War-related Trauma : Increasing the American GI ’ s Resilience through Marriage

Studies have shown PTSD has a negative impact on close relationships among Vietnam War veterans. Recently, studies have replicated these findings in the Operation Iraqi Freedom (OIF) and Operation Enduring Freedom (OEF) cohort. Currently, over half of the military is married and veterans are returning from combat with elevated rates of PTSD. Thus, investigating which symptom clusters influence marital satisfaction of the veteran the most is important for assisting social workers and other mental health professionals in identifying and prioritizing treatment goals. The current study identifies which of the four PTSD symptom clusters impacts marital satisfaction the most in returning combat veterans using regression analysis. The emotional numbing cluster negatively impacted marital satisfaction whereas the hyper-arousal cluster positively impacted it. Using all 17 Post-traumatic Disorder Checklist-Military (PCL-M) questions as possible predictors of veterans’ marital satisfaction, regression analysis revealed five of the questions account for 26 percent of the variance in marital satisfaction. Clinical implications and recommendations are explored.


INTRODUCTION
When veterans return from Operation Iraqi Freedom (OIF) or Operation Enduring Freedom (OEF), they may come back with experiences of war that often lead to the development of post-traumatic stress disorder (PTSD) (Seal, et al., 2008).For purposes of this article, veteran(s) is an all inclusive term that refers to Active Duty personnel and those who have separated from their respective military component who have been deployed in support of Operation Iraqi Freedom (OIF) and/or Operation Enduring Freedom (OEF).To be diagnosed with PTSD, according to the Diagnostic and Statistical Manual for Mental Disorders (DSM-IV TR) (American Psychiatric Association [APA], 2000), a person must meet several requirements.The first criterion is that: The person has been exposed to a traumatic event in which both of the following were present: (1) the person experienced, witnessed or was confronted with an event or events that involved actual or threatened death or serious injury, or a threat to the physical integrity of self or others (2) the person's response involved intense fear, helplessness or horror (p.467).
The person must also have symptoms from three clusters present for longer than one month: (B) re-experiencing, (C) avoidance/emotional numbing, and (D) hyper-arousal.Re-experiencing symptoms are intrusive memories about the trauma; nightmares; feeling as though the trauma is reoccurring; intense psychological distress; and physiological reactivity.Avoidance/emotional numbing symptoms are efforts to avoid thoughts, activities, and memories of the trauma; decreased interest in activities; emotional detachment; restricted affect; and a sense of a foreshortened future.Hyper-arousal symptoms are characterized as difficulty falling asleep; irritability; difficulty concentrating; hyper-vigilance; and an exaggerated startle response.The last criterion for PTSD is "the disturbance causes clinically significant distress or impairment in social, occupational, or other important areas of functioning" (p.468).Seal et al. (2008) report that about 21.8% of returning OIF and OEF veterans have PTSD.When a veteran is married-over half of the US military Active Duty population is (Defense Manpower Data Center [DMDC], 2008)-PTSD not only affects the veteran but also the marital dyad and the entire family system (Monson, Taft, & Fredman, 2009).Research shows secondary/vicarious traumatization, caregiver burden, ambiguous loss, and intergenerational transmission can develop in close relationships when an individual exhibits PTSD symptoms (Monson, Fredman, & Dekel, 2010).Maintaining and strengthening military marriages is integral in protecting veterans and their families from negative outcomes additional to PTSD including intimate partner violence, divorce, and suicide.The primary purpose of this study was to identify which post-traumatic stress symptom clusters impact marital satisfaction the most among returning OIF/OEF veterans to assist social workers and other mental health professionals in prioritizing treatment goals.

LITERATURE REVIEW
Researchers have been studying adaptation in close relationships after trauma for some time (e.g., Hill, 1949;Lavee, McCubbin, & Patterson, 1985;McCubbin, & McCubbin, 1991;McCubbin & Patterson, 1982;Segal, 1986).Whisman and Beach (2010) assert "studies have suggested that marital satisfaction is lower among people with psychiatric disorders, including anxiety disorders" (p.17).The majority of what is known about PTSD and close relationships is derived from research conducted on Vietnam veterans and their partners (Monson, et al., 2010).For example, Jordan et al. (1992) examined data from the National Vietnam Veterans Readjustment Survey (NVVRS) and concluded approximately 60% who met PTSD criteria reported medium-high to high levels of marital issues and poor family adjustment.Additionally, significant others of Vietnam veterans reported less satisfaction in their life (Jordan, et al., 1992).Finally, researchers (Jordan, et al., 1992;Kulka, et al., 1990) report male Vietnam veterans diagnosed with PTSD were more likely to experience divorce than those not diagnosed with PTSD.However, McLeland, Sutton, and Schumm (2008) caution divorce rates may not be the best way to determine if deployments cause marital dissolution and suggest "using marital or relationship-satisfaction scores may be a more sensitive way to assess the immediate effect of predeployment or deployment stress on close relationships" (p.842).
Literature has shown the more severe the PTSD experienced by the male veteran, the lower the marital satisfaction (Allen, Rhoades, Stanley, & Markman, 2010;Renshaw, Rodrigues, & Jones, 2008).PTSD can have a deleterious impact on a close relationship with an ultimate culmination being physical aggression as Stith, Green, Smith, and Ward (2008) concluded in a meta-analysis investigating marital satisfaction and marital discord as precursors to intimate partner violence.Their results indicate "that decreased marital satisfaction and increased marital conflict are positively associated with physical aggression in intimate relationships" (p.158).

Theoretical Framework Explaining the Role of PTSD in Marital Satisfaction
There are two main theories of PTSD in an interpersonal context: the Couple's Adaptation to Traumatic Stress Model (CATS Model) (Nelson Goff & Smith, 2005) and cognitive behavioral interpersonal theory (Monson, et al., 2010).The CATS Model shows how individuals and/or couples function and cope after a traumatic event; for this study, the traumatic event is combat exposure of the veteran and the long-term effects that exposure has on the veteran and the veteran's marriage.The CATS Model posits that how the couple copes with the trauma is contingent on three variables: predisposing factors/resources, level of functioning in each partner, and the couple's functioning (Monson, et al., 2010;Nelson Goff & Smith, 2005).Monson and colleagues (2010) have also illustrated that cognitive behavioral theory explains relationship issues where PTSD is present.Monson et al. (2010) assert domains like safety, trust, power, esteem, and intimacy can be impacted by a trauma, thereby affecting close interpersonal relationships.Essentially, as the veteran manifests PTSD symptoms at the individual level in the cognitive, behavioral, or affective realm, the same symptoms can be seen at the dyadic level.Both theories highlight the interpersonal exchange of PTSD in close relationships, yet leave the question: are certain symptoms more deleterious to the couple than others?

The Role of Individual PTSD Clusters in Close Relationships
Prior to 1998, there were few studies examining the role individual PTSD clusters play in close relationships (Riggs, Byrne, Weathers, & Litz, 1998).Riggs et al. (1998) were able to conclude each PTSD cluster can uniquely impact the interpersonal dyad.For example, if a veteran is re-experiencing trauma via nightmares, the spouse might not be inclined to sleep in the same room, possibly resulting in further isolation and emotional separation which could lead to the marriage deteriorating.Similarly, if a veteran is avoiding situations or becoming emotionally numb, this could lead the veteran to become less intimate with the spouse as has been confirmed in current research (Nunnink, Goldwaserf, Afari, Nievergeltf, & Baker, 2010;Solomon, Dekel, & Zerach, 2008).Riggs et al. (1998) concluded that the emotional numbing was the only significant predictor of relationship distress among Vietnam veterans diagnosed with PTSD.However, in their limitations, they caution their results may not be generalizable to other traumatized populations, which can include different war cohorts such as OIF/OEF.

PURPOSE OF THE STUDY
From an interpersonal therapeutic context, identifying which clusters are creating the most impairment for OIF and OEF veterans in relation to their marriages-a key source of support for them-is of paramount concern to assist mental health professionals in prioritizing treatment goals such as strengthening military marriages.The primary purpose of this study was to fill the gap in previous research by identifying which PTSD clusters impact marital satisfaction among returning OIF/OEF veterans.The hypotheses that guided this study included:  H1) Veterans who score above 50 on the Post Traumatic Stress Disorder Checklist-Military (PCL-M) will have lower marital satisfaction than veterans who score 49 or below on the PCL-M. H2) Emotional numbing as measured by the PCL-M will be the strongest predictor of OIF/OEF veterans' marital satisfaction among the four different clusters. H3) A model exists to predict RAS scores based on individual PCL-M questions.

METHOD Procedure for Data Collection and Analysis
This exploratory study used an anonymous survey to investigate variables that influence marital satisfaction among OIF and OEF veterans.All procedures were approved by the University of Texas at Arlington Institutional Review Board.These authors contacted veteran service organizations to inquire if they would allow us to collect data from their members.After written permission was obtained from the organizations, these authors posted the survey to their private discussion boards, chat rooms, and via email.Only organizations that verified OIF or OEF service via Department of Defense Form 214 were allowed to participate in this study.Data were collected for approximately forty days and included demographic questions, the Relationship Assessment Scale, the Combat Exposure Scale and the Post-traumatic Stress Disorder Checklist-Military version.All data were analyzed using the Statistical Package for Social Sciences (SPSS) version 18.0; statistical significance was assessed at the .05level.

Relationship Assessment Scale (RAS)
The Relationship Assessment Scale (RAS) is a seven-item measure that assesses satisfaction in close relationships (Hendrick, 1988).The cumulative scores range from one to five.Hendrick, Dicke, and Hendrick (1998) report "Scores over 4.0 would likely indicate non-distressed partners, whereas scores closer to 3.5 for men and between 3.5 and 3.0 for women would indicate greater relationship distress and possibly substantial relationship dissatisfaction" (p.141).The RAS has good discriminant validity and has been found to have good convergent validity with both the Dyadic Adjustment Scale (DAS) (.80 and .88)and the Kansas Marital Satisfaction Scale (KMSS) (.64 for men and .74for women) (Hendrick, et al., 1998).

Combat Exposure Scale (CES)
The Combat Exposure Scale (CES) was developed by Keane et al. (1989) and is a seven-item self-report instrument designed to measure level of combat exposure.The items are rated from one to five and measure extent and severity of active combat experiences using a score weighted for the severity of each item.Scores can range from 0 to 41 with a higher score indicating heavier exposure.The test-retest reliability of the CES is .97(Keane, et al., 1989) showing excellent stability.For internal consistency, Cronbach's Alpha was calculated for the CES to be .85(Keane, et al., 1989).Additionally, Keane et al. (1989) concluded that the CES has good discriminant validity.

Post-traumatic Stress Disorder Checklist Military (PCL-M)
There are several versions of the PTSD Checklist (PCL) that can be modified to different populations including civilians, spouses, and military personnel.The PTSD Checklist-Military (PCL-M) is a Likert scale standardized assessment instrument with 17 items assessing PTSD symptomology derived from the Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (APA, 1994).Each item relates to criteria for the PTSD diagnosis.Pratt, Brief, and Keane (2006) examined the PCL-M and concluded that it has good convergent and discriminant validity along with high internal consistency and test-retest reliability.The PCL-M is used to identify persons within the military population who might have a diagnosis of PTSD, requiring further clinical assessment for confirmation of the diagnosis (Weathers, Litz, Herman, Huska, & Keane, 1993).
The DSM first recognized PTSD as a mental disorder in 1980 (APA, 1980) and currently has three symptom clusters (re-experiencing, avoidance/emotional numbing, and hyper-arousal) (APA, 2000).For some time researchers have recognized the avoidance and emotional numbing features of PTSD as separate and distinct (Asmundson, Stapleton, & Taylor, 2004;Foa, Riggs, & Gershuny, 1995;Riggs, et al. 1998).For the purposes of this study, the avoidance/ emotional numbing traits are separated into two groups by generating means for each PCL-M question and coding them into the following clusters: questions 1-5 (re-experiencing), 6-7 (avoidance), 9-11 (emotional numbing), and 13-17 (hyper-arousal), thereby creating cluster means.The decision to code PCL-M questions into these clusters is in replication of Riggs et al.'s (1998) coding strategy and is recommended by staff from the National Center for PTSD (B. Litz, personal communication, February 16, 2011).

Participants
The sample consisted of 119 participants.All were married and had been deployed to OIF or OEF at least once.One hundred and one were males (84.9%) and eighteen were females (15.1%).The majority of respondents were White (n = 107, 89.9%).Four other categories accounted for the remaining 10.1% (Hispanic, n = 4; Other, n = 4; More than one race, n = 3; and African American, n = 1).Because the majority of the participants were White males, separate analyses based on race and sex would have lacked statistical power so no sub-group analyses were conducted.Length of marriage ranged from 1 to 37 years with a mean of 9.18 years (SD = 7.96).Of the respondents, 56 (47.1%) had been deployed once; 38 (31.9%) twice; 14 (11.8%)three times; and 11 (9.2%) four or more times to OIF/OEF.Length of deployment was collapsed into total months deployed.The minimum was two months and the maximum was 39 months.The mean was 16.58 months (SD=8.52).Five respondents did not complete this question.Ninety-five (79.8%) of the respondents had been stateside more than one year, the rest had been home one year or less and two did not complete this question.

Assessment Instruments
All 119 participants completed the RAS, CES and PCL-M.The distributions of these scales all approached normality.For the RAS, the minimum score was one and the maximum was five; the mean RAS score was 3.65 (SD = 1.02).The data suggest that eighty-one participants (68.1%) are likely not martially distressed (3.5 or higher) and thirty-eight participants (31.9%) are likely distressed in their marital relationships (3.49 or lower).The mean CES score was 16.46 (SD= 9.9) indicating light to moderate combat exposure (Keane et al., 1989) with scores ranging from 0 to 41; 48% of the respondents reported moderate or high combat exposure.The minimum score for the PCL-M was 17 and the maximum was 85, the mean PCL-M score was 43 (SD= 17.61).Forty-five participants (37.8%) had a PCL-M score of 50 or greater suggesting a PTSD diagnosis, requiring further clinical assessment for confirmation of a PTSD diagnosis.

Hypothesis 1
The first hypothesis was veterans who score above 50 on the PCL-M will have lower marital satisfaction than veterans who score 49 or below; this was tested using t-tests.The mean PCL-M score for those 50 and above was 61.46 (SD = 10.13) and the 49 and below group's mean PCL-M score was 31.81 (SD = 9.56).The mean marital satisfaction score for the 50 and above cohort was 3.45 (SD= 1.14), suggesting greater marital distress than the 49 and lower group whose mean RAS score was 3.78 (SD= .92).Homogeneity of variance was assessed using Levene's Test.Equal variances were not assumed (F (78.6) = 4.41, p< .05).The t-test indicated a rejection of hypothesis 1 since there was no statistically significant difference in marital satisfaction between veterans who had a clinical cutoff on the PCL-M of a score above 50 versus those who scored 49 or below (t (78.6) = -1.64,p=.11).

Hypothesis 2
The second hypothesis was that the emotional numbing cluster as measured by the PCL-M will be the strongest predictor of OIF/OEF veterans' marital satisfaction among the four different clusters.A regression model using the enter method with RAS scores as the dependent variable and each PTSD symptom cluster mean as the four independent variables was conducted.The regression model was statistically significant (F (4) =5.75, p<.001) and had an Adjusted R squared value of .14, which accounts for 14% of the RAS variability.The emotional numbing and hyper-arousal clusters were both statistically significant predictors of marital satisfaction.The beta weight of the emotional numbing cluster was larger (ß = -.73)than the hyper-arousal cluster (ß = .49),thus supporting hypothesis two.The emotional numbing beta weight suggests that as emotional numbing increases, marital satisfaction decreases.The hyper-arousal beta weight can be interpreted that as hyper-arousal increases, so too does marital satisfaction.Please refer to Table 1.

Hypothesis 3
The third hypothesis is a model exists to predict RAS scores based on individual PCL-M questions.Due to the results of hypothesis 1, all participants were included in the testing of this hypothesis as opposed to the original plan of only including participants who scored 50 or higher on the PCL-M.
A backward stepwise regression model with all 17 PCL-M questions as predictors of marital satisfaction was conducted.The statistically significant model (F (5) =9.07, p<.001) had an Adjusted R squared value of .26,which accounts for 26% of the RAS variability.Five PCL-M questions (10, 11, 14, 15, and 16) were statistically significant: two relating to emotional numbing and three to hyper-arousal.The beta weights for feeling distant from others (ß = -.29),emotionally numb (ß = -.44), and irritable or angry (ß = -.27),suggest inverse relationships with marital satisfaction-as each increases, marital satisfaction decreases.The beta weights for difficulty concentrating (ß = .35)and being alert or watchful (ß = .44)can be interpreted that as these increase, so too does marital satisfaction.Please refer to Table 2 for a complete listing.

Limitations
Data were not gathered from the spouse's point of view; this was done to ensure anonymity to the veteran.If data had been gathered from the spouse, anonymity could not have been ensured.Future studies should aim to collect data from the spouse and service member.Another limitation is that data for this article were gathered at one point in time (post deployment).Data should optimally be gathered pre, during, and post deployment in an attempt to establish a baseline level of marital satisfaction and assess change over time, especially given the CATS model of PTSD in an interpersonal context where the couples' predisposing factors/resources, level of functioning in each partner, and the couple's functioning impact the dyad (Monson, et al., 2010;Nelson Goff & Smith, 2005).

CONCLUSIONS AND IMPLICATIONS FOR SOCIAL WORK PRACTICE
This study is a significant contribution to the growing body of literature on how PTSD among returning OIF/OEF combat veterans affects interpersonal functioning.Almost thirty-eight percent of veterans in the current study had a score of 50 or higher on the PCL-M suggesting a PTSD diagnosis requiring further clinical assessment for confirmation of the diagnosis; this is considerably higher than previous research (e.g.Hoge, et al., 2004;Seal et al., 2008).Research investigating PTSD has generally analyzed PTSD symptoms using the three clusters set forth by the DSM-IV (TR).However, recent research has concluded that the avoidance/emotional numbing cluster is in fact two distinct clusters (Asmundson, et al., 2004;Foa, et al., 1995;Riggs, et al. 1998).This study assesses the impact of PTSD, conceptualized using the four cluster organization of symptoms, on veterans' marital satisfaction.The goal of this study is to assist social workers and other mental health professionals in prioritizing treatment goals related to the four clusters so as to increase veterans' marital satisfaction quickly, equipping the veteran with an important resource-the marital relationship-for support and healing while decreasing the likelihood of intimate partner violence, divorce, and suicide.

Conclusions
Our first conclusion is that, though it has been supported in the literature, veterans with PTSD have more problems in their marriages than their fellow, trauma-exposed veterans without PTSD (Monson, et al., 2009), the results of this study do not support this finding.In our sample 37.8% scored a 50 or above on the PCL-M, suggesting that a PTSD diagnosis and marital satisfaction scores among those with PTSD were not statistically different than those for the veterans who scored 49 or below on the PCL-M.One possible reason for this is the moderate combat exposure reported by this sample.Those participants with a score of 50 or greater on the PCL-M on average endorse moderate combat exposure.Their mean score on the CES was 20.51 (SD= 9.86) whereas those with a score of 49 or below on the PCL-M reported light to moderate combat exposure (M=14; SD= 9.15).Renshaw et al. (2009) showed combat exposure was directly related to PTSD, which was related to lower marital satisfaction.Thus, since the combat exposure on average was moderate, it could account for the results not being significant.
Our second conclusion is that emotional numbing and hyper-arousal are the two clusters with the most impact on veterans' marital satisfaction.This differs from previous research (Riggs, et al., 1998) where the avoidance/emotional numbing cluster was the only significant predictor of relationship distress among Vietnam veterans.The data of the present study shows that as the veterans' emotional numbing increases, the veterans' marital satisfaction decreases; for the hyper-arousal cluster, as hyper-arousal increases, so too does the veterans' marital satisfaction.
Finally, our third conclusion is that certain aspects of the emotional numbing and hyper-arousal clusters are more influential on marital satisfaction than others as indicated in the regression analysis of individual questions on the PCL-M.Specifically, the emotional numbing cluster components of feeling "distant or cut off" from others and emotionally numb or "being unable to have loving feeling for those close to you," and feeling "irritable" or having "angry outbursts" all had inverse relationships with marital satisfaction meaning the less severe these components, the higher the marital satisfaction.For the "difficulty concentrating," and being "watchful" components of the hyper-arousal cluster, the more severe these are, the higher the marital satisfaction.Together, these five components explained a little more than a quarter of the variance in marital satisfaction with the remaining items in emotional numbing and two questions regarding hyperarousal not being significant to marital satisfaction.

Implications for Practice and Future Research
The findings of this article are important to social workers and other front line mental health providers who treat OIF/OEF combat veterans and their family members.Even if a veteran does not meet full diagnostic criteria for PTSD, it is important for a clinician to assess for sub-clinical symptoms.The PCL-M is widely used and could easily be adapted for clinical practice with couples.It is recommended that the clinician administer the PCL-M to the veteran and a modified version to the spouse upon intake.The PCL-M can easily be modified to obtain the spouse's perceptions of the veteran's post-traumatic stress symptoms; this has been done previously in a research setting (Renshaw, et al., 2008) and could easily be adapted for a clinical setting.This would allow for a more holistic picture of PTSD in the couple's interpersonal relationship.Using the results of our PCL-M regression model, the clinician could look at the individual PCL-M questions, especially the 5 that were significant in this study, and be provided with a quick overview of the couple's presenting issues.
A clinician can use these findings to quickly hone in on and prioritize areas for intervention; this is especially important for veterans seeking couple's therapy because of the extensive impact that the emotional numbing and hyper-arousal clusters have on marital satisfaction.In work with veterans prior to OIF/OEF, behavioral/cognitivebehavioral therapy has been the most effective approach to working through couplerelated issues (Monson, et al., 2009).However, current research shows that effectiveness can be achieved using a generic approach or by using a more PTSD-focused approach (Monson, et al., 2009) leaving the clinician in a quandary as to which modality to embrace.Thus, one example of how the findings of this study assist in prioritizing treatment goals relates to using behavioral/cognitive-behavioral therapy while enhancing communication when the veteran is exhibiting emotional numbing.When emotional numbing is present, communication or quality of communication between spouses decreases leading to stress, in turn resulting in lower marital satisfaction.Thus, a primary social work implication is that when emotional numbing is present, a couple's counseling treatment goal should be increasing quality of communication to maintain or increase marital satisfaction.This has been found to be an effective strategy with Vietnam veterans (Cahoon, 1984;Sweany, 1987).
Two additional social work implications from our research relate to the effects of hyper-arousal.Hyper-arousal includes irritability and angry outbursts, difficulty concentrating, and increased watchfulness; this study supports that two of these aspectsirritability and angry outbursts-have an inverse relationship with marital satisfaction.The authors posit that the irritability and angry outbursts may be perceived as a threat by the spouse who shuts down and withdraws.Thus, social work interventions with couples should focus on anger management.Contrastingly, our findings support that the other aspects of hyper-arousal-decreased concentration and increased watchfulness (items 15 and 16 on the PCL-M)-increase marital satisfaction, perhaps eliciting more empathy from the spouse.Specifically decreased concentration by the veteran might not be viewed as a threat thereby allowing the spouse to assist without threat of harm.Being watchful or on guard has the same possibilities for eliciting empathy from the spouse without the spouse feeling threatened.Clinically, social workers may be wise to prioritize emotional numbing, anger and irritability over increasing concentration and decreasing watchfulness given the results of this study.

Table 2 . Prediction of Marital Satisfaction by Post Traumatic Stress Disorder
a Relates to the emotional numbing cluster.b Relates to the hyper-arousal cluster.