Introduction

Intimate partner violence (IPV) is a serious public health problem, with a recent large-scale survey estimating that one in three women (39.2 million) and over one in four men (31.9 million) in the USA experience physical IPV victimization over their lifetimes [1]. Physical IPV includes any acts of physical aggression directed at one’s intimate partner (e.g., pushing, hitting), and psychological IPV includes attempts to isolate one’s partner from others, denigrating one’s partner, refusing to discuss problems important to one’s partner, and intimidating one’s partner during disagreements, among others. Consequences of IPV can be severe and include physical injury, health conditions stemming from chronic stress, and psychological sequelae such as depression and posttraumatic stress disorder (PTSD) [2,3,4]. In order to support the prevention of IPV, it is crucial to have an understanding of how experiencing trauma can influence one’s risk of using IPV.

A large evidence base highlights trauma exposure and PTSD symptoms as key risk factors for the use of IPV. For example, studies of men presenting to court-mandated IPV interventions have found high rates of trauma exposure, with a range of 78 to 94% reporting exposure to at least one traumatic event, with participants commonly reporting lifetime exposure to multiple different types of trauma [5,6,7]. Additionally, in two of these studies, PTSD symptoms were associated with greater use of IPV [6, 7], whereas one found that PTSD symptoms were associated with reporting less IPV [5]. However, meta-analyses examining the link between PTSD symptoms, IPV, and psychological aggression have found positive relationships with small-to-medium effect sizes [8, 9]. This review will provide an overview of recent (past 3 years) research that helps explain how trauma and PTSD may confer increased risk for IPV, as well as recent developments in trauma-informed IPV intervention and prevention for military veterans.

Social information processing

McFall’s [10] social information processing (SIP) model, later adapted to the study of IPV by Holtzworth-Munroe [11], has been used as a framework for understanding how trauma and PTSD symptoms may influence risk of IPV use. The model describes a series of sequential steps in which incoming social stimuli are converted into responses that may be effective or ineffective. The first stage of the model represents decoding skills, where sensory information is received, perceived, and interpreted. PTSD-related deficits at this stage may include heightened vigilance to threat cues that make escalation of conflict more likely (e.g., interpretation of a partner speaking loudly from across the house as yelling). The second stage of the model involves decision skills, in which a person searches for possible responses, tests the match between various responses and the task demands, selects a response, searches their repertoire for exemplars of the response, and assesses the utility of implementing the response. PTSD-related deficits at this stage may include reduced ability to generate helpful, non-aggressive responses, overestimating the utility of an aggressive response, or faster selection of the first response generated before evaluating other possible responses. The final stage of the model represents encoding skills, which involve executing the response as intended and monitoring the environment for discrepancy between the intended and observed effects (which then involves decoding skills). PTSD-related deficits at this stage may include executing the response not as intended (e.g., body or voice trembling, causing a person to appear more aggressive than intended) or biased perception in monitoring the effects of the response on the environment. McFall’s [10] conceptual model has been more recently reformulated to include more distinct stages, bidirectional relationships between each stage and a central database of social schemas, scripts, and knowledge [12], and the influences of emotion processes [13].

Recent work has examined the connection between PTSD symptoms and IPV using the SIP model. A study by Taft et al. investigated the relationships between PTSD, biases in the decoding stage of social information processing, anger expression, and IPV [14•]. With a sample of 92 male US veterans of the wars in Iraq and Afghanistan, they administered assessments of PTSD symptoms and IPV use, as well as the articulated thoughts in simulated situations (ATSS), a laboratory task that measures cognitions during anger arousal [15]. Participants were asked to listen to audiotaped scenarios, imagine that they are in the scenario described, and verbalize their thoughts and feelings when prompted. Examples of the scenarios include the participant playing a game with a competitive couple and the participant overhearing his partner flirting with a male acquaintance. Responses were later coded for cognitive biases (e.g., making assumptions in the absence of evidence, dichotomous thinking) and hostile attributions (i.e., blaming the cause of an event on the deliberate and hostile intentions of another person). Mediation models were tested for variables that showed significant bivariate associations. Consistent with hypotheses, cognitive biases during the ATSS mediated the relationship between PTSD symptoms and general anger expression. Additionally, hostile attributions made during the ATSS were significantly positively associated with IPV use, although no mediation models were tested with IPV as the outcome. In line with prior research, PTSD symptoms reflecting hyperarousal (e.g., hypervigilance to threat cues) showed the strongest relationship with IPV use. This study provides partial support for the idea that biased perception and interpretation of social information are related to PTSD symptoms, anger expression, and violence.

A study by LaMotte Taft, Weatherill, and Eckhardt with the same sample of veterans examined deficits at the decision stage of McFall’s [10] model and their associations with PTSD symptoms and IPV [16•]. Assessing these deficits involved presenting participants with a series of problematic marital situation vignettes, asking them to describe with what they would say or do in reaction to each situation, and later rating the social competency of these responses according to a manual developed from the initial studies with this paradigm [17, 18]. A competent response was defined as one that would help solve the problem and would prevent similar problems from happening in the future, whereas an incompetent response was defined as one that would not solve the problem and would likely make the situation worse. In partial support of hypotheses, results indicated that decision-stage SIP deficits mediated the relationship between PTSD symptoms and psychological IPV but not physical IPV. Additionally, when entering the different PTSD symptom clusters into a regression predicting these SIP deficits, only emotional numbing emerged as a unique predictor. Although this finding needs replication, it raises the possibility that feelings of detachment and inability to experience positive emotions diminish one’s motivation for and ability to identify and appraise responses intended to promote a positive outcome for the couple. This may be a distinct process from PTSD hyperarousal symptoms interfering with decoding-stage deficits, which past theoretical work has highlighted [19].

A third study from the same sample examined whether the presence of trauma cues potentiated the relationship between PTSD symptoms and aggressive tendencies [20]. Participants were administered versions of the ATSS both before and after the presentation of an audio recording of their personalized trauma narrative taken from an earlier PTSD interview (the order of these ATSS variations was counterbalanced). Aggressive tendencies were measured by counting the number of physical partner aggression articulations (i.e., expressions of physically aggressive intentions toward the partner in the scenarios) and verbal partner aggression articulations (i.e., statements intended to insult or demean the partner in the scenarios) made during the task. Because they depict the participant’s intended action in the scenario, these can be conceptualized as occurring at the end of the decision stage of McFall’s model [10]. Results indicated that having higher PTSD symptoms predicted making greater physical partner aggression articulations only after trauma cue presentation, whereas higher PTSD symptoms predicted making greater verbal partner aggression articulations both before and after trauma cue presentation. These findings offer some insight into how decision-making skills, for those with PTSD symptoms, may be particularly impaired after reminders of a traumatic event. Overall, these studies suggest that PTSD symptoms relate to deficits in both interpretation of social cues and selection of non-aggressive, more helpful responses to these social cues.

Other work suggests that trauma can have particular impact on cognitions around certain themes, distorting related social information processing and contributing to maladaptive anger expression and IPV use. For example, a study by Germain, Kangas, Taylor, and Forbes examined trauma-related cognitions as mediators of the relationships between PTSD symptoms and anger expression variables among treatment-seeking Australian veterans [21]. All variables were assessed via self-report questionnaires. They found that negative beliefs about oneself and the world mediated the relationships between PTSD symptoms and anger expression variables. Specifically, negative beliefs about oneself were associated with aggressive anger expression, unconstructive suppression of angry feelings, and poorer anger control, whereas negative beliefs about the world were only associated with aggressive anger expression. Another study among a community sample of couples found that, for men, the cognitive schema of mistrust mediated the relationships between the number of lifetime trauma exposures and use of physical and psychological IPV [22]. For women, there was a significant relationship between trauma exposure and use of IPV, but this was not mediated by mistrust. Together, these studies indicate that trauma-related disruption of certain cognitive schemas is associated with problematic expression of anger and use of violence.

The role of anger

Anger is an emotional state that appears helpful in understanding and addressing the connection between PTSD and IPV use, as several recent studies have demonstrated. For example, with data from the National Vietnam Veterans Readjustment Study, Novaco and Chemtob found that anger mediated the relationship between PTSD symptoms and use of violence, as well as moderated this relationship such that veterans with PTSD were more violent only if they were also high in anger [23•]. Similarly, Berthelot et al. found that trait anger mediated the relationship between childhood maltreatment, PTSD symptoms, and IPV use among a sample of adults presenting for treatment for sexual dysfunction [24]. A large-scale (N = 2420) survey of US Army soldiers found that the indirect effect of combat exposure on aggression via PTSD symptoms was conditional on trait anger, with this indirect effect present for those with high but not low levels of trait anger [25].

Together, these studies suggest that experiencing frequent anger reactions is a key determinant in the connection between PTSD symptoms and aggression. Additional work has begun to explicate how anger may interact with other cognitive processes in this relationship. For example, among a sample of community heavy drinkers, Massa, Eckhardt, Sprunger, Parrott, and Subramani assessed trauma-related negative cognitions about the world, anger/hostility, rumination (i.e., focused attention on one’s own distress and its possible causes/consequences), and IPV use [26]. They found that anger/hostility mediated the relationship between negative cognitions about the world and IPV use, and that rumination moderated this indirect effect, such that it was strengthened at higher levels of rumination. This finding suggests rumination as a possible treatment target for ending IPV use among those with problematic anger.

Treatment

Recent developments in trauma-informed IPV treatment

The Strength at Home Men’s Program (SAH-M) was developed in response to the empirical literature on trauma-related deficits and the lack of existing IPV interventions that target such deficits [27]. The program is a 12-week intervention consisting of 2-h sessions conducted in a group therapy format by a co-therapist team. Contrasting with common community approaches to IPV intervention, which involve providing psychoeducation to larger groups, the SAH-M groups were restricted to a size of six to eight clients in order to promote its function as intensive group therapy [27]. SAH-M follows a cognitive-behavioral therapy (CBT) framework, while incorporating strategies to target trauma-related SIP deficits and anger reactions that increase risk of violence [27]. Importantly, the program does not frame PTSD as a “cause” of IPV that diminishes personal responsibility for abusive behavior, but rather as a risk factor to be discussed and addressed. Through targeting social information processing mechanisms in a trauma-informed and motivational manner, group members take more responsibility for abusive behavior and are better able to recognize and change problematic patterns in their thinking and behavior, thus reducing their IPV [27].

In session 1 of SAH-M, clients are introduced to the group structure, expectations, and philosophy, as well as brainstorm and discuss the “pros” and “cons” of abusive behavior to increase motivational readiness to change. Session 2 is focused on understanding various forms of abuse and common reactions to trauma that influence relationships, as well as continued goal-setting. Sessions 3 and 4 focus on understanding and addressing the anger response through enhanced self-monitoring and use of “Time-Outs,” in which the client temporarily leaves to cool down before resuming the difficult relationship discussion. Session 5 examines anger-related thinking and trauma’s influence on threat appraisal, and clients brainstorm active coping-thoughts that counter their unhelpful, anger-producing thoughts. This is designed to target decoding skill deficits from the SIP model. Session 6 focuses on the importance of managing stress and matching coping skills to the avoidable or unavoidable nature of the stress. Sessions 7 through 11 address problematic communication styles (e.g., passive or aggressive communication, avoiding expression of one’s feelings) and promote positive communication strategies (e.g., active listening, assertiveness, and emotional expression). This is designed to target decision skill deficits from the SIP model by enhancing generation and proper evaluation of helpful, non-aggressive responses to conflict. Finally, in session 12, clients review the changes they have made, identify barriers to change, and describe goals and strategies for continued change. Throughout treatment, there are practice assignments for clients to complete between sessions [27].

The efficacy of SAH-M was recently evaluated in a randomized controlled trial (RCT) among a sample of 135 veterans/service members [28••]. Participants were randomized to the SAH-M condition or an enhanced treatment as usual (ETAU) condition in which they were provided referrals for other mental health and IPV services. Participants in the ETAU condition were offered enrollment in the SAH-M program after a 6-month period of time. Primary study outcomes included physical and psychological IPV. Collateral data on participants’ abusive behavior use was obtained from 82% of their female partners. These variables were assessed at pre-treatment, posttreatment, and a 3-month follow-up period. With intent-to-treat analysis, they found a significant time-by-condition effect, with the SAH-M showing greater reductions in physical and psychological IPV. Additionally, when looking at four distinct forms of psychological IPV on another measure, they found that the SAH-M condition showed greater reductions in restrictive engulfment behavior (e.g., trying to keep a partner from seeing certain friends or family members). This was the first RCT to support the efficacy of an IPV intervention in a military or veteran population.

A subsequent study analyzed outcomes from this RCT including data from participants that eventually received the SAH-M program after being in the ETAU condition [29]. First, the authors analyzed whether or not changes in IPA from treatment were any different between the original experimental group and the control group that eventually received the program. Findings indicated that the later application of the SAH-M program to the control group was as effective for physical or psychological IPV as the initial application to the experimental group. Additionally, PTSD symptoms strongly predicted physical and psychological IPV. PTSD did not interact with receipt of treatment in predicting IPV, but slightly decreased the treatment effect for psychological IPV. Thus, although PTSD symptoms did not substantially impact the effectiveness of the SAH-M program, further reductions in abuse-related outcomes may be supported by adjunctive treatment of PTSD symptoms.

Alongside SAH-M, the Strength at Home Couples (SAH-C) program was developed to prevent IPV among military couples in which the male veteran/service member had not yet used physical IPV [27]. SAH-C is a 10-week program conducted in a group format with couples. Similar to SAH-M, it is trauma-informed and is designed to prevent IPV through targeting SIP deficits. Treatment phases of SAH-C are largely similar to those of SAH-M, covering the influence of trauma on relationships, conflict management, and communication skills. The main differences from the SAH-M program are that SAH-C focuses somewhat less on anger recognition and management, and focuses more on dyadic exercises [27].

A recent RCT evaluated the efficacy of the SAH-C program in preventing IPV [30••]. The sample was made up of 69 male veteran/service members and their female partners. Participants were randomized to receive either the SAH-C program or a supportive prevention (SP) group. In the SP group with couples, therapists encouraged a mutually supportive environment with a focus on relationship issues and preventing IPV, but offered no structured skill-training that is given in the SAH-C program. Physical and psychological IPV were the primary outcome measures. The SAH-C condition showed significantly higher treatment retention than did the SP condition, with participants almost twice as likely to complete SAH-C. Results indicated that with intent-to-treat analyses, participants in the SAH-C condition showed lower physical and psychological IPV than did participants in the SP condition in all post-baseline assessments. These trends showed small-to-medium effect sizes.

A separate study examined the effectiveness of a variant of SAH-C called the Strength at Home Friends and Family (SAH-F) intervention [31]. Session content closely followed the SAH-C program, but was not limited strictly to romantic relationships. Of the 70 veterans in the study, 72% participated with a spouse/romantic partner, and 28% participated with a friend, relative, or other loved one. This study examined physical and psychological aggression toward the other dyad member across pre-treatment, program completion, and follow-up time points. No control group was used in this study. Findings showed that for both veterans and loved ones, use of psychological aggression significantly decreased from pre-treatment to completion and follow-up. Use of physical aggression did not significantly change from pre-treatment to completion and follow-up. Regarding secondary outcomes, loved ones, but not veterans, reported significant increases in perceived relationship quality following the intervention. Additionally, both veterans and loved ones reported significant reductions in depressive and PTSD symptoms following the intervention. Together, these two studies offer preliminary support for the usefulness of SAH-C in preventing aggression and reducing important risk factors for aggression.

Conclusions and directions for future work

There have been notable advancements in our understanding of the connection between trauma, PTSD symptoms, and the use of abusive behaviors in relationships, as well as the effectiveness of trauma-informed IPV intervention/prevention among military veterans/service members. This is in contrast to the larger IPV intervention literature, which commonly focuses on psychoeducational programs that are centered on power-and-control, and which scarcely employs RCT methodology [32]. However, there are several areas for expansion in future work. Regarding basic research, it would be helpful to further refine assessment tools in order to distinguish between individual processes involved in the SIP model. This could also allow more nuanced analysis of the interaction between these different stages in determining relationship behaviors. Further development of trauma-informed clinical materials for IPV intervention may benefit from addressing additional factors that literature identifies as related to these processes (e.g., rumination). With promising early efficacy research of the SAH-M program among a military sample, it will be helpful to determine how well these findings generalize to civilian populations of court-mandated clients, which also show high rates of trauma exposure [5,6,7]. Though the program is designed to prevent IPV via improvements in social information processing, the mechanisms of the treatment have yet to be shown. Additionally, it will be beneficial to determine whether adding any adjunctive services (e.g., PTSD-focused treatment) enhance treatment gains. More RCT research is also needed to support the efficacy of the SAH-C program to prevent IPV among military members, as it can be more difficult to end the violence in relationships once it begins. Further research in this area has the potential to make a large impact on the way that we approach IPV intervention/prevention and improve abuse-related outcomes.