Intergenerational Transmission: Observed Negative Communication Mediates Dyadic Associations Between Childhood Maltreatment and Marital Quality

ABSTRACT

Childhood maltreatment (CM), a complex trauma, has long been recognized as a detrimental predictor of marital quality in adulthood. Although research has almost exclusively focused on the associations of CM with a range of relationship outcomes from an individual perspective, the dyadic effects of CM on couple interactions and marital quality remain understudied. Relying on the Couple Adaptation to Traumatic Stress model, the present study examined whether CM of each partner is associated with their own and their partner's relationship quality through observed negative couple interaction. The study included 115 mixed‐gender Israeli married couples, recruited using a convenience sampling method. During a laboratory session, couples completed self‐report questionnaires assessing CM and relationship quality and engaged in a 10‐min couple conflict discussion. Actor‐partner Interdependence Models revealed an indirect effect from CM of women on the marital quality of women and men. Specifically, women's CM was associated with more expressions of negative communication by women, which in turn was associated with lower marital quality reported by men and women. The study points to systemic intergenerational effects of CM in women on adult relationships through negative couple communication, underscoring the need to assess CM history in couple therapy and to focus on improving communication patterns to prevent the intergenerational transmission of relational dysfunction.

1. Introduction

Childhood maltreatment (CM) is a prevalent public health issue, affecting millions globally (Stoltenborgh et al. 2015). It includes physical, sexual, emotional, and/or neglecting behaviors committed toward children in a relationship of trust and power, and can severely impact the child's well‐being and development (World Health Organization 2016). Moreover, a large body of evidence demonstrates its lasting effects across life including the ability to establish healthy intimate relationships (for a review, see Li et al. 2019; Vaillancourt‐Morel et al. 2024; Zamir 2022). Intimate relationships are a major source of happiness and a significant developmental milestone in adulthood (Karney and Bradbury 2020), and are implicated in mental and physical well‐being (Choi and Marks 2008; Robles et al. 2014). Adults with a history of CM are at risk for a range of relationship disruptions including a greater likelihood of divorce (Colman and Widom 2004), ineffective communication, conflicts, intimate partner violence (IPV), and lower levels of marital satisfaction (for a review, see Li et al. 2019; Zamir 2022).

CM may not only spill over across generations, perpetuating poor relationships with loved ones, but it can also cross over to affect the quality of one's partner's relationship quality (for a review, see Vaillancourt‐Morel et al. 2024; Zamir 2022). An intimate relationship is a system in which reciprocal processes affect the two dyad members (Kelley and Thibaut 1978; Kenny et al. 2006). Whereas theory and research mostly focused on the effects of CM on marital functioning (i.e., actor effect), not much is known about the mechanisms by which CM affects a partner's marital functioning (i.e., partner effect). Such knowledge may help inform prevention interventions designed for couples with a history of CM. Thus, this study examined whether negative couple communication mediated the associations between CM and one's own and one's partner's marital quality.

1.1. A Theoretical Model: The Couple Adaptation to Traumatic Stress

CM is a form of complex trauma, chronic stressors, including physical or psychological threats, mostly occurring within the caregiving system of children (Courtois and Ford 2012). One of the major lasting implications of CM is trauma‐related distress including depression, anxiety, and post‐traumatic stress (Charak et al. 2020; Cloitre et al. 2009). The Couple Adaptation to Traumatic Stress (CATS; Goff and Smith 2005) posits that trauma‐related distress experienced by a partner with a trauma history can be transmitted to their spouse, a process termed secondary traumatization. This process is bi‐directional, such that trauma‐related distress of each dyad member can amplify the other's symptoms. The physiological arousal associated with trauma‐related distress can hamper effective couple communication, triggering hostility, aggression, withdrawal, and other negative behaviors. Such disruptions may hinder the security needs of each partner and further reduce their perceived marital quality (Goff and Smith 2005).

CM is also theorized to alter the perception and interpretation of the world, influencing how individuals perceive and engage with their partners. Survivors of CM frequently struggle with persistent negative attributions toward themselves (e.g., guilt, self‐blame) and others. For instance, they may interpret neutral or ambiguous partner behavior as critical or rejecting. They also tend to perceive a threat in relationships (Goff and Smith 2005). These maladaptive cognitive patterns may lead to aggressive behaviors or social withdrawal in intimate interactions (Dodge 2006; van Reemst et al. 2016). Negative communication may, in turn, trigger negative or aggressive responses from partners, potentially perpetuating a destructive cycle of conflict within the relationship (Gottman and Notarius 2000).

Consistent with the CATS model, research has indicated a direct link between CM and lower relationship quality (DiLillo et al. 2009; Labella et al. 2017; Lassri et al. 2023; Vaillancourt‐Morel et al. 2024; Zamir and Lavee 2016), defined as the way romantic partners evaluate the goodness of the characteristics of their relationship (Glenn 1990). Emerging evidence indicated associations of CM with one's own and one's partner's lower relationship quality (Maneta et al. 2015; Riggs et al. 2011; Vaillancourt‐Morel et al. 2024). However, questions remain on whether negative communication underlies these dyadic associations.

1.2. Couple Communication

Couple communication refers to the behaviors couples express while interacting (Kelly et al. 2003). Research has shown that partners with a history of CM are more likely to report on marital conflicts (Belt and Abidin 1996; Kim et al. 2009), during which they tend to utilize a more volatile problem‐solving style (Knapp et al. 2017) and express more domineering, distant, and controlling behaviors (for a review see Zamir 2022). Also, their interactions tend to be less positive, warm, and intimate (Mullen et al. 1994; Whisman 2014). A few studies indicated that CM is associated with one's partner's lower communication quality (Busby et al. 2011; Knapp et al. 2017; Whisman 2014). Having constructive communication may benefit the two dyad members. Research evidence indicated that the communication behaviors of each partner are associated with their own and their partner's relationship quality (Segrin et al. 2009). Considering the CATS model (Goff and Smith 2005) that negative communication stimulated by trauma‐related symptoms may hinder relationship quality, this study examined whether observed negative couple communication, expressed during marital interactions, mediates the actor (i.e., one's own) and partner (i.e., one's partner's) associations of CM with marital quality.

Research linking CM and communication has focused mostly on global evaluations of communication quality or intimacy (e.g., Davis et al. 2001; DiLilio and Long 1999). Importantly, extensive research spanning over the past three decades has identified specific interactive behaviors including contempt, criticism, withdrawal, defensiveness, and negative affect (e.g., anger) as predictors of marital happiness and stability (for reviews see Gottman and Notarius 2000; Heyman 2001; Kelly et al. 2003). Because research on CM and couple communication almost entirely employed self‐report measures (for a review see Zamir 2022) which are sensitive to recall bias and common method variance (Podsakoff et al. 2003), we employed observations of couple communication assessing communication behaviors predictive of marital quality (for reviews see Gottman and Notarius 2000; Heyman 2001; Kelly et al. 2003). In addition, given that CM is a significant risk factor for IPV perpetration (for a review see Vaillancourt‐Morel et al. 2024; Li et al. 2019), we assessed psychologically abusive behaviors and verbal and non‐verbal communication intent at inflicting mental or emotional pain or imposing control over a partner (Breiding et al. 2015).

1.3. The Current Study

Over the past few decades, research has consistently demonstrated the detrimental and direct links between CM and marital functioning (Vaillancourt‐Morel et al. 2024; Zamir 2022). Building on this foundation, the present study aims to extend the literature by exploring an interpersonal mechanism linking each partner's CM history to their own and their partner's marital quality. Although previous work has largely adopted an individual‐focused approach (Zamir 2022), this study employs a dyadic framework. Specifically, the CATS model posits that negative couple communication serves as a key pathway through which CM impacts the marital quality of both partners (Goff and Smith 2005). Prior research has supported dyadic associations between CM and self‐reported couple communication (e.g., Busby et al. 2011; Knapp et al. 2017; Zamir 2022; Zamir et al. 2018), as well as between couple communication and overall relationship quality (Segrin et al. 2009). To advance this line of research, the current study adopts a nuanced and ecologically valid approach by using real‐time observational methods to assess couple communication patterns. Hence, we tested whether observed negative couple communication mediated the actor and partner effects of CM on marital quality. We hypothesized that (a) exposure to CM in each partner would be associated with their own and their partner's greater negative communication behaviors. In turn, (b) greater negative communication of each partner would be associated with their own and their partner's lower marital quality.

2. Method

2.1. Participants

A sample of 115 mixed‐gender first‐time married couples (same‐sex marriage is not sanctioned in Israel) was recruited through convenience sampling. The mean age for men was 32.02 (SD = 6.95), and for women, it was 30.82 (SD = 4.23). On average, the couples had been married for 5.57 years (SD = 1.65), with an average of 1.33 children per family (SD = 1.65). Most men (81.7%) and women (83.5%) were born in Israel, and all identified as Jewish and native Hebrew speakers. Additionally, most participants had earned at least a bachelor's degree (72% of women and 53.9% of men). Family income varied, with 14.8% of couples reporting a monthly gross income of fewer than 1400 USD, 27% reporting incomes between ~1400 and 2700 USD, 21.7% between ~2700 and 4000 USD, 13% between ~4000 and 6700 USD, 8.7% between ~6700 and 8000 USD, and 5.2% between ~8000 and 11,000 USD. For comparison, the average monthly gross income for a family in Israel in 2023 was ~5500 USD (Israel Central Bureau of Statistics 2023).

2.2. Procedure

Participants were recruited via social media advertisements directing them to a brief electronic screening survey to assess their eligibility (i.e., married for the first time, in a mixed‐sex marriage, and native Hebrew speakers). Those meeting the criteria were invited to provide their contact details and subsequently received a phone call from a research assistant for a brief intake interview, during which they were briefed about the study. Interested participants were then scheduled for a visit to the Family Resilience Laboratory at the Hebrew University of Jerusalem. Upon arrival, couples provided informed consent and proceeded to complete separate electronic self‐report questionnaires, followed by engagement in a problem‐solving discussion task. Couples received compensation of ~55 USD. The procedures and instruments used in this study were approved by the Ethical Review Board of the School of Social Work and Social Welfare, Hebrew University of Jerusalem. Data collection took place from August 2021 to July 2022.

2.3. Measures

2.3.1. Childhood Maltreatment

The Childhood Trauma Questionnaire Short‐Form (CTQ‐SF; Bernstein et al. 2003) is a 28‐item retrospective inventory measuring childhood physical, sexual, and emotional abuse, as well as physical and emotional neglect. It comprises five items for each type of CM and three items assessing underreported CM (e.g., “People in my family said hurtful or insulting things to me”) Each partner rated the frequency of their CM experiences up to age 18 on a 5‐point Likert‐type scale, ranging from 1 (never true) to 5 (very often true). Scores are summed for each subscale. CM cut‐off scores were obtained using the guidelines provided by CTQ developers (Bernstein and Fink 1998): a score of 5 or higher indicates sexual abuse, 8 or higher for physical abuse, 9 or higher for emotional abuse, 9 or higher for emotional neglect, and 8 or higher for physical neglect. Participants endorsing at least one form of CM were classified as exposed to CM (coded 1), whereas those who did not were classified as not exposed (coded 0).

The CTQ has demonstrated adequate internal consistency, validity, and reliability (Bernstein et al. 2003). In the current study, Cronbach's alpha coefficients indicated excellent internal consistency for both men (α = 0.91) and women (α = 0.90), except for physical neglect in women, which exhibited poor reliability (α = 0.53) and was consequently excluded from the final score for women. A paired‐samples t test comparing CM sum scores between men (M = 28.81, SD = 9.85) and women (M = 28.90, SD = 8.97) did not indicate a significant difference, t(114) = 0.08, p = 0.94. In addition, a chi‐square test assessing the presence versus absence of CM exposure between men and women revealed no significant differences, χ ²(1, N = 116) = 0.28, p = 0.97, suggesting that the overall severity and prevalence of CM exposure were comparable in men (57.4%) and women (53.9%).

2.3.2. Marital Satisfaction

The Dyadic Adjustment Scale (DAS—7; Hunsley et al. 2001) is a 7‐item scale that includes (a) three items assessing the degree of dyadic agreement on a scale from 0 (always disagree) to 5 (always agree); (b) three items assessing the degree of dyadic cohesion on a scale from 0 (never) to 5 (more often); and (c) one item assessing global satisfaction on a scale from 0 (extremely unhappy) to 6 (perfect). The scale score was an equally weighted average of response scores for all the items. The DAS‐7 has good criterion and discriminant validity (Hunsley et al. 2001). In this study, Cronbach's alpha was adequate for men (α = 0.83) and women (α = 0.78).

2.3.3. Demographic Measures

Participants filled out brief demographic questions (e.g., age, education, income, relationship length, number of children, socioeconomic status).

2.3.4. Problem‐Solving Discussion Task

We assessed couple communication quality through a 10 min discussion using a conflict discussion paradigm adapted from Bullard et al. (2010) for the Family Interaction Tasks (FITs) procedure (Forgatch and DeGarmo 1999). First, partners were asked to rate a current conflict from a checklist of frequent marital conflicts (Cowan and Cowan 1990) in order of severity. Next, couples were instructed to attempt to solve the conflict issue that was ranked as the most severe in 10 min, during which they were left alone in an observation room while their interaction was recorded.

2.3.5. Coding Couple Interactions

The coding of the interactions involved a team of five coders who underwent a comprehensive training program totaling 30 h, supplemented by weekly recalibration meetings to maintain consistency and minimize observer drift. Inter‐rater reliability was evaluated for 20% of the videos, showing satisfactory inter‐coder agreement for all codes in both men and women (ICC = 0.71–0.90).

We employed the Specific Affect Coding System (SPAFF) developed by Gottman (Coan and Gottman 2007) to assess communication behaviors. The SPAFF integrates verbal and non‐verbal cues such as voice tone, context, facial expressions, gestures, and body movements. Although the original SPAFF utilizes a micro coding approach, our study adopted a macro coding system. Each coder was instructed to watch each interaction twice and rate 10 behaviors representing positive (affection, humor, interest, validation/empathy) or negative (anger, contempt, criticism, stonewalling, defensiveness) communication on a scale from 1 (never) to 7 (always). The tenth code, psychological abuse, was adapted from the well‐established Rapid Marital Interaction Coding System (Heyman 2004), encompassing psychological aggression during the interaction including expressions of disgust, belligerence, threatening, domineering, and gaslighting behaviors.

We expected the 10 items of the communication codes to load onto two factors, positive and negative communication for men and women. We conducted a confirmatory factor analysis (CFA) using AMOS 25 (Arbuckle 2013). A two‐factor CFA provided a good fit to data for women, χ ² = 32.31, df = 34, p = 0.550, CFI = 1.00, RMSEA = 0.00, but only a mediocre fit to the data for men, χ ² = 53.05, df = 37, p = 0.02, CFI = 0.93, RMSEA = 0.70, which was not adequate. Inspection of loading coefficients indicated that one item, “stonewalling”, in men (Factor Loadings = 0.23) and women (Factor Loadings 0.24) reached a poor loading coefficient. Stonewalling represents a passive‐aggressive behavior, whereas the rest of the behavioral codes represent an active form of negativity (Factor Loadings: 0.33–0.816). Therefore, we decided to exclude this item. The revised model of nine items had better fit indices in men, χ ² = 37.33, df = 26, p = 0.70, CFI = 0.96, RMSEA = 0.06, and women, χ ² = 21.24, df = 26, p = 0.73, CFI = 1.00, RMSEA = 0.00, with item loadings for the negative scales ranging between 0.31–0.79 in men and 0.31–0.88 in women and item loadings for the positive scale ranging between 0.49–0.65 in men and 0.35–0.63 in women. Cronbach's alpha for the negative communication scale was satisfactory in men (α = 0.77) and women (α = 0.79). We computed a negative communication scale by averaging anger, contempt, criticism, defensiveness, and hostility.

2.4. Data Analysis

The preliminary analysis and mediation analyses were conducted using SPSS 29. Hypotheses were tested utilizing an Actor–Partner Interdependence Model (APIM), which considers the interdependence between dyad members. This model allows for the examination of the effects of each partner on their own outcome (actor effect) as well as on their partner's outcome (partner effect; Kenny et al. 2006). We employed the APIM Extended to Mediation (APIMeM) via MEDYAD, an SPSS macro (Jacob et al. 2019). This macro enables the estimation of dyadic mediation models with one or multiple mediators. Indirect effects were estimated using 5000 bootstrapped samples, producing bias‐corrected 95% confidence intervals (CI). If the CI includes zero, the inference is that the population indirect effect is zero; if zero falls outside the CI, the indirect effect is considered non‐zero (Coutts et al. 2019). There was no missing data in this study.

3. Results

3.1. Preliminary Analysis

Descriptive statistics and zero‐order correlations between all study variables are presented in Table 1. A point‐biserial correlation testing the relationships between CM in men and women and the study variables indicated that CM in women was positively correlated with more negative communication in men and women. Men's stonewalling was positively associated with women's stonewalling behaviors, as well as with men's and women's negative communication. It was also negatively associated with marital quality reported by men and women. In women, stonewalling was positively associated with their negative communication behaviors. Men's and women's negative communication were strongly and positively correlated, with men's negative communication associated with lower marital quality, whereas women's was associated with their own and their partner's lower marital quality. Given the high correlations between negative communication in men and women (r = 0.68, p < 0.01), we tested multicollinearity, which indicated VIF of 1.86, tolerance of 0.54, pointing to nonsignificant multicollinearity.

TABLE 1.

Means, standard deviations, sample size, and bivariate correlations among the study variables.

	CM men	CM women	Stonewalling men	Stonewalling women	Negative CMN men	Negative CMN women	Marital quality men	Mean	SD
CM men	—
CM women	0.15	—
Stonewalling men	0.11	−0.05	—					1.56	1.10
Stonewalling women	0.04	0.04	0.28	—				1.42	0.90
Negative CMN men	0.05	0.19*	0.19*	0.13	—			3.6	1.46
Negative CMN women	0.06	0.27**	0.25*	0.19*	0.68**	—		3.74	1.50
Marital quality men	−0.02	−0.08	−0.19*	−0.07	−0.17	−0.30**	—	26.11	4.46
Marital quality women	−0.08	−0.14	−0.18*	−0.01	−0.27**	−0.40**	0.53**	26.67	3.80

Abbreviation: Negative CMN = total negative communication.

*p < 0.05. **p < 0.01.

Next, a Pearson correlation coefficient was computed to assess the associations between socio‐demographic variables, including age, family income, education level, durations of marriage, number of children, with the mediating and outcome variables, namely stonewalling, negative communication, and marital quality, respectively. No significant correlations were identified and therefore no socio‐demographic variables were incorporated as covariates in the model.

In the following step, we tested the study hypotheses—whether negative communication behavior mediated the actor and partner associations between CM and marital quality. Because stonewalling was not found significantly linked with CM among men and women, it was not included in the mediation analysis.

Figure 1 shows the results of the APIM revealing that women's CM was associated with higher levels of negative communication in women, and with nearing significance in men. Women's negative communication was linked to lower marital quality for both men and women. However, CM history in men was not related to the study variables (see Table 2).

FIGURE 1.

APIM mediation analysis of negative couple communication on the actor and partner links of CM and marital quality. Negative CMN = negative communication. CM (0 = no CM 1 = CM history). Entries are partially standardized structural coefficients. Dashed paths are not significant. ϯ = 0.054. *p < 0.05. **p < 0.01.

TABLE 2.

Regression coefficients.

	Coefficient (B)	SE	Lower CI	Upper CI
Total and direct effects of CM on MQ
Total actor effect men	−0.09	0.09	−1.80	1.62
Total actor effect women	−0.96	0.71	−2.38	0.46
Total partner effect men	−0.38	0.72	−1.81	1.06
Total partner effect women	−0.69	0.85	−2.39	0.99
Direct actor effect men	−0.05	0.83	−1.70	1.60
Direct actor effect women	−0.33	0.86	−1.68	1.83
Direct partner effect men	−0.20	0.69	−1.68	1.02
Direct partner effect women	0.00	0.85	−1.68	1.70
Path
CM men > CMN men	0.08	0.27	−0.47	0.63
CM men > CMN women	0.05	0.28	−50	0.60
CM women > CMN men	0.52	0.27	−0.08	1.06
CM women > CMN women	0.79	0.27	0.25	1.34
CMN men > MQ men	0.17	0.38	−0.58	0.93
CMN men > MQ women	−0.015	0.31	−0.63	0.60
CMN women > MQ men	−1.00	0.37	−1.75	−0.25
CMN women > MQ women	−0.94	0.01	−1.56	−0.33
Indirect effects
CM men > CMN men > MQ men	−0.01	0.21	−0.25	0.27
CM men > CMN men > MQ women	−0.00	0.08	−0.16	16
CM men > CMN women > MQ men	−0.05	0.30	−0.65	0.59
CM men > CMN women > MQ women	−0.04	0.28	−0.59	0.54
CM women > CMN women > MQ women	−0.76	0.34	−1.50	−0.17
CM women > CMN women > MQ men	−0.80	0.41	−1.74	−0.11
CM women > CMN men > MQ women	−0.00	0.17	−0.36	0.36
CM women > CMN men > MQ men	0.09	0.24	−0.35	0.64

Note: Bold paths are significant.

Abbreviation: CMN = total negative communication.

Testing the indirect associations indicated a significant indirect link between women's CM and marital quality for both women, B = −0.10, SE = 0.05, 95% CI [−0.20, −0.02], and men, B = −0.09, SE = 0.05, 95% CI [−0.19, −0.02], through women's negative communication. The total and the direct actor‐partner associations between CM and marital quality were not significant (see Table 2). Overall, the model explained 3% and 7% of the variance in men's and women's negative communication and 9% and 15% of the variance in marital quality for men and women, respectively.

In sum, negative couple communication mediated the association between women's CM exposure and marital quality for both partners. No such effects were observed in men.

4. Discussion

This study aimed to broaden the knowledge about dyadic mechanisms through which a history of CM may hinder marital quality. Employing the CATS model (Goff and Smith 2005), we examined whether negative couple communication mediated the relationships between each partner's CM history and their own and their partner's marital quality. Our hypotheses were partly supported. As expected, negative couple communication mediated the indirect associations between exposure of women to CM and marital quality in both men and women. Specifically, women reporting a CM history expressed more negative communication behaviors during a conflict discussion, which, in turn, were associated with lower marital quality perceived by men and women. Contrary to our expectations, we did not find evidence for similar patterns when men were exposed to CM.

The study illuminates an indirect path through which women's CM may reverberate across generations, shaping not only their own marital experience but also their partner's. In other words, even men without a history of CM were found at risk of facing lower marital quality when paired with women survivors of CM. These results converge with research showing that all forms of CM experienced by one partner can disrupt the marital functioning of both dyad members (for a review, see Vaillancourt‐Morel et al. 2024; Zamir 2022). However, the current findings are novel, suggesting a possible mechanism wherein women's negative communication accounts for the link between their CM experiences and the quality of marriage in both men and women.

Intimate partners constantly reciprocate thoughts, feelings, and behaviors (Kelley and Thibaut 1978). Resonating on this idea of interdependence, our findings align closely with the CATS model, postulating that complex trauma (i.e., CM) may indirectly affect the relationship quality of both dyad members through a mutual process taking place in an intimate relationship. CM may result in post‐traumatic distress, implicating emotion regulation difficulties such as hyperarousal, which compromise the capacity for regulated behaviors while interacting with a romantic partner. The persistent negative communication dynamics involving expressions of anger, contempt, hostility, criticism, and defensiveness can gradually undermine the sense of security within the relationship and ultimately the perceived quality of the marriage (Goff and Smith 2005).

This study highlights the crucial role of negative communication in perpetuating adverse relationship patterns from childhood into marriage. Effective marital communication is essential for resolving conflicts, an ordinary marital challenge (Gottman 1994). Furthermore, negative marital dynamics can cascade to affect parent–child interactions (Conger et al. 2002; Gewirtz et al. 2017), potentially perpetuating intergenerational transmission of CM. Whereas research has predominantly focused on the link between CM and marital quality or satisfaction, with limited attention given to marital communication (Zamir 2022), the current findings stress the need to focus more on interaction dynamics related to CM. Negative communication can escalate to IPV (Heyman et al. 2023), further impair relationship closeness (Bodenmann 2005), and render the mental and physical well‐being of those involved (Pickover et al. 2017).

The findings suggest that women may be more susceptible to the effects of CM experiences in terms of negative communication behaviors. It should be noted that the literature on gender differences in the effects of CM on marital functioning is inconclusive, with some evidence indicating significant or stronger effects in women, whereas others indicate the opposite (Zamir 2022). Moreover, variations in research methodologies, such as using different measures (like self‐reports of overall communication quality), focusing on specific CM, or research designs (an individual model), further complicate reaching a clear conclusion (Zamir 2022).

The divergence across genders observed in our findings may be attributed to gender socialization processes taking place in abusive homes during childhood. Girls are socialized to prioritize interdependence within relationships. As they grow into adulthood, societal norms expect women to invest in nurturing and maintaining close relationships, often taking on more caregiving responsibilities (Cross and Madson 1997). When interacting with their spouse, women typically initiate conflict discussions, motivated to solve the problem in order to preserve the relationship (Christensen et al. 2006). Women also play a central role in de‐escalating conflicts (Gottman et al. 1998). However, exposure to violent and ineffective problem‐solving behaviors modeled in a maltreating environment during childhood may compromise the ability to acquire adaptive communication skills (Bandura and McClelland 1977; Shipman et al. 2007), and subsequently manage constructive conflicts with their partner later in life. Without effective de‐escalating processes, couples risk prolonging cycles of negative reciprocity, wherein negative communication from one partner triggers a negative response from the other, potentially exacerbating tensions and hindering conflict resolution (Gottman et al. 1998).

Considering that physiological arousal associated with CM‐related distress was postulated to elicit negative interaction behaviors (Goff and Smith 2005), it could be that physiological responses during marital conflict have differential implications for men and women. For example, research has shown that women's physiological changes during couple conflict are more strongly associated with relationship outcomes, including negative behaviors (Gottman and Levenson 1999; Kiecolt‐Glaser and Newton 2001), reflecting their greater sensitivity to stress response during negative marital interactions. Similarly, while women who were dissatisfied in their relationships had higher adrenocorticotropic hormone levels, indicating a stress response, at the beginning of a conflict discussion, compared to those who were satisfied, no difference was found among men based on relationship satisfaction (Kiecolt‐Glaser et al. 2003). Given that we assessed couple communication behaviors under a stressful task—a conflict discussion, and given that CM may intensify physiological arousal (Goff and Smith 2005), it could be that women with CM histories were more inclined to exhibit negative communication.

Altogether, the model underscores the central role of CM history in women in the relationship functioning of both dyad members. This is also reflected in our finding showing that CM in women was correlated not only with women's expressions of negative communication but also with men's. This link, however, reached only marginal significance in the dyadic model. The results of the present study, while necessitating further validation, raise questions about the mechanisms underlying the effects of men's CM on men's and women's marital functioning.

4.1. Limitations and Future Directions

The present study offers insight into a dyadic mechanism by which a history of CM in women predicts lower marital quality of men and women, and through women's communication patterns. However, it is important to acknowledge several limitations of the findings. First, the retrospective cross‐sectional design constrains the ability to establish causality. Additionally, although the adverse effects of CM on relationship quality were repeatedly documented in prospective and retrospective studies (Labella et al. 2017; Zamir 2022), relying on recollections of traumatic events from a distant past may introduce bias. Second, our investigation focused primarily on negative communication during conflict interactions, which, while validated and reliable (e.g., Gottman et al. 1998), may benefit from additional ecological validity. For instance, future studies could employ methods such as daily diary studies, involving multiple informants. Furthermore, our focus on negative communication limits our understanding of the role of positive interaction dynamics in the context of CM. Intimate partners are a major source of support, protecting against the adverse effects of trauma (Brewin et al. 2000; Ozer et al. 2003). Incorporating observations of supportive and other positive communication patterns could offer a more comprehensive understanding of couple functioning in the context of CM.

Our study sample included mostly academically educated individuals who may demonstrate higher levels of resilience in coping with CM. The high‐functioning nature of our sample could account for the lack of significant direct associations between CM and marital quality that we observed. Future research should include more diverse and socio‐demographically representative samples, potentially including clinical populations with documented histories of childhood abuse. Furthermore, although post‐traumatic symptoms were postulated to impede communication quality among couples affected by CM (Goff and Smith 2005), our community sample, comprised of high‐functioning individuals, may not exhibit clinical‐level post‐traumatic symptoms. It is plausible that communication challenges in CM cases were acquired through social learning dynamics, such as negative communication patterns and core beliefs about relationships developed within the family of origin. Thus, future research should focus on the mechanism underlying the effects of CM on communication behaviors to validate the CATS model.

Finally, given that our sample included mixed‐sex Israeli married couples, replication of our model is needed within LGBTQ communities as well as other cultures. Caution is warranted when generalizing our findings in a different sociocultural context. Israel is a society adopting traditional family‐centered values with modern, egalitarian norms (Lavee and Katz 2003). For instance, while Jewish Israeli and American populations share some core values regarding family life, Israeli culture places a greater emphasis on legal marriage (Gavriel‐Fried and Shilo 2015). Importantly, while research has demonstrated mostly universal patterns related to conflict issues, it is possible that, similarly to traditional cultures (Dillon et al. 2015), Israeli women may be more accepting of conventional gender roles including the household division of labor. Furthermore, compared to Western societies, Israeli couples tend to be more involved in and influenced by extended family systems and receive more parental support (Lavee and Katz 2003). It could be that individuals with a history of CM, who may lack access to this resource, would be more vulnerable to relational stress. It is also important to note that Israel provides comprehensive social services, including legal protections against CM and universal access to healthcare (e.g., psychotherapy), which may help mitigate the long‐term impact of CM on adult relationships.

4.2. Clinical Implications

The present study underscores the critical role of negative communication expressed by women as a key mechanism through which CM may hamper the marital system in non‐clinical couples. Although these findings warrant further replication, and although generalizability to clinically distressed couples should be made cautiously, they may offer insights for clinical interventions. Specifically, fostering healthy communication skills may benefit couples with a history of CM. Evidence‐based couple therapies have been shown to improve positive interpersonal skills and marital satisfaction (Hawkins et al. 2008). Notably, our findings indicate that even in a non‐clinical population, women's negative communication may hinder healthy relationships. Given that CM remains prevalent yet underreported and untreated (Stoltenborgh et al. 2015), many individuals may reach adulthood without adequate support. We therefore recommend that interventions for distressed couples may benefit from systematically assessing trauma histories. When CM is identified among women, targeted couple communication skills training could facilitate more supportive and constructive interactions, promoting healthier relationship dynamics. In addition, our findings highlight the potential value of prevention interventions designed to strengthen couple communication skills in this at‐risk population.

5. Conclusion

Employing the CATS model, the current study examined whether negative couple communication mediates the actor and partner associations between CM and marital quality. Our findings were partly supported, indicating that a history of CM in women is associated with more expressed negative couple communication among women, and in turn with lower marital quality in men and women. This model reveals a process by which women's experiences of CM can reverberate and impair the marital well‐being of both partners, above and beyond the effects of a man's experiences of CM, suggesting broader intergenerational and systemic effects of CM on adult relationships.

Conflicts of Interest

The authors declare no conflicts of interest.

Zamir, O. , Adar O., Cohen D. B., Goldberg C., Regev G. M., and Shapira M.. 2025. “Intergenerational Transmission: Observed Negative Communication Mediates Dyadic Associations Between Childhood Maltreatment and Marital Quality.” Family Process 64, no. 3: e70059. 10.1111/famp.70059.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.