Introduction

Intimate partner violence (IPV), a pervasive problem affecting millions of people every day in the United States, has multiple definitions, but can generally be defined as assault, physical, sexual or psychological abuse, by a current romantic partner or former spouse (Wilson et al. 2012). It is estimated that a staggering one in three women (35.6%) and one in four men (28.5%) have experienced physical abuse, rape, and/or stalking by an intimate partner in their lifetime within the United States (Black et al. 2011). Research has found that almost 24% of young adults, between the ages of 18–28, in relationships have been exposed to some form of violence (Whitaker et al. 2007). Moreover, among couples seeking outpatient treatment, it is estimated that 36–58% couples experience male to female physical assault in the past year, while 37–57% of couples have experienced female to male physical assault (Jose and O’Leary 2009). Further, while only 34% of individuals injured by intimate partners seek medical care (National Coalition Against Domestic Violence, n.d), IPV lifetime costs are $3.6 trillion to the United States population alone, making it a significant social problem (Center for Disease Control 2018).

IPV and Individual Symptomology

Research indicates that IPV has multiple impacts on an individual’s mental health and can be a risk factor for IPV victimization and perpetration. In a meta-analysis of risk markers for IPV victimization, Spencer et al. (2019b) found that the strongest ontogenetic risk markers were related to mental health (Spencer et al. 2019b). Specific mental health correlates have been identified for both IPV victimization and perpetration, including depression, anxiety, post-traumatic stress disorder, antisocial personality disorder, and borderline personality disorder for both men and women (Spencer et al. 2019a). Further, other researchers have found that, for men, IPV victimization was associated with increased odds of experiencing a disruptive behavior disorder, a substance use disorder and multiple psychiatric experiences (Afifi et al. 2009). For women, IPV victimization was associated with increased odds of experiencing an anxiety disorder, a disruptive behavior disorder, a substance use disorder, and a psychiatric disorder (Afifi et al. 2009). Moreover, in a study on rural couples, Renner et al. (2014) found that women and men with IPV histories are 2.4 and 3, respectively, times more likely to display depressive symptoms than those without IPV histories.

Interestingly, the association between mental health issues and IPV is not solely for those victimized. Renner et al. (2014) further found that the association between IPV and depressive symptoms to be the most robust not only for women who reported being the victim of both physical and emotional IPV, but also for men who reported perpetrating both physical and emotional IPV (Renner et al. 2014). Additionally, they found an association between perpetration and victimization, in that men who perpetrated physical IPV were more likely to be victims of physical IPV, and women who perpetrated physical IPV were more likely to be victims of both physical and emotional IPV (Renner et al. 2014).

IPV and Relational Functioning

The research on the association between relationship functioning and IPV has generally found that marital satisfaction is lower when IPV is reported, particularly when the violence is severe (Leonard et al. 2014). Further, this results has been replicated when women were the perpetrators of IPV, as well as men (Leonard et al. 2014). The literature has identified several maladaptive relationship dynamics associated with IPV, particularly in the recent literature. Specifically, research has identified that IPV is associated with higher frequency of self and partner contempt (Sommer et al. 2016), verbal conflict, infidelity (Johnson et al. 2015; Kaufman-Parks et al. 2017), jealousy, control (Kaufman-Parks et al. 2017), and dominance (Goussinsky et al. 2017) in relationships. Further when looking at the odds of each of these relationship dynamics, Kaufman-Parks et al. (2017) found that frequency of verbal conflict was the most significant predictor of IPV. Maladaptive communication, specifically, disrespectful communication has additionally been found to be related to increased IPV perpetration (Goussinsky et al. 2017). Contrarily, trust (Johnson et al. 2015; Kaufman-Parks et al. 2017) and commitment (Johnson et al. 2015) in relationships have been found to be protective factors against IPV.

Current Study

While these studies demonstrate a link between IPV and both individual symptomology and relationship functioning, the research on this topic has largely been done using individual perspectives. Few studies have examined the dyadic nature of IPV for two primary reasons. First, there are safety concerns that arise when addressing this sensitive issue conjointly, such that discussions about relational issues could trigger further IPV experiences (Hurless and Cottone 2018; Karakurt et al. 2016; Paananen et al. 2018; Stith et al. 2012; Vall et al. 2016). However, recent research using physiological measures of arousal has demonstrated that victims of IPV may feel more comfortable and not experience as much autonomic arousal when discussing IPV in session as previously thought (Paananen et al. 2018). Second, there is a perception that there is (or should be) a clear perpetrator and victim of IPV, and that research exploring the relational dynamics of IPV might lead to victim blaming, while simultaneously excusing the accountability of the person engaging in the violence (Hurless and Cottone 2018; Karakurt et al. 2016; Stith et al. 2012). Different typologies of IPV have been identified, existing on a continuum of severity, with the most severe IPV characterized by patterns of fear and control, while the least severe often occurring in the context of arguments that has escalated out of control and more frequently presenting as bidirectional (Hurless and Cottone 2018; Karakurt et al. 2016; Stith et al. 2012). It has been consistently identified that severe IPV cannot be treated conjointly; however, several models of conjoint therapy have been developed for mild to moderate IPV, demonstrating that this issue can be addressed from a relational perspective (Hurless and Cottone 2018; Karakurt et al. 2016; Stith et al. 2012). This study adds to the extant relational IPV literature by exploring the relationship between IPV and individual symptomology as well as IPV and relational functioning using dyadic level data from heterosexual couples presenting to treatment at a university community clinic.

Specifically in this study, we hypothesized that the individual symptomology scores would be higher when IPV was reported and that relationship adjustment scores would be lower when IPV was reported.

  • Research Question 1: How does the presence (or absence) of IPV impact the relationship between partners’ self-reported individual symptomology in couple dyads? It is hypothesized that there will not be a significant main effect for individual symptomology, meaning that partners’ self-reported individual symptomology scores will be not be significantly different. Further, it is hypothesized that there will be a significant interaction effect between partners’ self-reported individual symptomology scores and IPV grouping, meaning that partners’ self-reported individual symptomology will be higher in couples who screen positive for IPV than couples who do not screen positive for IPV.

  • Research Question 2: How does the presence (or absence) of IPV impact the relationship between partners’ self-reported relational functioning in couple dyads? It is hypothesized that there will not be a significant main effect for relational functioning, meaning that partners’ self-reported relational functioning scores will be not be significantly different. Further, it is hypothesized that there will be a significant interaction effect between partners’ self-reported relational functioning scores and IPV grouping, meaning that partners’ self-reported relational functioning scores will be lower in couples who screen positive for IPV than couples who do not screen positive for IPV.

  • Research Question 3: How does the presence (or absence) of IPV impact the relationship between partners’ self-reported individual symptomology and relational functioning in couple dyads? It is hypothesized that there will be a significant interaction between partners’ self-reported individual symptomology and relational functioning. It is also hypothesized that there would be a significant interaction between partners’ self-reported individual symptomology and relational functioning based on IPV group, such that the relationship between the variables will be more apparent when IPV is present.

Methods

Participants and Procedure

This current study protocol was reviewed and approved by the institution IRB. Archival data from clients who presented to at least one session of conjoint couple’s therapy for a non-substance use presenting concern between March 2016 and December 2016 were used for the present study. Original data were collected from participants who self-referred for services at a graduate training clinic at a Midwestern University. Both members of the couple presented to the initial session for the purpose of couple therapy and completed written instruments, during which information about relevant variables of interest were queried. Each member of the couple completed the instruments independently of one another using an electronic touch screen tablet (or paper version if the participant was unable to use the tablet). Of note, clinic protocol follows the completion of paper screening devices with a conjoint appointment to verbally assess the relationship, as well as an individual appointment with each partner to verbally assess for safety, including suicidality, homicidality, substance use, IPV, and prior trauma history. If either partner endorses safety concerns, including patterns of fear and control or severe physical IPV, such that one or both partners require medical intervention within the past 30 days, individual therapy rather than conjoint therapy is recommended. If either partner endorses experiences of mild to moderate IPV within the past 6 months, both partners must agree to contract for safety in order to continue conjoint therapy.

For the present study, data were downloaded, de-identified, cleaned, and structured to meet the methodological parameters of the current study. A total of 107 couples sought out services. Only 2 couples identified as participating in a same-sex relationship and were excluded due to the disproportionately small sample size. 72 of the remaining couples had complete data files, which comprised the current sample. No couples were removed from the dataset unless they were missing data.

Measures

Demographics

Respondents self-reported on sex (male or female); race/ethnicity (Caucasian/white; African-American/Black; Asian-American, Latino; Arab American; other; more than one race/ethnicity; prefer not to answer); and highest level of education completed (less than high school, completed high school, some college, associated degree, bachelors degree, graduate degree, or prefer not to answer). Due to the distribution of the data, some of the categories were condensed and are referenced in Table 1.

Table 1 Demographic characteristics by IPV grouping
Full size table

IPV

IPV was assessed using the HITS (Sherin et al. 1998). The acronym stands for: Hurts, Insults, Threatened with Harm, and Screamed at them. The HITS assessment consists of four associated questions assessing for IPV victimization, inquiring “How often does your partner: physically hurt, insult or talk down to you, threatens you with harm, and scream or curse at you”. The scores range from 4 to 20, with a score of 4 identifying that none of the behaviors ever occur. Any score above 10 indicates a positive screen for IPV. The HITS was determined to have good internal reliability (α = .75) for this sample. For the current study, the HITS summary score was calculated for each member and each member was categorized as screening positive or negative. Couples were categorized into two groups: IPV positive (at least one partner screened positive) and IPV negative (neither partner screened positive). For the current sample, in the IPV negative group, scores ranged from 4 to 9, with a mean score of 6.08 (SD = 1.50) for male partners and 5.86 (SD = 1.50) for female partners. In the IPV positive group, scores ranged from 6 to 16, with a mean score of 10.09 (SD = 2.86) for male partners and 10.59 (SD = 2.59) for female partners.

Individual Symptomology

Individual symptomology was measured using the Brief Symptom Inventory (BSI; Derogatis 1993). The BSI was originally developed to help healthcare professionals assess for psychological problems at intake and to monitor progress throughout treatment. It is made of 53 items, ranked on a on a 5-point scale ranging from 0 (not at all) to 4 (extremely) characterizing the intensity of distress during the past 7 days. There are nine symptom subscales and three global indices that can be used to inform treatment. For the current study, the Global Severity Index (GSI) score was used, as it aims to assess overall distress. The GSI score is derived from averaging (based on the number of questions answered) the questions from the nine symptom dimensions plus four additional items not included in any of the dimension scores. This score is then converted to a T score. The clinically significant cutoff score for the GSI, is a T score greater than .63, with lower scores indicating less distress. The GSI was determined to have good internal reliability (α = .97) for this sample. For the current sample, in the IPV absent group, scores ranged from 0 to 3.85, with a mean score of .67 (SD = .66) for male partners and .02 to 2.53, with a mean score of .82 (SD = .62) for female partners. In the IPV present group, scores ranged from 0 to 2.02, with a mean score of .68 (SD = .56) for male partners and .13 to 2.6, with a mean score of 1.02 (SD = .79) for female partners. In comparing these data to the clinical cutoff scores, results that when IPV was present, 45.5% (n = 10) of men and 54.5% (n = 12) of women met the threshold for clinically significant distress, while 42% (n = 21) of male partners and 48% (n = 24) of female partners met the threshold when IPV was absent.

Relational Functioning

Relational functioning was assessed using the Revised Dyadic Adjustment Scale (RDAS; Busby et al. 1995). This assessment consists of 14 questions and aims to measure relationship adjustment. It uses a 6 point Likert, ranging from never occurs/strongly disagree to always occurs/strongly agree, depending on the question. There are three subscales (consensus, satisfaction, and cohesion) within the RDAS. For the current study, the total score was used to assess overall relationship quality. Higher scores equate to greater relational adjustment, with a clinical cutoff score 47 or below indicating relational distress. The RDAS was determined to have good internal reliability (α = .86) for this sample. For the current sample, in the IPV absent group, scores ranged from 18 to 62, with a mean score of 41.65 (SD = 11.12) for male partners and 16 to 63, with a mean score of 42.63 (SD = 10.67) for female partners. In the IPV present group, scores ranged from 29 to 61, with a mean score of 39.00 (SD = 7.85) for male partners and 11 to 48, with a mean score of 32.36 (SD = 10.35) for female partners. When looking at relational distress cutoffs, results showed that 66% (n = 33) of the female partners and 62% (n = 31) of the male partners scored in the distressed range when IPV was absent, while 95% (n = 21) of the female partners and 90.9% (n = 20) of the male partners scored in the distress range when IPV was present.

Data Analysis Plan

First, the dataset was cleaned and data screening procedures were employed. For the purposes of dyadic data analysis, data was categorized into partner (women and men) scores on each item. Next, descriptive statistics were examined to contextualize the sample. To address the research questions, a repeated-measures MANOVA were conducted, with partner as the repeated variable, BSI and RDAS scores for each partner as the dependent variable, and IPV grouping (present vs. absent) as the between subjects factor (Maguire, 1999). This statistical analysis measure was utilized rather than more advanced statistical measures (e.g. APIM) due to the smaller sample size for the research questions analyzed.

Results

Descriptive Statistics

Univariate analyses were used to contextualize the sample by sex and detailed statistics are located in Table 1. Regarding race/ethnicity, the vast majority of the sample self-identified as either Caucasian/white or African American. The majority of the sample reported a household income of $50,000 or less. Most of the sample reported their educational attainment at least some college or above. Sample characteristics can also be found in Table 1 organized by IPV screening group.

Regarding screening positive for IPV, 15 female partners and 15 male partners screened positive on the HITS. Regarding couple IPV screens, 50 (69.4%) couples had both partners screen negative for IPV, 8 couples (11.1%) had both partners screen positive for IPV (concordant), and 14 (19.4%) couples had one partner screen positive for IPV (discordant). Due to the small sample size for positive screens (both concordant and discordant couples), these categories were collapsed for the purposes of the analyses, resulting in 22 (30.6%) couples screening positive for IPV. See Table 2 for detailed statistics regarding IPV groupings.

Table 2 Outcome variables by IPV grouping
Full size table

Research Question 1

The first research question was to examine if there are differences in individual symptomology in couple dyads based on IPV group. It was hypothesized that there would not be a significant main effect for BSI scores, but there would be a significant interaction effect between BSI scores and IPV grouping. Results from the repeated measures MANOVA showed a significant main effect for BSI scores, F (1, 69) = 1320.49, p < 0.01, which does not support the first part of the hypothesis. This suggests that partners’ came into couple therapy with differing levels of individual symptomology, regardless of IPV grouping. More specifically, female partners reported higher levels of individual symptomology than male partners (mean difference = .14), regardless of IPV grouping.

Results from the repeated measures MANOVA also showed a significant interaction between BSI scores and IPV grouping, F (1, 69) = 9.77, p < 0.05, which supports the second part of the hypothesis. This means that the presence of IPV is associated with differences in the relationship between partners’ self-reported individual symptomology. Specifically, female individual symptomology scores were higher when IPV was present than when IPV was absent (mean difference = .20), while there was little discernable difference within male BSI scores between IPV screening groups (mean difference = .01; see Fig. 1). This means that female partners are reporting higher levels of distress when IPV is present than when it is absent. Further, as can be seen in Fig. 1, female partner scored much higher on the BSI than male partners overall, suggesting greater levels of distress.

Fig. 1
figure 1

Individual functioning difference based on IPV screening group

Full size image

Research Question 2

The second research question was to examine if there were differences in relational functioning in couple dyads based on IPV group. It was hypothesized that there would not be a significant main effect for RDAS scores, but there would be a significant interaction effect between RDAS scores and IPV grouping. Results from the repeated measures MANOVA did not show a main effect for RDAS, F (1, 69) = 2.42, p < 0.12, which is consistent with the first part of the hypothesis. This means that partners’ entered couple therapy reporting similar levels of relational functioning, regardless of IPV grouping.

Results from the repeated measures MANOVA additionally showed a significant interaction between RDAS scores and IPV grouping: F (1, 69) = 4.97, p < 0.5, which supports the second part of the hypothesis. Specifically, as can be seen in Fig. 2, a crossover effect can be observed in that female relational functioning score lower than male relational functioning scores when IPV was present than when it was absent. Further, there is a much larger discrepancy between male and female relational functioning scores when IPV is present (mean difference = 6.64) than when it is absent (mean difference = .98; see Fig. 2). Lastly, relational functioning scores for both partners’ are lower when IPV is present, suggesting overall greater levels of relational distress, than when IPV is absent.

Fig. 2
figure 2

Relational functioning difference based on IPV screen grouping

Full size image

Research Question 3

The third research question was to examine the relationship between partners’ self-reported individual symptomology and relational functioning within couple dyads based on IPV group. It was hypothesized that there would be a significant interaction between partners’ BSI scores and RDAS scores. It was also hypothesized that there would be a significant interaction between partners’ BSI scores and RDAS scores based on IPV group. Results from the repeated measures MANOVA did not show a significant interaction between partners’ BSI scores and RDAS scores: F (1, 69) = 3.12, p = .08. This finding does not support the first part of the hypothesis.

Results from the repeated measures ANOVA also showed a significant interaction between partners’ RDAS scores and BSI scores and IPV grouping: F (1, 69) = 5.07, p < 0.05. This finding supports this part of the hypothesis. These results mean that while partners’ self-reported individual symptomology and relational functioning seem to vary independently, when IPV is taken into consideration, the range and standard deviation of the two variables is narrower when IPV is present than when it is absent, making the interrelationship between the two variables observable.

Discussion

This study examined the impact of IPV on individual symptomology and relational functioning in a clinical sample of heterosexual couples. The present study focused on self-reported IPV experiences for couples presenting to therapy. Results showed that several hypotheses were supported, demonstrating that IPV had a significant impact on individual symptomology, relational functioning, and the relationship between individual and relational functioning in couple dyads.

IPV and Individual Symptomology

Several interesting results were noted in the results related to the relationship between individual symptomology and IPV. Specifically, female partners consistently reported higher individual symptomology than the male partners, regardless of IPV grouping. Further, consistent with prior research finding that female partners are at risk for a variety of mental health issues when they experiencing IPV (Afifi et al. 2009; Renner et al. 2014; Spencer et al. 2019a, b), the current study also found that female partners reported elevated individual symptomology when IPV was reported than when it was not reported. Contradictory to prior research that has identified that male partners are also at risk for a variety of mental health symptoms when experiencing IPV (Afifi et al. 2009; Spencer et al. 2019a, b); however, the male partners in the current study did not report elevated individual symptomology when IPV was reported than when it was not reported. One possible explanation for this sex discrepancy is that that male partners who have more internalized acceptance of IPV as an appropriate relationship behavior are less likely to report resulting mental health problems (Kaura and Lohman 2007). This study did not examine acceptance of IPV, so future research should include this as a possible mediating viable in the relational research on IPV.

Another possible explanation is that due to the subject matter of the research, in conjunction with presenting for conjoint therapy and the stigma around IPV perpetration and victimization discussed earlier, particularly for male partners, the male partners who participated in this study may not have felt comfortable sharing their individual symptomology (Stith et al. 2012). Prior research has also identified that male partners are less likely to report individual symptomology regardless, so this study may not have accurately captured the depth of the mental health challenges the male partners experienced (Affleck et al. 2018). It may be that the male partners need to be engaged in conversations on IPV victimization in a different way to allow them to fully utilize the mental health services available to them.

Results also found that reports of individual symptomology were significantly related, regardless of IPV context. From a biopsychosocial model, it can be understood that the individual biological functioning and psychological functioning of the members of the couple impacts the social functioning of the relationship as a whole. Systemically, we understand that together, the impacts are greater than when we look at each individual separately. The results from this study demonstrate that individual symptomology within the context of mild to moderate IPV may be intertwined into the relational functioning. Research should continue to be developed to understand what drivers create and perpetuate this IPV as a relational construct, so that programs can be developed to prevent and mitigate this public health concern.

IPV and Relational Functioning

The results from this study demonstrated that IPV has a negative impact on partners’ perception of relational functioning. Specifically, both partners’ scored lower scores on relational functioning when IPV was reported than when it was not reported; however, the difference was much more robust for female partners, which is consistent with prior research (Afifi et al. 2009; Renner et al. 2014). Further, these results provide additional support to prior research that has demonstrated that IPV is related to relational functioning (Goussinsky et al. 2017; Johnson et al. 2015; Kaufman-Parks et al. 2017), as male and female partners’ relational functioning scores were significantly different when IPV was reported, with little discernable difference in partners’ scores when IPV was not reported. This suggests that when IPV was not reported, partners’ reported similar perceptions of their relationship functioning; however, when IPV was reported, partners’ were no longer in sync with respect to their perception of their relationship functioning. This may be a clear indicator of the need to assess for IPV, when partners’ have vastly differing perceptions of their relational functioning (i.e. consensus, satisfaction, and cohesion).

Implications for Practice/Policy

These results provide support for the need to address not only the intrapersonal impact of IPV, but also the interpersonal impact of IPV. For instance, results showed that almost 1/3 of the sample (30.6%) screened positive for IPV, which is consistent with finding on the general population (Black et al. 2011l). Given the clinical nature of this sample, the authors were surprised that the rates of IPV were not higher as prior research has found IPV to be of increased prevalence within clinical samples (Jose and O’Leary 2009). It becomes less surprising as we tease out the data more. Specifically, only 8 (11.1%) male partners and 9 (12.5%) female partners in this study reported never experiencing any types of violence in their relationship and only partners in 2 (4%) relationships were concordant in their report of never experiencing any IPV within their relationship. Further, when examining individuals who reported IPV, but did not screen positive, it was found that 49 (68.1%) male partners and 48 (66.7%) female partners positively endorsed verbal IPV (i.e. insults or talks down, screams at them). While these reports did not reach a clinical threshold based on the HITS criteria, it is important to still address this type of violence within the couple therapy as it can lead to later physical IPV (Vall et al. 2016), Prior research has demonstrated that couple therapy involving dialogical dialogue may be effective in addressing psychological IPV, while maintaining safety, within couple therapy (Vall et al. 2016).

These results are supportive of the literature’s suggestion that IPV must be assessed with all clients, regardless of whether it is part of the presenting issues for couples (Todahl et al. 2008; Todahl and Walters 2011) and that clinicians must be equipped to navigate the impacts on individuals within the relationship, the couple as a whole, and the therapeutic process (Crnkovic et al. 2000). This routine assessment should follow all appropriate protocols, recognizing that when IPV presents within the context of fear and control that couple therapy is not recommended (Stith et al. 2012).

Couple therapy, when done safely, can help couples with mild to moderate IPV identify the systemic impact of IPV and work toward understanding the others’ perspective, bringing them back into sync (Hurless and Cottone 2018; Karakurt et al. 2016; Stith et al. 2012). Several considerations must made when engaging in couple therapy, such as clear and firm boundaries regarding stopping couple therapy if there are incidents of IPV, contracting for safety, ongoing assessment of conflict, and engaging in specific interventions to assist with de-escalation. Engaging in IPV couple therapy should be thought of in a similar way to trauma therapy, in that therapists help clients develop IPV-resistant interpersonal (i.e. I statements, time-outs, psychoeducation on their specific cycle of violence) and intrapersonal (i.e., coping strategies, self-care, etc.) skills prior to delving into the content of the IPV relationship trauma (Substance Abuse and Mental Health Services Administration 2014).

Limitations and Future Directions

A limitation of this study was the limited sample size due for the measures and analyses used, such that this study was unable to capture these more nuanced relationships. While almost one-third (31.6%) of the sample reported experiencing IPV victimization, the sample size was not large enough to compare the differences between couples who concordantly and discordantly reported IPV. Additionally, the sample for this study was homogenous, in that it was heterosexual couples from one clinical training clinic in an urban area, which decrease the generalizability of these results. Future research should broaden the collection of data to other non-training community samples and multiple sites across a diversified geographical area. Future research should specifically focus on same-sex couples as prior research has demonstrated that there may be nuanced differences in the experience of IPV from that of heterosexual couples (Miller et al. 2000). Further, future research should compare clinical and nonclinical samples because examining how IPV impacts individual and relational functioning in a nonclinical samples will provide additional insights into this issue and further inform clinical practice.

Lastly, this study may have been limited by the manner in which IPV was assessed, as only self-reported perception of IPV victimization was assessed. Future research should examine IPV in more nuanced ways, such that the impacts of both IPV victimization and perpetration on individual symptomology and relational functioning can be further teased out. Despite these limitations, the results from this study adds to the clinical research on IPV by providing clinicians with a better understanding of how IPV can impact individual functioning and relational functioning within couples.