Associations among PTSD Symptoms, Fear of Emotion, and Couple Communication Difficulties: A Between-Person Dyadic Analysis

Abstract

This study investigated between-person associations among PTSD symptoms, fear of emotion, and perceived couple communication difficulties in a dyadic context among 64 trauma-exposed, mixed gender community couples (N = 128 individuals) using the Actor-Partner Interdependence Mediation Model. Individuals with higher PTSD symptoms endorsed greater fear of their emotions (β_Men = .72; β_Women = .49), and those with greater fear of their emotions reported lower levels of constructive couple communication (β_Men = −.19; β_Women = −.21) and higher levels of self-demand/partner-withdraw communication (β_Men = .20; β_Women = .25) and partner-demand/self-withdraw communication (β_Men = .26; β_Women = .33) with their partners. Additionally, women whose partners had higher PTSD symptoms endorsed greater fear of their emotions (β = .30). The most robust indirect PTSD-communication links were between (a) individuals’ PTSD symptoms and their perceptions of partner-demand/self-withdraw communication when accounting for associations with fear of their emotions and (b) men’s PTSD symptoms and women’s perceived partner-demand/self-withdraw communication accounting for associations with women’s fear of their emotions. Men with higher PTSD symptoms also reported greater self-demand/partner-withdraw communication (β = 0.28), independent of their fear of emotion. Couple-based treatments for PTSD that promote emotional tolerance and are sensitive to gender differences in how PTSD symptoms relate to each partner’s perception of the man-demand/woman-withdraw communication pattern may improve trauma survivors’ relationship functioning and increase the potential for relationships to serve as a conduit for recovery from PTSD.

Posttraumatic stress disorder (PTSD; American Psychiatric Association, 2013) symptoms are associated with a host of close relationship problems. These include relationship discord, parenting impairments, and aggression towards partners and children (Giff et al., 2019; Marshall et al., 2019; Taft et al., 2011), as well as partner and child psychological distress (Lambert et al., 2012; 2014). A growing body of research has implicated couple communication difficulties in the linkages between PTSD symptoms and many of these outcomes, including couple relationship distress and parenting difficulties (e.g., Allen et al., 2010; Fredman et al., 2019; Giff et al., 2019). Both members of the couple physically experience the impact of PTSD, as literature demonstrates that dysfunctional communication patterns mediate the link between PTSD symptoms and both partners’ negative health sequelae (e.g., heightened cardioreactivity; Smith et al., 2021). Furthermore, how traumatized individuals communicate with their partners to resolve relationship difficulties has important prognostic implications for recovery from PTSD (Evans et al., 2009; Fredman et al., 2017). Notably, difficulty with communication is one of the most common reasons people seek couple therapy in general (Doss et al., 2004) and a prominent reason that veterans with PTSD desire family involvement in their care (Batten et al., 2009).

To optimize outcomes for trauma survivors in intimate relationships, it is important to both characterize the types of communication difficulties that trauma survivors experience with their partners as well as predictors of those communication difficulties. Communication behaviors in couples are typically categorized as positive or negative. Constructive communication, a type of positive dyadic communication behavior, and demand/withdraw, a negative dyadic communication pattern, are robust predictors of couple relationship adjustment and distress (Heavey et al., 1996; Litzinger & Gordon, 2005; Schrodt et al., 2014). Constructive communication refers to a collaborative approach to problem solving in which both partners discuss the problem, listen to the other’s perspective, suggest compromises, and express their feelings. Demand/withdraw refers to a communication pattern in which one partner nags, criticizes, complains, or otherwise attempts to initiate change (i.e., “demands”), and in response the other partner avoids or retreats from the interaction (i.e., “withdraws”); further, the more one partner withdraws, the more the other partner is inclined to demand (Christensen, 1987). This communication pattern can also be conceptualized as one partner having a negatively valenced approach orientation while the other partner has an avoidance orientation toward the partner or problem (Crenshaw et al., 2017). Constructive communication is positively associated with relationship satisfaction (Heavey et al., 1996). In contrast, the demand/withdraw behavior pattern is associated with relationship distress (Schrodt et al., 2014) and intimate partner aggression (Jarnecke et al., 2016).

There is a consistent finding that PTSD symptoms are associated with poorer quality of couples’ communication. This has been demonstrated in military samples consisting primarily of male service members or veterans (Allen et al., 2010; 2018; Knobloch-Fedders et al., 2017; Miller et al., 2013) and in mixed gender civilian samples (Fredman et al., 2017; 2019). This has also been found when the communication is observed (Knobloch-Fedders et al., 2017; Miller et al., 2017) or self-reported (Allen et al., 2010; 2018; Fredman et al., 2017; 2019). Further, this association has been observed cross-sectionally (Allen et al., 2010; Fredman et al., 2019; Knobloch-Fedders et al., 2017; Miller et al., 2013) and longitudinally (Allen et al., 2018; Fredman et al., 2017).

At the same time, much remains unknown about the associations between PTSD symptoms and couple communication. Within these studies, couple communication is typically characterized in a manner that decontextualizes one partner’s behavior from the other partner’s behavior, despite couple communication being an inherently transactional process. That is, instead of characterizing patterns of contingent behavioral exchanges between partners, existing research largely focuses on total frequency counts of positive and negative behaviors by each member of the couple observed during conflict discussions (e.g., Miller et al., 2013) and individuals’ self-reports of the extent to which they or their partners each individually engage in particular negative relationship behaviors and the frequency and severity of arguments with their partners generally (e.g., Allen et al., 2010; 2018; Fredman et al., 2019). One way that the field has attempted to recognize the interdependence of partner behaviors is by examining the degree to which one member of the couple’s PTSD symptoms are associated with the other partner’s communication behavior or perceptions of the couple’s conflict communication (i.e., between-person cross-partner effects; Fredman et al., 2019; Miller et al., 2013). The next step for the field is to measure interactional communication patterns, such as demand-withdraw, in addition to examining such cross-partner effects, to permit more insight into how PTSD symptoms and couple communication processes are related.

Additionally, even though observed communication is generally regarded as the gold standard in the field, there may be particular value in examining perceived (i.e., self-reported) communication patterns when studying couple communication in the context of PTSD. The psychopathology of PTSD includes cognitive biases that play out in interpersonal domains (Ehlers & Clark, 2000). Often, these include biased appraisals of threat from external stimuli, which may include one’s intimate partner (Monson et al., 2010). Consequently, appraisals of threat may lead to enhanced perception of partner demanding behaviors that would not be captured by objective external raters. Furthermore, patients’ reports of communication problems (as well as other patterns of family dysfunction) predict attenuated response to individual psychotherapy for PTSD (Evans et al., 2009), thus underscoring the clinical importance of considering individuals’ perceptions of couple communication patterns.

The psychopathology of PTSD is also centered on cognitive biases that include perceptions of threat from internal stimuli. Cognitive-behavioral conceptualizations of PTSD (e.g., Ehlers & Clark, 2000; Foa & Kozak, 1986; Weiss et al., 2018) highlight that emotional arousal itself, whether negatively or positively valenced, can serve as a trauma cue. This may result from learning due to temporal associations between trauma exposure and emotional arousal, as well as excessively negative and/or de-contextualized appraisals of the traumatic event(s) and the individual’s emotional reactions to it. During and shortly after a traumatic event, some individuals experience a range of negative emotions (e.g., fear, disgust, sadness; Rizvi et al., 2011). To alleviate their distress in the short-term, they may attempt to avoid both external (e.g., people, places, sights, sounds) and internal trauma reminders (e.g., emotional arousal). Because avoidance temporarily alleviates psychological distress, individuals are more likely to continue avoiding stimuli that elicit distress. However, this avoidance impedes the ability to develop a cogent narrative regarding the traumatic event and its aftereffects (e.g., emotional experiencing) and can contribute to catastrophic beliefs about emotions themselves (e.g., “Feeling anxious means I am in danger,” “If I start feeling sad, I’ll never stop,” “I will lose control if I feel angry”; Ehlers & Clark, 2000). Indeed, a well-established body of literature indicates that PTSD symptoms are associated with fear of one’s emotions (Arditte-Hall & Mitchell, 2022) and with emotion regulation strategies that aim to suppress or minimize emotional experiencing (Seligowski et al., 2015).

Romantic relationships are a key context for the experience of emotions. Given that individuals with elevated PTSD symptoms tend to fear their own emotion (Arditte-Hall & Mitchell, 2022) and that relationship disagreements tend to elicit strong emotions (Shrout et al., 2023), individuals who exhibit greater fear of emotion may engage in less constructive communication patterns with their partner during conflict. Common to multiple empirically supported general couple therapies for relationship distress (e.g., Baucom et al., 2020; Christensen et al., 2014; Johnson et al., 2023) is the notion that emotions play a functional role in couple behavior, including couples’ communication behavior. Although the specific techniques vary across interventions, all of these therapies seek to reduce demand/withdraw and other negative couple communication patterns (e.g., mutual avoidance) by modifying the way individuals perceive, respond to, and ultimately tolerate their own emotions.

The cognitive-behavioral interpersonal theory of PTSD (Monson et al., 2010) posits that, among couples in which one or both partners experience PTSD symptoms, demand-withdraw and mutual avoidance communication patterns exist by virtue of the psychopathology of the disorder and how it manifests interpersonally. According to this model, cognitive biases for threat include fear of one’s own emotions (Williams et al., 1997), which makes it difficult to regulate them. Behaviorally, this can manifest in dysregulation of the fight or flight response, with verbal and physical aggression and other negatively valenced approach behaviors (e.g., criticism, blame) representing dysregulation in the “fight” part of the fight or flight response and avoidance and withdrawal representing dysregulation in the “flight” part of this response. Within the dyadic context, this can play out with trauma survivors perceiving their partner as a stressor against which they need to fight (e.g., “demand”) and/or from which they need to flee (e.g., “withdraw”) to alleviate their distress in the moment (Monson et al., 2010). Partners may respond by attempting to engage a withdrawn trauma survivor using a range of approach strategies that include trying to discuss a problem, nagging, criticizing, and threatening (“demanding”) and/or by withdrawing themselves. These behaviors may occur in response to partners’ perceptions of the trauma survivor’s demanding behavior or as a preemptive attempt to avoid conflict. Both partners may also engage in chronic avoidance of conflict resolution attempts given past unsuccessful attempts. These couple-level transactions give rise to stable patterns of behavior and couples’ characterizations of such. However, whether fear of emotion accounts for any of the association between PTSD symptoms and dysfunctional couple communication has never been formally tested.

Current Study

The goal of the present study was to examine between-person associations among PTSD symptoms, fear of emotion, and perceived couple communication patterns, and, more specifically, the degree to which fear of one’s own emotions in a dyadic context may partly account for the link between PTSD symptoms and perceived couple communication. We tested this question in a community sample of trauma-exposed mixed gender couples. PTSD symptoms were measured using the gold standard clinical interview for both members of the couple and measured dimensionally, as has been recommended (Broman-Fulks et al., 2006). We hypothesized that individuals with higher PTSD symptoms would endorse greater fear of emotion (Hypothesis 1a) and that individuals with greater fear of emotion would perceive more difficulties resolving problems with their partners (i.e., lower levels of constructive communication and higher levels of demand/withdraw communication; Hypothesis 1b). Further, we expected that there would be evidence of an indirect association between individuals’ PTSD symptoms and their perceptions of couple communication difficulties when accounting for associations with fear of their own emotions (Hypothesis 2; see Supplemental Figure 1). Cross-partner associations and gender differences were also examined given the existence of some, albeit inconsistent, prior demonstration of differences between men and women with respect to linkages between PTSD symptoms and relationship functioning (e.g., Lambert et al., 2012; Liebman et al., 2023). However, no formal hypotheses were made, treating these instead as research questions.

Method

Participants

The current study included 64 married or cohabiting couples (N = 128 individuals) in mixed gender relationships who participated in a larger study that aimed to examine associations between trauma and relationship functioning. Participants were recruited from rural or semi-rural communities between 2008 and 2010 via local advertisements (e.g., newspaper advertisements, flyers in mental health clinics) that targeted couples with at least one partner who had experienced a stressful life event. Each partner completed separate telephone screens and were eligible to participate if they were cohabiting, at least 18 years old, and at least one partner met screening criteria for probable PTSD using the Posttraumatic Checklist-Civilian version (i.e., score ≥ 44; Weathers et al., 1993) based on criteria in the Diagnostic and Statistical Manual of Mental Disorders 4^th edition (DSM-IV; American Psychiatric Association, 1994). Couples were excluded if neither partner endorsed probable PTSD (n = 122), they were no longer interested in participating (n = 8), or they had ended their relationship (n = 1). To be representative of the local rural to semi-rural population, couples were also excluded if they had more than $100,000 in combined annual income and/or either partner had over six years of post-high school education (n = 3).

Most participants identified as non-Hispanic White (85.9%), followed by Black or African American (6.3%), biracial/multiracial (3.9%), non-White Hispanic or Latino (2.3%), and White Hispanic or Latinx (1.6%). On average, participants were 37.2 years old (SD = 12.6, range = 18–64). Most participants were employed full-time (40.7%; 31.4% part-time; 28.0% unemployed), and the average monthly individual income was $1,731.24 (SD = $1,522.18). Average education was 14.30 years (SD = 2.32). Most participants reported being married (71.7%); average relationship length was 12 years and 1 month (range = 6 months to 45 years).

Procedures

After a telephone screening, eligible couples completed either one eight-hour or two four-hour in-person lab sessions. During these sessions, each participant completed self-report questionnaires related to trauma and relationship health, interviews with clinically trained graduate students, and lab tasks. Participants were compensated $175 for completing the full study protocol. The study had approval from the Pennsylvania State University Institutional Review Board, and all participants provided informed consent.

Measures

PTSD Symptom Severity

The Clinician Administered PTSD Scale (CAPS; Blake et al., 1995) is the gold standard semi-structured interview used to assess severity of PTSD symptoms according to the DSM-IV (American Psychiatric Association, 1994) criteria. Each DSM-IV PTSD symptom is rated on a 5-point scale from 0 to 4 for the frequency and intensity that were experienced by the participant in the past month. Ratings are then summed, resulting in a composite score ranging from 0 to 136, such that higher scores indicate greater severity of PTSD symptoms. Psychometric properties of the CAPS are strong (Blake et al., 1995; Weathers et al., 2001), and there was high internal consistency (Cronbach’s α = .94) and interrater reliability for 10% of interviews (intraclass correlation = .92) in the current study sample.

Fear of Emotion

The Affective Control Scale (ACS; Williams et al., 1997) is a 42-item measure assessing the extent to which individuals fear different aspects of emotional experiences, such as experiencing the emotion itself, loss of emotional control, or behavioral reactions to the emotions. Participants rate each item (e.g., “It scares me when I am nervous”) using a 7-point Likert response scale ranging from 1 (very strongly disagree) to 7 (very strongly agree). Scores are summed such that higher scores reflect greater fear of emotion. The measure demonstrates evidence for construct validity and test-retest reliability (Williams et al., 1997). For the current study, we used a 27-item psychometrically optimized version of the ACS based on factor analysis in a trauma-exposed sample (Arditte-Hall & Mitchell, 2022), which demonstrated high internal consistency in the current study sample (Cronbach’s α = .95).

Couple Communication

The Communication Patterns Questionnaire (CPQ; Christensen, 1987; Christensen & Sullaway, 1984) was initially developed as a 35-item measure assessing individuals’ perceptions of how they and their partner communicate with each other before, during, and after a conflict. The CPQ has good convergence with observational measures of couples’ communication (Hahlweg et al., 2000). In the current study, we used a 23-item version derived from factor analysis to produce a more reliable and valid measure relative to the original (Crenshaw et al., 2017). This includes three subscales reflecting the individual’s perception of different types of couple communication when problems arise in the relationship: constructive communication (e.g., “Both my partner and I try to discuss the problem”), self-demand/partner-withdraw (e.g., “I nag and demand while my partner withdraws, becomes silent, or refuses to discuss the matter further”), and partner-demand/self-withdraw (e.g., “My partner nags and demands while I withdraw, become silent, or refuse to discuss the matter further”). Participants rate each item using a 9-point Likert response scale ranging from 1 (very unlikely) to 9 (very likely). For each subscale, responses are summed such that higher scores reflect greater perceptions that the specific communication pattern occurs in the relationship. The CPQ demonstrates evidence for reliability, construct validity, sensitivity to change, and factor structure (Crenshaw et al., 2017). Internal consistency was good for the constructive communication (Cronbach’s α = .83), self-demand/partner-withdraw (Cronbach’s α =.82), and partner-demand/self-withdraw (Cronbach’s α =.78) subscales in the current study.

Analytic Plan

Data were analyzed using the Actor-Partner Interdependence Mediation Model (APIMeM; Ledermann et al., 2011) in a path modeling framework using Mplus 8.3 (Muthén & Muthén, 1998–2017). The percentage of missing values across the five variables ranged from 0.78% to 6.25%. Little’s Missing Completely at Random test (Little, 1988) confirmed that the data were missing completely at random (χ2(7) = 7.80, p = .35). To account for missing data and the non-normal distribution of two study variables (men's PTSD symptoms and men's self-demand/partner-withdraw communication), we employed robust maximum likelihood estimation (MLR) in Mplus. No extreme outliers were identified for any variables. Model goodness of fit was interpreted based on recommendations by Hu and Bentler (1999): (a) non-significant χ2 test; (b) root mean square error of approximation (RMSEA) ≤ .06; (c) standardized root mean squared residual (SRMR) ≤ .08; and (d) comparative fit index (CFI) and the Tucker-Lewis index (TLI) greater ≥.95.

The indirect associations between individuals’ PTSD symptoms and their perceptions of couple communication behaviors (constructive communication, self-demand/partner-withdraw communication, partner-demand/self-withdraw communication) when accounting for associations with their own fear of emotion were modeled for men and women. The indirect associations between individuals’ PTSD symptoms and their partners’ perceptions of couple communication behaviors when accounting for associations with individuals’ own fear of emotion and their partners’ fear of emotion were also modeled, as were the indirect effects from individuals’ PTSD symptoms to individuals’ perception of couple communication behavior when accounting for associations with their partners’ fear of emotion. These indirect associations were modeled while simultaneously accounting for the direct associations between individuals’ PTSD symptoms and their own and their partners' perceptions of couple communication behaviors (direct actor and partner effects, respectively). Covariances between the predictors and among the residuals were also modeled.

The analyses were conducted in the following order: First, we tested a fully saturated model with all paths freely estimated. Second, gender differences were tested using the χ2 difference test adjusted for MLR to assess if imposing equality constraints across gender significantly deteriorated the model fit. When the χ2 difference test was not significant, indicating a lack of difference in the strength of the path between genders, paths were constrained to be equivalent across genders to obtain a more parsimonious model. The joint significance test suggests that, when the path from X to M and the path from M to Y are both significant, the indirect effect from X to Y through M is also significant (Fritz & MacKinnon, 2007), and it has been recommended for testing indirect effects in single- and two-mediator sequential chains (e.g., Taylor et al., 2007). However, others have found it to be both too liberal and too conservative across various circumstances; thus, it has been recommended that the joint significance test be used in conjunction with other tests (Fossum & Montoya, 2023). Consequently, bootstrapping with 5000 samples (Preacher & Hayes, 2008; Tibbe & Montoya, 2022) was used to evaluate the robustness of the indirect actor and partner effects indicated as significant by the joint significance test. A 95% confidence interval that does not contain zero would indicate a significant indirect effect.

Results

Descriptive Statistics and Correlations

Descriptive statistics and bivariate correlations among study variables are presented in Table 1. Compared to men, women endorsed significantly higher PTSD symptoms (t = 5.07, p <.001) and greater fear of emotion (t = 3.49, p <.001). However, there were no gender differences in men’s and women’s reports of constructive communication (t = −1.17, p = .24), self-demand/partner-withdraw communication (t = 1.19, p = .24), or partner-demand/self-withdraw communication (t = 1.14, p = .26).

Table 1.

Correlations, Means, and Standard Deviations for Study Variables

	1	2	3	4	5	6	7	8	9	10
1. Men’s PTSD Symptoms	–
2. Women’s PTSD Symptoms	−.16	–
3. Men’s Fear of Emotion	.74**	−.08	–
4. Women’s Fear of Emotion	.23	.33**	.10	–
5. Men’s CC	−.06	.14	−.03	.08	–
6. Women’s CC	−.11	.07	−.12	.16	.47**	–
7. Men’s SDPW	.47**	−.28*	.26	.01	−.40**	−.23	–
8. Women’s SDPW	.19	−.13	−.09	.28*	−.20	−.49**	.49**	–
9. Men’s PDSW	.22	−.17	.21	.00	−.49**	−.31*	.41**	.07	–
10. Women’s PDSW	.02	.13	.14	.35**	−.11	−.54**	.00	.26*	.49**	–
M	26.06	46.16	74.04	89.98	53.33	50.46	24.04	26.68	28.50	30.81
SD	24.33	20.30	23.57	26.33	12.49	14.23	11.52	12.61	11.26	10.96

Note. N = 64 dyads (128 individuals). PTSD = posttraumatic stress disorder; CC = constructive communication; SDPW = self-demand/partner-withdraw communication; PDSW = partner-demand/self-withdraw communication.

^*p < .05.

^**p < .01.

Path Model

The model fit the data well: χ2(11) = 10.69, p = .47; RMSEA = .00 [.00, .128]; SRMR = .07; CFI = 1.00, TLI = 1.00. The APIMeM with significant standardized parameters is depicted in Figure 1. Table 2 includes a full list of all unstandardized paths. Covariances among residuals are available upon request from the first author.

Figure 1. The Actor-Partner Interdependence Mediation Model with Standardized Parameters Depicting the Associations among PTSD Symptoms, Fear of Emotion, and Perceived Couple Communication Patterns.

Note. N = 64 dyads (128 individuals). PTSD = posttraumatic stress disorder. Dashed lines represent non-significant paths. For ease of presentation, parameter estimates for non-significant paths are not displayed but are reported in Table 2.

⁺ p < .10. * p < .05. * p < .01. *** p < .001.

Table 2.

The Actor-Partner Interdependence Mediation Model with Unstandardized Path Parameters

	Men		Women
	B (SE)	p value	B (SE)	p value
Actor Effects
PTSD ⟶ Fear of Emotion	0.689 (0.08)	< .001	0.689 (0.08)	< .001
Fear of Emotion ⟶ CC	−0.104 (0.05)	.042	−0.104 (0.05)	.042
Fear of Emotion ⟶ SDPW	0.105 (0.05)	.047	0.105 (0.05)	.047
Fear of Emotion ⟶ PDSW	0.126 (0.05)	.005	0.126 (0.05)	.005
PTSD ⟶ CC	0.078 (0.08)	.326	0.078 (0.08)	.326
PTSD ⟶ SDPW	0.138 (0.07)	.043	−0.116 (0.07)	.108
PTSD ⟶ PDSW	−0.005 (0.06)	.929	−0.005 (0.06)	.929
Partner Effects
Partner’s PTSD ⟶ Fear of Emotion	−0.015 (0.07)	.821	0.347 (0.11)	.001
Partner’s Fear of Emotion ⟶ CC	−0.042 (0.05)	.409	−0.042 (0.05)	.409
Partner’s Fear of Emotion ⟶ SDPW	−0.034 (0.04)	.412	−0.034 (0.04)	.412
Partner’s Fear of Emotion ⟶ PDSW	0.042 (0.05)	.359	0.042 (0.05)	.359
Partner’s PTSD ⟶ CC	0.049 (0.07)	.478	0.049 (0.07)	.478
Partner’s PTSD ⟶ SDPW	−0.129 (0.06)	.029	0.073 (0.07)	.306
Partner’s PTSD ⟶ PDSW	−0.071 (0.06)	.196	−0.071 (0.06)	.196

Note. B = unstandardized coefficient; SE = standard error; PTSD = posttraumatic stress disorder; CC = constructive communication; SDPW = self-demand/partner-withdraw communication; PDSW = partner-demand/self-withdraw communication.

Indirect Effects

Men and women with higher PTSD symptoms reported significantly greater fear of emotion, with no evidence of a gender difference in this association (Δχ2(1) = 2.30, p = .129). There was also a partner effect that did differ by gender (Δχ2(1) = 8.61, p = .003) such that female partners of men with higher PTSD symptoms reported significantly greater fear of emotion, whereas no association was observed between women’s PTSD symptoms and men’s fear of emotion. Men and women with greater fear of emotion reported significantly lower levels of constructive communication, higher levels of self-demand/partner-withdraw communication, and higher levels of partner-demand/self-withdraw communication with their partners, with no gender differences in these associations (Δχ2(1) = 0.02 to 0.78, all ps > .377).

Results from bootstrapping provided further evidence of a significant indirect association between individuals’ PTSD symptoms and their perceptions of partner-demand/self-withdraw communication when accounting for associations with fear of their own emotion (B = .087 [.015, .157]) for both men and women. In contrast, results from bootstrapping did not provide evidence of an indirect association between individuals’ PTSD symptoms and their perceptions of either constructive communication (B = −.071 [−.145, .008]) or self-demand/partner-withdraw communication (B = .073 [−.005, .176]) for men or women.

The indirect association between men’s PTSD symptoms and women’s perceptions of partner-demand/self-withdraw communication when accounting for higher levels of women’s fear of emotion was also further supported by results from bootstrapping (B = .044 [.004, .097]). In contrast, results from bootstrapping did not provide evidence of an indirect association between men’s PTSD symptoms and women’s perceptions of constructive communication (B = −.036 [−.089, .003]) or self-demand/partner withdraw communication (B = .037 [−.002, .100]).

Direct Effects

There were two significant direct paths between PTSD symptoms and perceived couple communication patterns. Specifically, there was an actor effect for men such that men with higher PTSD symptoms reported significantly greater self-demand/partner-withdraw communication; however, the same association was not found for women, with a significant gender difference (Δχ2(1) = 52.19, p < .001). There was also a partner effect that differed by gender (Δχ2(1) = 9.41, p = .002) such that male partners of women with higher PTSD symptoms reported significantly lower self-demand/partner withdraw communication, whereas no association was observed between men’s PTSD symptoms and women’s reports of self-demand/partner-withdraw communication.

Discussion

The goal of the current study was to extend prior literature and theory on the associations between PTSD symptoms and couple communication difficulties at the between-person level by investigating whether fear of emotion is an individual difference variable that may help to account for the linkages between PTSD symptoms and perceived communication difficulties. This was done in a dyadic context, which permitted the examination of cross-partner effects and gender differences in actor and partner effects. Accordingly, this study provides an empirical test of associations among PTSD symptoms, fear of emotion, and perceived couple communication described in the cognitive-behavioral interpersonal theory of PTSD (Monson et al., 2010) but never explicitly tested. We also used a mixed gender, trauma-exposed community sample rather than a military sample with male service members/veterans, as is commonly done, to effectively examine the roles of PTSD and fear of emotion for both genders.

Results demonstrated that, on average, individuals who have higher PTSD symptoms have greater fear of emotion and that individuals who have greater fear of emotion perceive poorer communication with their partners in the form of lower levels of constructive communication and higher levels of self-demand/partner-withdraw communication and partner-demand/self-withdraw communication. For some people, when PTSD manifests in the form of negative cognitions about the experience and consequences of feeling strong emotions, over time, they may come to experience their relationships as characterized by maladaptive communication patterns. Although there is a well-established literature demonstrating that PTSD and fear of emotion are linked (Arditte-Hall & Mitchell, 2022; Seligowski et al., 2015), this is the first study to empirically demonstrate that fear of emotion may exist within an interpersonal context among couples grappling with PTSD, with implications for how couples attempt to resolve relationship conflict. Moreover, the most robust evidence of an indirect association between individuals’ PTSD symptoms and perceived communication patterns when accounting for associations with individuals’ fear of emotion was for partner-demand/self-withdraw communication. As such, these results suggest that fear of emotion may be an aspect of the psychopathology of PTSD that helps to account for why individuals with higher levels of PTSD tend to be those who view their partners as a threat from which they need to retreat (Monson et al., 2010).

In examining the associations among PTSD symptoms, fear of emotion, and perceived communication difficulties within a dyadic context, we observed a gender difference in the direct association between individuals’ PTSD symptoms and their perception of communication difficulties with their partners. Specifically, men with higher PTSD symptoms reported engaging in higher levels of the self-demand/partner-withdraw communication pattern with their partners, whereas these were not directly linked for women. This pattern of findings is consistent with those observed by Jarnecke et al. (2016) in their community sample of mixed gender couples, in which they found that husbands’ externalizing symptoms predicted husbands’ reports of self-demand/partner-withdraw communication but no association for wives.

There was also a cross-partner association between individuals’ PTSD symptoms and their partners’ fear of emotion that differed by gender. Specifically, women whose male partners had higher levels of PTSD symptoms reported greater fear of emotion, whereas there was no such association for men. Further, there was formal evidence of an indirect association between men’s PTSD symptoms and women’s higher levels of perceived partner-demand/self-withdraw communication when accounting for associations with women’s fear of emotion. Relative to other anxiety-based disorders, anger is a prominent feature of PTSD (Olantunji et al., 2010), and PTSD-related hyperarousal symptoms, which include anger and irritability, are associated with men’s aggression towards their partners (e.g., Taft et al., 2007). As previously noted, men with higher levels of PTSD symptoms reported engaging in more self-demand/partner-withdraw communication, and the man demand-woman withdraw pattern of communication is associated with the perpetration of aggression by men towards their female partners (Jarnecke et al., 2016). As such, this indirect association may indicate that being partnered with a man with higher levels of PTSD symptoms and the emotional arousal experienced in this context are frightening for women and one reason that they withdraw in the face of their partners’ negatively valenced approach behaviors. This interpretation aligns with research on partner accommodation of PTSD symptoms (e.g., colluding in avoidance and safety behaviors, refraining from expressing one’s thoughts and feelings about relationship matters; Fredman et al., 2014; Renshaw et al., 2020; Thompson-Hollands et al., 2024). Given the negative consequences of their male partners’ PTSD symptoms, accommodating these symptoms may help women feel less anxious in the moment (e.g., Fredman et al., 2023) but inadvertently maintain psychological and relational distress over the longer term (Fredman et al., 2022). These findings are also consistent with work demonstrating that associations between one member of a couple’s PTSD symptoms and the other partner’s distress are stronger for women than men (Lambert et al., 2012). Some research in samples consisting primarily of male trauma survivors with female partners has indicated that this may occur via partner accommodation (Fredman et al., 2022), particularly accommodation of trauma survivors’ anger symptoms (Thompson-Hollands et al., 2024). Given that women tend to be more attuned to relationships than men when under stress (Taylor et al., 2000), future research examining gender differences in the putative role of fear of emotion in the linkages among PTSD symptoms in one member of a couple and the other partner’s accommodative behaviors and psychological and relational distress is warranted.

Communication difficulties are a primary reason couples seek marital therapy (Doss et al., 2004). Findings from this study suggest that PTSD, a modifiable individual difference variable, may increase couples’ risk for communication difficulties directly for men and indirectly through fear of emotion for both men and women. Accordingly, they provide several clinical implications for the assessment and treatment of couples. With respect to assessment, when couples with PTSD present for treatment, clinicians should assess PTSD symptoms in both members of the couple, along with associated features such as fear of emotion. Further, in addition to the common practice of having couples engage in a communication task in which the therapist observes the couples discuss a relationship problem, findings from this study suggest that clinicians should also assess couples’ perceptions of their communication (e.g., by self-report), especially demand-withdraw communication. With respect to treatment, disorder-specific couple therapies for PTSD that explicitly target fear of emotion and provide instruction in communication and problem-solving skills are particularly well suited for treating the disorder and associated communication difficulties (e.g., Monson & Fredman, 2012; Morland et al., 2022; Sautter et al., 2015). There may also be cases in which couples present for general couple therapy for the treatment of relationship distress and one partner has clinically elevated symptoms of PTSD but a disorder-specific couple therapy for PTSD is not indicated. If so, the partner with PTSD could be referred for evidence-based individual psychotherapy for PTSD (e.g., Foa et al., 2019; Resick et al., 2023), and evidence-based general couple therapies (e.g., Baucom et al., 2020; Christensen et al., 2014; Johnson et al., 2023) could be used concurrently to address fear of emotion and maladaptive communication patterns.

This study was designed to investigate questions about how PTSD symptoms relate to generalized beliefs about the dangerousness of one’s emotions and how these beliefs relate to individuals’ perceptions of communication difficulties with their partners, which likely represent impressions based on accumulated experiences over months to years. This ordering of variables was chosen for several reasons. First, a prior study using an autoregressive cross lagged design to disentangle directionality between PTSD symptoms and couple communication difficulties over a three-month period (Fredman et al., 2017) demonstrated that, at least during the acute posttrauma period, total PTSD symptoms drive self-reported couple communication difficulties rather than vice versa. Second, cognitive-behavioral models of PTSD (e.g., Ehlers & Clark, 2000; Foa & Kozak, 1986) posit that the psychopathology of the disorder – in particular, maladaptive appraisal of emotional arousal as trauma cue -- contributes to fear of one’s emotions and that such appraisals are maintained in the absence of corrective information resulting from behavioral and experiential avoidance. Considered within the context of romantic relationships, avoidance symptoms account for the prospective associations between PTSD symptoms and perceived couple communication difficulties (Fredman et al., 2017). Third, these patterns of associations among individuals’ trait-like characteristics converge with findings from within-person micro timescale emotion dynamics during couples’ conflict conversations. In prior work with this sample (Fredman et al., 2023), we have shown that, during laboratory-based conflict discussions with their romantic partners, individuals with higher PTSD symptoms (particularly reexperiencing and avoidance symptoms) had greater difficulty regulating their momentary emotional arousal, potentially because emotional arousal itself can serve as a trauma reminder. If individuals with higher levels of PTSD indeed fear their emotional arousal in the moment during conflict with their partners and they hold generalized beliefs about the dangerousness of their own emotions, they may also find it harder to communicate constructively with their partners around areas of relationship disagreement in general. Because this study used cross-sectional data, directionality cannot be formally discerned or asserted. That said, we are not aware of any models proposing an alternative directionality.

The between-person approach used in the current study answers questions regarding how people with elevated PTSD symptoms differ in their perception of communication difficulties with their partners. To answer questions about mediation at the within-person level (e.g., whether, within a given individual, their PTSD symptoms cause them to fear their own emotions, and this, in turn, causes them to perceive communication difficulties with their partner), future studies would benefit from repeated measures data analyzed with the random-intercept cross lagged panel model (e.g., Hamaker et al., 2015). This would permit the investigation of within-person mediation processes having accounted for trait-like, between-person associations among variables (e.g., Fredman et al., 2022). Within-person mediation could also be tested in an intervention context using a randomized controlled trial or in another context in which fear of emotion is experimentally manipulated.

There may be limitations regarding the generalizability of the study’s findings. This sample consisted of couples selected to match the demographics of the surrounding semi-rural area in which the data were collected and, as a result, the sample was largely non-Hispanic White. This sample also consisted of mixed gender couples, as the gender balanced design was used in the parent study to examine gender differences in the associations between PTSD and couple functioning. Although methodological strengths in some respects, it also means that the pattern of findings observed in this sample may not generalize to same sex or gender diverse couples or couples living in more urban areas. To increase generalizability, future studies should include couples who are more diverse with respect to race, socio-economic status, gender composition, and place of residence. This could allow for examination of other factors (e.g., gendered socialization of emotions) that may influence associations among these constructs. Further, although we do not anticipate that the nature of the associations among PTSD symptoms, fear of emotion, and perceived couple communication difficulties would differ using total PTSD symptoms assessed under DSM-5 (American Psychological Association, 2013) rather than DSM-IV (American Psychological Association, 1994), replication of study results with DSM-5-asessed PTSD symptoms would add further confidence in the pattern of findings.

In conclusion, this was the first study to empirically demonstrate that fear of emotion exists within an interpersonal context among couples grappling with PTSD, with implications for how couples attempt to resolve relationship conflict. Findings demonstrate that individuals with higher PTSD symptoms may perceive, and experience, more difficulties resolving relationship conflict with their partners by virtue of being people who have greater fear of their emotions. In addition, men with higher PTSD symptoms tend to be those in relationships characterized by more demanding behaviors from men and more withdrawing behaviors from women, with gender differences in how fear of emotion predicts these between-person differences. Couple-based treatments for PTSD that promote emotional tolerance and attend to gender differences in how PTSD symptoms relate to individuals’ and partners’ perceptions of demand-withdraw communication patterns may improve romantic relationship functioning among trauma survivors and increase the potential for relationships to serve as a conduit for recovery from the disorder.

• Individuals with higher PTSD symptoms report greater fear of their emotions

• Those with greater fear of emotion report more dysfunctional couple communication

• Women whose male partners have higher PTSD symptoms report more fear of emotion

• Men with higher PTSD symptoms report more self-demand/partner-withdraw communication