Social support accounted for the highest indirect effect size explaining 72% of the association between childhood maltreatment and resilience while simultaneously considering all other intermediate variables. The more youth experienced cumulative forms of childhood maltreatment, the less they perceived social support and exhibited resilience, regardless of their emotional regulation, attachment avoidance/anxiety, relationship satisfaction, or mentalization. These results are consistent with previous literature on social support evolving as a potential intervening variable of resilience in young adults who experienced childhood maltreatment (Pepin & Banyard, 2006). Hence, social support from romantic partners and family and friends seem to play an important role, above other mechanisms, in buffering the impact of childhood maltreatment and promoting resilience in young adults. As well, as was expected according to the vast literature on the field, social support acted as a significant intervening variable between childhood maltreatment and psychological symptoms in our sample (Struck et al., 2020; Zerach & Elklit, 2020).

Emotion Regulation

The more youth reported childhood maltreatment, the poorer were their emotion regulation skills, and worse were their symptoms and ability for resilience. Our results are in line with others who suggest that emotion regulation predicts resilience in adolescents (Mestre, Núñez-Lozano, Gómez-Molinero, Zayas, & Guil, 2017). This study goes one step further, as a paucity of studies has assessed the association between childhood maltreatment, emotion regulation, and resilience in young adults (Francoeur et al., 2020). A previous study from our team also found that the more youth reported childhood maltreatment, the less they presented emotion regulation abilities and, in turn, the more they showed psychological symptoms (Francoeur et al., 2020). This study adds to the literature linking childhood maltreatment and emotion regulation difficulties in children, adolescents, and adults (Gruhn & Compas, 2020; Jennissen et al., 2016; Kim & Cicchetti, 2010).

Attachment

As previous results indicated, the more youth reported numerous forms of childhood maltreatment, the more they presented attachment avoidance and attachment anxiety, but only attachment avoidance explained maltreatment’s negative association with resilience (Bradley & Hojjat, 2017; Dodd et al., 2015). Some authors suggest that individuals with an anxious romantic attachment style would exhibit greater distress but rebound faster than individuals with an avoidant romantic attachment style (Marshall, Bejanyan, & Ferenczi, 2013). The inhibited distress in relationships in individuals with an avoidant attachment appears to make them at higher risk of reporting less resilience than their peers showing more attachment anxiety.

While considering all other intermediate variables, childhood maltreatment predicted symptoms through attachment anxiety but not avoidance. The more people experienced childhood maltreatment, the more they presented attachment anxiety, and in turn, the more they presented psychological symptoms, which is consistent with previous findings (Chi Kuan Mak et al., 2010). The anxious romantic attachment style accounted for the highest indirect effect size explaining 51% of the association between childhood maltreatment and symptoms. This was expected according to the literature on the topic (Mallinckrodt & Wei, 2005; Ng & Hou, 2017; Rholes & Simpson, 2004; Zheng et al., 2020). Indeed, prior studies have shown that attachment anxiety in youth in a romantic relationship was associated with more conflict behaviors, which were associated with dissolution and more symptoms (Brassard, Lussier, & Shaver, 2009; Marchand-Reilly, 2012; Mikulincer and Shaver, 2016). More studies are warranted to explain why attachment avoidance did not explain the relationship between childhood maltreatment and symptoms as found in other studies (Dion et al., 2019; Muller et al., 2012). It may be that it shares a common variance with the other intervening variables in the model, which were more significantly associated with psychological symptoms than avoidant attachment was.

Relationship Satisfaction

As expected, childhood maltreatment was associated with symptoms through relationship satisfaction. Hence, the more individuals reported childhood maltreatment, the less they reported being satisfied in their relationship, and the more they presented with psychological symptoms. This finding is congruent with a previous study that revealed a similar intervening association while studying postpartum depressive symptoms in mothers who had experienced childhood maltreatment (Cao et al., 2021). However, poorer relationship satisfaction did not explain the association between childhood maltreatment and low resilience in young adults in romantic relationships. To our knowledge, this is the first study to investigate relationship satisfaction as an intermediate variable of the association between childhood maltreatment and resilience.

Mentalization

Mentalization only directly predicted the level of resilience, as expected (Bak, Midgley, Zhu, Wistoft, & Obel, 2015). However, mentalization did not explain the association between childhood maltreatment and resilience. In a previous study (Francoeur et al., 2020), childhood maltreatment was only related to breakup adjustment through self-related mentalization. In the present study, we did not consider the different forms of mentalization given the number of intervening variables included in our model. The association between childhood maltreatment and resilience through mentalization thus needs further investigation. Studies should assess the different forms of mentalization separately: self and other related mentalization and the motivation to mentalize (Dimitrijević et al., 2018). As well, childhood maltreatment did not predict psychological symptoms through mentalization, which is surprising according to the literature in the field (Adler, Gervinskaitė-Paulaitienė, Čekuolienė, & Barkauskienė, 2021; Belvederi Murri et al., 2017; Berthelot et al., 2019; Taubner et al., 2016). This failure to explain maltreatment’s association with symptoms through mentalization may be due to our inclusion of other concurrent intervening variables. Some authors suggest that the link between childhood abuse and clinical symptoms in non-affective psychotic disorders via mentalization is observed only for severe childhood abuse (Weijers et al., 2018).

Clinical Implications

Our study’s findings have several implications for clinical practice. First, they reinforce prior knowledge that childhood maltreatment has several social, thus economic, short and long-term consequences. Results revealing an association between cumulative childhood maltreatment and psychological adjustment indicate that interventions aimed at reducing symptoms and enhancing resilience should address relational and personal difficulties to maximize their effectiveness. This could be done in clinical practice by considering personal childhood traumas as well as the relational challenges in the current romantic relationship that lead to relationship dissatisfaction and distress. Our results also indicate that facilitating emotion regulation strategies/skills, developing a larger social skills repertoire and developing youth’s social support network should be effective therapeutic goals to prevent psychological maladjustment.

Limitations

A cross-sectional design and retrospective measure of childhood maltreatment were used, entailing potential recall bias (Colman et al., 2016; Widom & Morris, 1997). Only self-report questionnaires were used, which entails common measure variance error (Hardt & Rutter, 2004; Reuben et al., 2016). The scale for mentalization was not validated in a French population, although in our study the measure showed good internal consistency (Dimitrijević et al., 2018). The characteristics of the romantic relationship, including its length and nature (e.g. cohabiting or not), were not considered, and we did not distinguish the type of childhood maltreatment, only their cumulative effect. The participants were a homogeneous group attending university and results cannot be generalized to the population of 18 to 25-year-old youths. There is also an overlap in the measure of resilience which contains items assessing social support in its construct. This could partly explain the strong association between childhood maltreatment and resilience via social support. Nevertheless, social support is a well-known strong predictor of psychological adjustment following childhood maltreatment (Kim-Cohen et al., 2006; Struck et al., 2020). It should also be noted that the study’s recruitment was during the Covid 19 pandemic and the first confinement, which may have affected participants’ distress. Further study should consider these additional factors, at least as control variables.

Conclusions

Social support and emotion regulation were the most constant and robust mediators of the association between cumulative childhood maltreatment and resilience or psychological symptoms in young adults aged 18 to 25 years in a romantic relationship when all intervening variables were considered. Romantic attachment and relationship satisfaction were less consistent in explaining the link between cumulative childhood maltreatment, resilience, and symptoms of distress. As measured in our study, mentalization did not significantly explain cumulative maltreatment’s association with either resilience or psychological symptoms when all other intervening variables were considered. Social support accounted for the highest proportion of resilience explained by cumulative childhood maltreatment, although this could be due to their similar constructs. Attachment anxiety accounted for the highest proportion of symptoms explained by cumulative childhood maltreatment.

Recommendations

Further research should study resilience as a process, instead of a consequence, to see whether it explains some of the association between childhood maltreatment and symptoms (Collazzoni et al., 2020; Masten et al., 1990; Vieira et al., 2020). Further investigations are needed to determine whether interventions to enhance resilience, interpersonal and intrapersonal functioning in young adults who experienced childhood maltreatment should focus on mentalization given discrepant results with previous findings. They should, however, concentrate on romantic attachment, relationship satisfaction, the perception of social support, and emotion regulation abilities.

Authors’ contribution statements

All authors made substantial contribution on this manuscript. The first author was responsible for writing the final draft of the manuscript. The second author was responsible for supervising the first author and made numerous editing of the manuscript. The other authors were responsible for reviewing the manuscript and approved the final draft before the submission process.

Funding

This research was funded by the Social Sciences and Humanities Research Council (No. 435-2019-0279).

Declarations

Competing interest

All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.

Ethics approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The study was approved by the University of Montreal (No. CEREP-20-086-R) and Centre intégré universitaire de santé et services sociaux de l’Est-de-L’Île-de-Montréal (No. 2020 − 1963) ethical committees.

Consent to participate

Informed Consent was obtained from all individual participants included in the study.