SOWC 6101-5 DISCUSSION WORKING WITH FAMILIES

ORIGINAL ARTICLE

Post-9/11 Service Era Veterans: Intimate Partner Violence and Substance Use

Roberto Cancio

Department of Sociology, Bellarmine College of Liberal Arts – Loyola Marymount University, Los Angeles, CA, USA

ABSTRACT Using structural equation modeling, this study considers variations of intimate partner vio- lence (IPV) among military families from the point of the perpetrator to test previously established empirical models on military subgroups in order to observe the impact of demo- graphic factors on the type of IPV most prevalent among Post-9/11 military families from the National Longitudinal Study of Adolescent to Adult Health (1994–2008): Waves I and IV in-home interviews (N¼ 499). Study findings indicate that the perpetration of physical and sexual IPV varies across race/ethnicity perpetrator profiles. Models for substance use and IPV patterns were not similar across military cohorts and or racial/ethnic groups.

KEYWORDS IPV; violence; military; veterans

Introduction

Military deployments place service members at high risk for intimate partner violence (IPV) perpetration, as well as place substantial strain on service members and their family (Cancio, 2018). Various studies using representative military samples find that military pop- ulations report greater rates of IPV perpetration than the non-military populations when controlling for age and race/ethnicity (Cancio, 2017; Cancio & Altal, 2019). The increase of military-related calls into the National Violence Hotline (457 in 2006 to over 1100 in 2010), potentially signifies that IPV is getting worse among military families (Women’s Justice Project, 2011). From a military perspective, conflict among service members and their families is closely related to higher levels of health care service use, lower military morale for both families and the service member, poorer job performance across actors, and inflated mission safety risk for service members involved with IPV (Fontana & Rosenheck, 2010). In addition, a var- iety of studies have posited that beyond injury and death, victims of IPV are more likely to report a var- iety of acute and chronic health conditions resulting in hospitalization, disability, and/or death (Black, 2011; Coker et al., 2002). However, among military IPV perpetrators, empirical research has yet to flesh out the various prevalent forms of violence that mani- fest distinctively through different cohorts of mili- tary families.

Background

It is important to understand the dynamic ways in which behavior and substance uses interact. In a study of veterans, Johnson, Giordano, Manning, and Longmore (2015) found that younger age was associ- ated with higher IPV perpetration among substance abuse inpatients. Focusing on race and age differen- tials, studies using representative samples of veterans suggest that the execution of IPV perpetration is more frequent among non-Whites than Whites, and is nega- tively associated with the age of the perpetrators (McCarroll et al., 2003; Rosen et al., 2002). Breiding, Chen, and Black (2014), expand on ethnicity by con- sidering other forms of violence both in outside the family unity. Breiding et al. (2014) find that Native American, Black, and multi-racial Non-Latino men, had a significantly higher prevalence of sexual assault, battery, or stalking compared with White Non-Latino men. Studies focused on IPV and military veterans have found that IPV does not manifest uniformly across groups of men of similar racial/ethnic groups (Cancio, 2015, 2017; Cancio & Altal, 2019).

Using a sample of Pre-9/11 veteran perpetrators, Cancio (2017) found that White veterans were more disposed to perpetuate sexual violence against their partners; meanwhile, Black veterans were more likely to carry out physical IPV. These findings were also reflective of the types of substance used by particular ethnic and racial groups. Cancio (2017) posited that a

CONTACT Roberto Cancio [email protected] Department of Sociology, Bellarmine College of Liberal Arts – Loyola Marymount University, Los Angeles, CA 90045, USA. � 2019 Taylor & Francis Group, LLC

SUBSTANCE USE & MISUSE 2020, VOL. 55, NO. 2, 241–251 https://doi.org/10.1080/10826084.2019.1662812

number of social determinants of mental and physical health, such as limited access to education, economic resources, and social services, likely play important roles in IPV types and perpetration rates. Finally, the stress associated with the financial situation of the family unit may also be related to the perpetration of IPV because higher IPV prevalence has been found among families with combined household income less than $50,000 than for families with a combined income over $75,000 (Breiding et al., 2014). Following the aforementioned literature on demographics, the importance of considering variations based on demo- graphic constructs offers a particular context from which substance use and IPV perpetrators manifest. These demographics also affect the availability and acceptability of substance types and rates of use.

Drugs/alcohol and IPV

The relationship between substance uses is only one of the several important factors that influence the risk for IPV, and vice versa (De Bruijn & De Graaf, 2016). Regardless of the direction of the relationship between substance use and IPV, the co-occurrence of IPV and substance use is substantial across a series of studies (see Ahmadabadi et al., 2019; Cancio, 2017; De Bruijn & De Graaf, 2016; Shorey, Stuart, Mcnulty, & Moore, 2014). Numerous studies have confirmed a strong association connecting substance use and IPV (Moore et al., 2008; Smith, Hormish, Leonard, & Cornelius, 2012). Substance use may also be influenced by other risk factors (e.g., violence in the family) and substance use may also influence other risk factors (e.g., belief that violence is appropriate). Nonetheless, illicit drug use is more correlated to IPV than alcohol (Choenni, Hammink, & van de Mheen, 2017). However, the higher the frequency of inebriation also increases IPV perpetration among men (Shorey et al., 2014). Therefore, regardless of what substance is consumed, higher levels of substance use are related to higher IPV perpetration.

Military service

Occupational-related stress manifests differently between civilian and military occupations. This differ- ence is increased by military stressors (e.g., military deployments, the ambiguity of family relations, com- bat, etc.) (Campbell & Nobel, 2009). From indoctrin- ation, military recruits are socialized to new social norms (i.e., values and norms of masculinity, military vocabulary, a new set of values, etc.) (Cancio, 2015).

Intense physical and psychological stressors stimulate military recruits so that push them to operate under new sets of norms, both emotional and physical. This social and psychological conditioning into military life includes hyperactive vigilance to perceived threats and emotional distancing from others (Grossman, 1995). Other military-related factors (e.g., exposure to com- bat, the number of deployments, and posttraumatic stress disorder) may also influence the use of various substances (Gewirtz, Polusny, DeGarmo, Khaylis, & Erbes, 2010) and IPV connections (Taft, Watkins, Stafford, Street, & Monson, 2011). Under these condi- tions, service members and veterans respond to per- ceived threats of hostility by using military-related reactions (i.e., violence) (Gonzalez, Novaco, Reger, & Gahm, 2016; Novaco & Chemtob, 2015).

Novaco and Chemtob (2015) have posited that vet- erans are more likely to employ an overly hostile interpretation of events and will override their ability to engage in self-monitoring to lower the risk for aggressive reactions. The presence of combat-related PTSD places a service member at a substantially higher risk for experiencing relationship conflict and perpetrating physical and mental IPV (Taft et al., 2011). Having higher PTSD severity is a common antecedent to more severe IPV perpetration (Heyman, Taft, Howard, Macdonald, & Collins, 2012). Furthermore, PTSD symptoms and impulsivity, as seen among combat veteran, can help account for aggressive behavior in the absence of active substance use (Heinz, Makin-Byrd, Blonigen, Reilly, & Timko, 2015). Higher IPV perpetration within military fami- lies has also been linked to PTSD severity and depres- sion among perpetrators (Taft et al., 2011). This suggests an added risk connected with these co-mor- bid diagnoses. This is because depression has been found to co-occur with PTSD (Seal et al., 2009). Existing research suggests that stress and depression among active duty personnel are associated with an increased risk of IPV perpetration. Marshall, Panuzio, and Taft (2005) found higher rates of depressive symptoms among Army IPV perpetrators, compared with those who were non-violent. Moreover, depres- sion severity (Rosen, Kaminski, Parmley, Knudson, & Fancher, 2003) and antisocial personality trait fre- quencies (Rosen et al., 2002) were associated with higher IPV perpetration (Judd, Schettler, Coryell, Akiskal, & Fiedorowicz, 2013). Other personality traits like authoritarianism and attitudes toward women did not distinguish IPV perpetrators from non-aggressive controls in a sample of active duty service members (Marshall et al., 2005).

242 R. CANCIO

Both depression and PTSD interfere with focus and concentration, making it challenging for an IPV per- petrator to effectively respond to offender interven- tions and or consciousness of reality. Hopelessness and helplessness are two conditions of depression. When an IPV perpetrator expresses the belief that there is nothing to lose, and feels jealousy, there is a potential increased risk for an IPV-related homicide (Campbell, 1992).

Service members and veterans experience changes in the ways in which they view the world after being exposed to extreme stress, combat, and possible war- related atrocities, including difficulties with respect to trusting others, developing closeness, and struggles with power and control (Taft et al., 2011). The social and cultural military environment informally allows for behavior against violence to become disinhibited by the context of a military culture of substance use and violent behaviors (Cancio, 2017). According to Cancio and Altal (2019), the military master status (i.e., warrior, hyperactive masculine, and violent) can become mixed-up in a state of inebriation, and acti- vate violently during partner altercations, as personnel are not able to make clear judgments while under the influence of alcohol and drugs. Inebriated partners apply the role of the enemy to their spouse/partner and address the quasi-threat with heightened violence.

As established by previous literature (see Cancio & Altal, 2019), a non-recursive theoretical model (Figure 1) was employed in hopes to test the validity of the model on other military populations and understand how other factors affect types of partner violence among a sample of perpetrators. This article explores the relationship that serving in the military plays for IPV perpetrators, the difference among mili- tary perpetrators to civilian perpetrators, and how

substance use is linked to IPV when comparing any effects of other substances simultaneously.

In an attempt to flesh out the particular IPV and substance use factors among Post-9/11 service veter- ans, this study aimed to: (a) Validate the dimensions of IPV and their indicators using structural equation modeling (SEM),1 (b) Calculate individual coefficients for these dimensions,2 (c) Test the determinants of IPV for each dimension, and (d) Look for differences between military and non-military families.

Methods

The study sample contains information about 449 active duty and veteran male3 participants from the National Longitudinal Study of Adolescent to Adult Health (Add Health) (1994–2008): Wave I and IV In- Home Interview in 2008 (Harris et al., 2009). Add Health is a longitudinal study of a nationally represen- tative sample following initial adolescents, between grades 7 and 12 in the United States during the 1994–95 school years. Participants have been followed through four in-home interviews, the most recent in 2008 when the sample was aged 24–32. This study used Wave I and IV data collection, conducted in 2008, includes in-home interviews with original respondents, now young adults. Inclusion criteria for military sample included responding if the participant has ever been in the military and or is currently serv- ing in the military (n¼ 449). Accounting for the rela- tive sample size,4 the analytic strategy took two approaches in order to address potential issues with the subsample (Wolf, Harrington, Clark, & Miller, 2013). First, as recommended by Muth�en and Muth�en (2002), we tested the stability of the results by running the analyses multiple times with a new, randomly

Figure 1. Theoretical simultaneous path model of the correlates of IPV.

SUBSTANCE USE & MISUSE 243

selected seed number5 (i.e., so that data generation began at a different point). Second, we designed mod- els in which the number of indicators per factor was greater than the minimum number of indicators required for model identification (Marsh, Wen, & Hau, 2004).

Measures

The following variables were used as exogenous indi- cators. Frequency of physical IPV was constructed by making the mean is zero and the standard deviation is one. This allowed for a standardized composite score for physical IPV. Sexual abuse was measured as a dichotomous variable, where respondents answered questions about forced sex on partners. Perpetrator demographic factors as reported by the perpetrator include military service (yes/no), military cohort (i.e., Post-9/11 (began military service between 2001 and 2008)), race/ethnicity (White, Black, and Latino), and previous exposure or victim of sexual childhood abuse (yes/no).

A measurement model was used to create an endogenous latent construct SES. SES is manifested by two observed variables: income and education. Income designates annual income of the perpetrator; this variable was constructed as a categorical ordinal variable using 12 categories with increments of $10,000 each, beginning with $5000 or less to $150,000 and more. Education was created using a single item intake question, education designates the highest grade completed at the time of the survey; this variable was constructed as a categorical ordinal vari- able using six categories beginning with less than a high school diploma and ending at post-graduate school. A measurement model was also used to create an endogenous latent construct mental health history. Mental health history is manifested by three observed variables: being diagnosed with depression, diagnosed with post-traumatic stress disorder, and diagnosed with panic or anxiety.

The observed substance use variables6 (i.e., drug use of a prescription drug without prescription, cocaine, methamphetamine use) were all dichotomized where zero indicated that the individual did not use the specific substance within the last 30 d, and one has used the substance (�1 times) within the last 30 d. Marijuana, heroin, and alcohol use were exam- ined as single-item indicators, zero indicated that the individual did not use the specific substance within the last 30 d, and one has used the substance (�1 times) within the last 30 d. Alcohol use and marijuana

use were operationalized so that one indicated that the substance was used more than 12 d per month, and zero indicating that the substance was used less than 8 d per month.

Analysis

The analytic strategies first employed descriptive sta- tistics (means, standard deviations, and category per- centages) using STATA 15 (Table 1). Second, STATA 15 confirmatory factor analysis and SEM were utilized to examine the total effects of the simultaneous equa- tion path model of veterans by race/ethnicity, followed by an examination of the similarity among military and non-military perpetrators (Table 2). In Table 3, SEM was used to assess the adequacy of measurement of the variables and to test whether Pre-9/11 veterans of IPV perpetration followed similar patterns of sug- gested literature on all veterans and whether Post-9/11 veterans of IPV perpetration followed similar patterns of suggested literature on all veterans. The researchers tested a priori specified hypothesis about the underly- ing structure of the simultaneous path model and the structure controlling for measurement error (Ullman, 2006). All estimated parameters were hypothesized a priori, intending to estimate a parsimonious and the- oretically based model (Figure 2). To compare the

Table 1. ADD health descriptive statistics estimates.

ADD health Military

veteran sample

Exogenous variables N¼6504 (n¼499) White 66.02% 61.3% Black 26.48% 16.8% Latino 11.43% 11.2% American Indian/

Native American 3.63% 3.73%

Asian/Pacific Islander 4.15% 3.1% Military veteran 6.90% 2.8% Pre-9/11 veteran (94–00) 69.40% (n¼311) Post-9/11 veteran (01–08) 30.60% (n¼138) Child abuse (sexual, �17) 5.03%

Endogenous variables Income (mean) $40,000 to

$49,999 (8) $40,000 to

$49,999 (8.19) Education

(Grad HS/GED or more) 85.10% 93.4%

Mental health history (diagnosed)

Depression 16.20% 15.88% PTSD 3.10% 15.88% Panic/anxiety 12.50% 11.21% Alcohol use (�12 day per mo) 13.34% 14.77% Marijuana use

(�12 day per mo) 35.27% 7.47%

Prescription drug use, not prescribed

17.30% 21.49%

Cocaine use 19.11% 23.36% Methamphetamine use 9.02% 8.41% Physical abuse 13.47% 8.41% Sexual abuse 2.90% 4.67%

244 R. CANCIO

general population to veterans and service members, this study tested the factor structures between military and non-military by race/ethnicity and cohort (Table 4). In Table 4, group comparison procedure was conducted for the different racial/ethnic groups examining veteran and non-veteran models together and then again grouping all veterans with non-veteran perpetrators. Finally, results from STATA regression analysis of the selected variables indicated that multi- collinearity did not need to be a concern in this model. Absolute values of correlations among the independent variables are less that 0.3; estimated regression coefficients are positive as expected. Furthermore, all standardized coefficients are relatively small, and the significance of the regression coeffi- cients is consistent with the significant R2.

Last, the goodness of fit was used to specify a base- line model that fits the data’s veteran perpetrators of IPV and substance use.7 The baseline model specifica- tion process took two sequential stages by the status of military service. Starting with the “propositional” model shown in Figure 1, the researchers first observed indications of model misspecification. To this end, modification indices were employed.8 The

second stage involved systematically trimming out of non-significant paths (i.e., coefficient estimates with p values > .05).9 At each step, interim evaluations were carried out in search of any relevant path once the model had been simplified. The process of finding the best-fitting model stopped when no additional paths were suggested by the modification indices, as all- remaining paths retained statistical significant given acceptable levels of model fit (Table 5). Figure 3 illus- trates the findings in Table 5, showing the distinctive baseline models by military cohort.

Table 2. Military and non-military theoretical simultaneous path model.

Non-military Military veteran

Child Abuse (Sexual) ! Mental Health �.2457� �.2288� Child Abuse (Sexual) ! SES �.4172��� �.4035��� Race/ Ethnicity ! Mental Health �.3203��� �.3434��� Race/ Ethnicity ! SES �.2483� �.2834�� Military Veterans ! Mental Health – �.1604 Military Veteran ! SES – �.1492 Mental Health ! Cocaine .8602��� .8764��� Mental Health ! Meth .6092��� .6087��� Mental Health ! Prescription Drugs .4355��� .4373��� Mental Health ! Alcohol .2657�� .2673�� Mental Health ! Marijuana .1025�� .1015 SES ! Cocaine �.3048�� �.3279�� SES ! Meth �.3035�� �.3200�� SES ! Prescription Drugs .1306 .1238 SES ! Alcohol .0894 .0808 SES ! Marijuana �.5765��� �.6213��� Cocaine ! Physical IPV .0707 .0705 Cocaine ! Sexual IPV �.0670 �.0714 Meth ! Physical IPV .0670 .06698 Meth ! Sexual IPV �.0015 .0010 Prescription Drugs ! Physical IPV .1077 .1078 Prescription Drugs ! Sexual IPV �.0979 �.0990 Alcohol ! Physical IPV �.0517 �.0517 Alcohol ! Sexual IPV �.0143 �.0143 Marijuana ! Physical IPV .0452 .0452 Marijuana ! Sexual IPV �.0636 �.0636 Goodness-of-fit statistics v2 92.93�� 90.70�� RMSEA 0.083� 0.084� DF 49 58 CFI 0.871 0.843

Note: N¼ 6504; SES: socioeconomic status; RMSEA: root mean square error of approximation.�p < .10, ��p < .05, ���p< .01.

Table 3. Theoretical simultaneous path model for each mili- tary cohort.

Post-9/11 Vets

Race/ Ethnicity ! Education .0001 Race/ Ethnicity ! Income �.0439 Race/ Ethnicity ! Depression �.1158��� Race/ Ethnicity ! PTSD �.0647� Race/ Ethnicity ! Anxiety .1318�� Veteran Cohort ! Education �.0366 Veteran Cohort ! Income �.0711�� Veteran Cohort ! Depression .0174 Veteran Cohort ! PTSD .1318��� Veteran Cohort ! Anxiety �.0057 Education ! Cocaine �.0763�� Education ! Meth �.1985��� Education ! Prescription Drugs .0463 Education ! Alcohol .0167 Education ! Marijuana �.1492��� Income ! Cocaine .0852�� Income ! Meth .0189 Income ! Prescription Drugs .0336 Income ! Alcohol .0356 Income ! Marijuana �.1049��� Depression ! Cocaine .0537 Depression ! Meth .0297 Depression ! Prescription Drugs .0324 Depression ! Alcohol �.0317 Depression ! Marijuana .0220 PTSD ! Cocaine .0524 PTSD ! Meth .0530 PTSD ! Prescription Drugs .0434 PTSD ! Alcohol �.0098 PTSD ! Marijuana .0077 Anxiety ! Cocaine .0445 Anxiety ! Meth .0220 Anxiety ! Prescription Drugs .0711� Anxiety ! Alcohol �.0324 Anxiety ! Marijuana .0037 Cocaine ! Physical IPV .0193 Cocaine ! Sexual IPV �.0099 Meth ! Physical IPV .0577 Meth ! Sexual IPV .0206 Prescription Drugs ! Physical IPV .0034 Prescription Drugs ! Sexual IPV .0361 Alcohol ! Physical IPV �.0761�� Alcohol ! Sexual IPV .0004 Marijuana ! Physical IPV .0421 Marijuana ! Sexual IPV .0698�� Goodness-of-fit statistics – v2 111.063 DF 45 RMSEA 0.166 CFI 0.167

Note: N¼ 449; SES: socioeconomic status; RMSEA: root mean square error of approximation.�p < .10, ��p < .05, ���p< .01.

SUBSTANCE USE & MISUSE 245

Findings

Table 1 reports the aggregating Wave I and Wave IV ADD Health data sets. In Table 2, we illustrate the results of the theoretical simultaneous path modeling (Figure 1) between military respondents and non-mili- tary respondents. In Table 2, only marginal support was found for the hypothesized models for non-

military members [v2 (df ¼ 49, N¼ 6504) ¼ 75.47, p < .05, Robust CFI ¼ 0.871, RMSEA ¼ .083] and military members [v2 (df ¼58, N¼ 6504) ¼ 90.70, p < .05, Robust CFI ¼ .843, RMSEA ¼ .084]. Overall RMSEA and CFI scores indicated that the models were reasonably acceptable (Awang, 2012).

We also examined the differences between IPV per- petration within race/ethnicity and military cohorts. SEM was used to examine the factor structures between military and non-military perpetration of physical and sexual IPV by race/ethnicity and cohort (Table 4). The factor structures showed different rela- tionships between military status and IPV type depending on the participant’s race/ethnicity and vet- eran cohort, especially among Black, Latino, and White participants. Black non-veterans as well as Black Post-9/11 veterans were more likely to carry out both physical and sexual IPV. Post-9/11 Latino veter- ans were less likely to inflict physical IPV whereas they were significantly more likely to commit sexual IPV than their civilian counter parts. Non-veteran Latinos were more likely to perpetrate both physical and sexual IPV. Regarding White participants, non-

Figure 2. Military and non-military theoretical simultaneous path model.

Table 4. Factor structures by race/ethnicity and veteran cohort. Physical IPV Sexual IPV v2 RMSEA AIC CFI

Black .0841 .0523 58.13 0.108 14718.71 0.449 Black Post-9/11 Veterans .0081 .0142 62.277 ��� 0.112 �4765.745 0.000 White �.0989��� �.0582��� 56.867 0.106 15179.58 0.531 White Post-9/11 Veterans �.0204 .0216 62.941 ��� 0.112 3090.602 0.037 Asian or Pacific Islander .0280� .0210 61.80 0.111 5756.15 0.062 Asian or Pacific Islander Post-9/11 Veterans �.0047 �.0021 62.380 0.112 �18704.933 0.000 American Indian .0406��� .0017 62.41 0.112 �704.41 0.091 American Indian Post-9/11 Veterans – – – – – – Latinos .0053 .0161 61.043�� 0.110 16335.135 0.021 Latinos Post-9/11 Veterans �.0165 .0474��� 62.153��� 0.111 �6445.049 0.146

Note: N¼ 449. RMSEA: root mean square error of approximation; CFI: comparative fit index; AIC: Akaike’s information criterion.�p < .10, ��p < .05, ���p< .01.

Table 5. Baseline model by military cohort, substance use, and IPV perpetration.

Post-9/11 Vets

Veteran Cohort ! Mental Health .3132�� Mental Health! Meth .4171��� Mental Health ! Marijuana .7033��� Mental Health ! Prescription Drugs .4734��� Meth ! Physical IPV .5871�� Meth ! Sexual IPV 0.2194 Prescription Drugs ! Sexual IPV 0.4125 Sexual IPV $ Physical IPV .1090��� Goodness-of-fit statistics v2 76.15��� DF 18 RMSEA 0.085 CFI 0.855

Note: N¼ 449. RMSEA: root mean square error of approximation; CFI: comparative fit index.�p < .10, ��p < .05, ���p< .01.

246 R. CANCIO

veterans were significantly less likely to perpetrate both physical and sexual IPV. Post-9/11 White veter- ans were less likely to perpetrate physical IPV whereas they were more likely to perpetrate sexual IPV. Race/ ethnicity served as a protective factor in API groups, in that API veterans were significantly less likely to perpetrate physical, and less likely to commit sexual IPV compared to API non-veterans.

Mental health issues predicted alcohol use, as well as cocaine, methamphetamine, and prescription drug use for both veterans and civilians (with the exception that mental health was only a significant predictor of marijuana among civilians). Alcohol use had statistic- ally identical relationships with both physical and sex- ual IPV among veterans and civilians. Finally, CFA post-hoc model modifications (Hutchinson, 1998) were performed in attempt to develop better fitting models (Table 5). The final model adequately fit the data forPost-9/11 veterans [v2 (df ¼18, N¼ 449) ¼ 76.15, p < .05, Robust CFI ¼.855, RMSEA ¼ .085]. Figure 3 illustrates the mental health, AOD use, and perpetration of physical and sexual IPV and the indi- vidual factor structures can be found in Table 5. Post- 9/11 veterans, veteran cohort significantly predicted mental health (standardized coefficient ¼ .31, p <

.01), which focused on depression and PTSD. The Post-9/11 veteran model involved the following AOD use: marijuana, methamphetamine, and prescription drugs. Higher mental health issues predicted greater use of marijuana (standardized coefficient ¼ .70, p <

.01), methamphetamine (standardized coefficient ¼

.42, p < .001), and prescription medications (standar- dized coefficient ¼ .47, p < .001). Moreover, physical IPV was significantly predicted by higher metham- phetamine use (standardized coefficient ¼ .59, p <

.01). Neither methamphetamine (standardized coeffi- cient ¼ .22, p > .05) use nor prescription drug use (standardized coefficient ¼ .41, p > .05) significantly predicted sexual IPV.

Discussion

The results of the analyses demonstrated that IPV in among veteran cohorts manifests differently and were significantly associated with the consumption of AOD use among male perpetrators. The central finding in the present study was that perpetration of physical and sexual IPV depends on the context of veteran cohort and race/ethnicity. Mental health, AOD use, and other demographic variables have different pre- dictive values and relationships with IPV depending on generational impacts related to veteran cohorts as well as ones race/ethnicity within the Post-9/11 vet- eran cohort. The present findings suggest that using drugs other than alcohol may be an important factor related to the perpetration of IPV when looking at differences between veterans and civilians. Recently, Cafferky, Mendez, Anderson, and Stith (2018) con- ducted a meta-analysis and found that both alcohol and drug use, such as amphetamines, cocaine, mari- juana, were similarly associated with IPV perpetration (r effect sizes ranged from .20 to .25). AOD uses seem to have similar associations with IPV among veterans and civilians. Furthermore, veteran status and AOD use among veterans did not associate stronger with physical IPV compared to sexual IPV. However, with- out considering how mental health and AOD use interacts with veteran cohort and race/ethnicity, the SEM models present similar findings for military members and civilians, which can mask important dif- ferences and unique interactions where they exist.

The present study finds that military status serves as a protective factor for API and American Indian participants, whereby veterans were less likely to per- petrate physical and sexual IPV toward their spouse/ partner and civilians were more likely to perpetrate physical and sexual IPV toward their spouse/partner. However, military status among White participants was a risk factor for IPV. White civilians were less likely to perpetrate physical and sexual IPV, but

Figure 3. Baseline model by military cohort, substance use, and IPV perpetration.

SUBSTANCE USE & MISUSE 247

White veterans were more likely to perpetrate physical and sexual IPV. Veteran cohort among White and Latino participants was a protective factor for physical IPV but a risk factor for sexual IPV. Both White and Latino non-veterans and Post-9/11 veterans, respective to their ethnic/racial groups, were not more likely to perpetrate physical IPV but more likely to perpetrate sexual IPV. Older veterans were less likely to perpet- rate physical IPV but Black civilians and Black Post-9/ 11 veterans were more likely to perpetrate physical and sexual IPV. These results point to importance of examining IPV perpetration within certain social and cultural contexts such as service era veteran cohorts, different mental illnesses and AOD uses associated with veteran cohorts, race/ethnicity and the inter- action among these characteristics.

This study’s findings focus on the theoretical rational about military veterans, service members, AOD use, and IPV perpetration. Although there were differences between race and ethnicity among service members and veterans, an overall characteristic of military perpetrators demonstrated that the perpetra- tion of physical and sexual IPV depends on the con- text of veteran cohort and race/ethnicity. While this study presents a significant contribution to the litera- ture on IPV perpetration and military cohorts, it has some limitations. First, comparing veteran cohorts is difficult and inevitably contentious, but it can be approached systematically by choosing categories for comparison (e.g., psychological impact) (Weissbecker, Sephton, Martin, & Simpson, 2008). This study com- pares veterans grouped by era of service and serves as in initial attempt to flesh out various contextual nuan- ces between veteran cohorts. Second, data limitations placed restrictions on other common antecedents of IPV and military families (i.e., social support, depres- sion, PTSD, and TBI as potential measures). For example, information regarding the status of mental health factors might explain why certain military eth- nic and racial groups had significantly distinctive con- fidence. Third, this study employed data from self- reported for IPV perpetration and AOD use. Furthermore, the study was not able to differentiate between types of military service (e.g., reservist, active duty, retired, or prior service). In addition, although important differences were described between Gulf War and Post-9/11 veteran, other military-specific contextual factors that were not available within this dataset (e.g., military occupation, rank, exposure to combat, pre-military existing conditions, etc.), could help to explain the differences found between these two cohorts. Knowing about the military-specific

context of an individual service member and or vet- eran might allow for a closer examination of the last- ing effects of exposure to the military lifestyle in the form of military career. Finally, the data did not include violence occurring between partners with female perpetrators or partners of the same gender.

Despite these limitations, this research is an important step in understanding the nature of military service and its impact on the outcomes of IPV perpet- ration. The effect of military service on the relation- ship of IPV perpetration suggests an expansion in the understanding of violence in the context beyond theo- ries of AOD use and race/ethnicity. These findings warrant further investigation to understand the differ- ences in patterns and dynamics of IPV associated with military service and non-military service–related IPV. These differences may highlight opportunities for potential interventions and have implications for the design of perpetrator programs. The central position of considering military service in the pathways to IPV perpetration spells out potential contributions to pub- lic health in response to military trauma and in the prevention of violence, especially in light of the two major US conflicts (i.e., Iraq and Afghanistan). In addition, there is no formal mechanism in existence for identifying IPV among military personnel and/or veterans. As a result, medical and behavioral health professionals may represent the first line of defense in recognizing victims and perpetrators of IPV and in linking these individuals to appropriate services (Hart & Andrew, 2013). The evidence from this study sup- ports the need to develop integrated prevention inter- ventions that incorporate mental and physical health components that build emotional resilience and cop- ing mechanisms of participants with military service. This may decrease the psychological impact of past trauma (i.e., military combat) while preventing future violence perpetration.

The collaboration of health practitioners and researchers can co-create a design of future integrated treatments that take a holistic approach and focus on addressing IPV perpetration. These findings highlight the need for the development of more positive inter- ventions that break the intergenerational transmission of violence in this context of past abuse and military service for perpetrators. The findings in this study serve as an initial step in drafting out the variations in military subgroups and may contribute to tailored interventions for IPV perpetrators.

For veterans, AOD use, and mental health and emotional wellness characteristics represent important correlates of IPV perpetration. PTSD, depression, and

248 R. CANCIO

anxiety symptomatology also represents a robust cor- relate of IPV for veterans; PTSD symptoms largely account for the relationship between warzone stressors and IPV, and PTSD comorbidity is associated with IPV. Such findings suggest that interventions targeting PTSD in particular may serve to ameliorate the effects of military stressor exposure on IPV perpetration. In addition, the uniqueness between IPV and AOD use among military cohorts directs attention to improving perpetrators’ treatment programs and future research should focus on developing different types of inter- ventions for different types of IPV perpetrators, rather than relying on a one-size fits all approach for mili- tary service members and veterans. Furthermore, there are limitations to comparing racial groups with respects to IPV. Future studies need to have a discus- sion of how historical racism and oppression may contribute to some of the observed substance use, mental health, and IPV differences. The uniqueness between IPV and substance use among military cohorts directs attention to improving perpetrators’ treatment programs and future research should focus on developing different types of interventions for dif- ferent types of IPV perpetrators, rather than relying on a one-size fits all approach for military service members and veterans. Also, future studies should examine micro-predictors of aggression in close prox- imity to the act that includes type and consumption pattern of specific psychoactive substances.

Future research should focus on extrapolating how other military-related factors (i.e., PTSD, depression, and TBI) contribute to IPV perpetration individually. Although there were relative ethnic and racial differ- ences on IPV perpetration, these findings cannot be generalizable to the whole military population and to racial and ethnic groups. Researchers on IPV perpet- ration should consider examining other possible social structural forces (e.g., military occupation, rank, exposure to combat, pre-military existing conditions, etc.) that may act as risk and or protective factors that push and pull current service members and or veteran to become IPV perpetrators.

Notes

1. Statistically, structural equation modeling (SEM) allows one to operationalize a conceptual theoretical framework and its validation indices can verify several dimensions simultaneously measuring various aspects of intimate partner violence (IPV).

2. The calculated scores for these dimensions are measures that can serve as a basis for studying violence in the military family.

3. For the purpose of this study, men are considered not because they are assumed to be perpetrators, but because men are twice as likely to be stressed for an upcoming deployment than women, men are more likely to be stressed by their absence to their family than women, men are more likely to be stressed by financial reasons than women, men are more likely to cope with stress by drinking, and having violent obsessions and thoughts than women (Bray, Fairbank, & Marsden, 1999). According to the U.S. Department of Defense, most veterans and military personnel are men, 85% respectively (U.S. Department of Defense, 2015).

4. Some evidence exists that simple SEM models could be meaningfully tested even if the sample size is quite small (Hoyle, 1999; Hoyle & Kenny, 1999; Marsh & Hau, 1999), but usually, N¼ 100–150 is considered the minimum sample size for conducting SEM (Anderson & Gerbing, 1988; Tabachnick & Fidell, 2001; Tinsley & Tinsley, 1987), making both veteran samples adequate.

5. The strategy was to focus on solution propriety in lieu of trying to attend only to statistical power by taking the mean minimum sample size based on analyses with several different seed numbers

6. Alcohol and substance use variables were not factored into any latent constructs in order to explicitly identify differences between the various categories.

7. Goodness of fit was evaluated using three indices (Kline, 2011). The root means square error of approximation (RMSEA) incorporates a penalty function for poor model parsimony (Hu & Bentler, 1999). Values under 0.06 suggest close approximate (adequate) fit, whereas values above 0.10 signify a poorly fitting model and that the model should be rejected (Kline, 2011). The comparative fit index (CFI) and goodness-of-fit index (GFI) represent incremental fit indices contrasting the hypothesized theoretical model to a more restricted nested baseline model (Kline, 2011). Both range from zero to one and values >0.9 are indicative of adequate fit (Tucker & Lewis, 1973).

8. A modification index reflects how much the overall model chi-square decreases or improves if a constrained parameter is freely estimated. Candidate paths involving conspicuous values (modification indices > 10) were then examined for the actual amount of model fit improvement and for the magnitude of the following freed estimated coefficients. The decision to explore and keep new paths also followed their theoretical meaningfulness.

9. This process adheres to the hierarchical principle advocated in the theoretical model. The trimming process thus moved from left to right, starting with paths beginning from<Race/Ethnicity> to<Physical /Sexual>.

Disclosure statement

No potential conflict of interest was reported by the authors.

SUBSTANCE USE & MISUSE 249

References

Ahmadabadi, Z., Najman, J. M., Williams, G. M., Clavarino, A. M., d’Abbs, P., & Smirnov, A. (2019). Intimate partner violence in emerging adulthood and subsequent sub- stance use disorders; Findings from a longitudinal study. Addiction, 114(7), 1264–1273. doi:10.1111/add.14592

Anderson, J. C., & Gerbing, D. W. (1988). Structural equa- tion modeling in practice: A review and recommended two-step approach. Psychological bulletin, 103(3), 411.

Awang, Z. (2012). Structural equation modeling using AMOS graphic. Penerbit Universiti Teknologi MARA.

Black, M. C. (2011). Intimate partner violence and adverse health consequences: Implications for clinicians. American Journal of Lifestyle Medicine, 5(5), 428–439. doi:10.1177/1559827611410265

Bray, R., Fairbank, J., & Marsden, M. (1999). Stress and substance use among military women and men. The American Journal of Drug and Alcohol Abuse, 25(2), 239–256.

Breiding, M. J., Chen, J., & Black, M. C. (2014). Intimate partner violence in the United States — 2010. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.

Cafferky, B. M., Mendez, M., Anderson, J. R., & Stith, S. M. (2018). Substance use and intimate partner violence: A meta-analytic review. Psychology of Violence, 8(1), 110–131. doi:10.1037/vio0000074

Campbell, D. J., & Nobel, O. B.-Y. (2009). Occupational stressors in military service: A review and framework. Military Psychology, 21(Suppl 2), S47–S67. doi:10.1080/ 08995600903249149

Campbell, J. C. (1992). If I can’t have you, no one can”: Power and control in homicide of female partners. In J. Radford & D. E. Russell (Eds.), Femicide: The politics of woman killing (pp. 99–113). New York, NY: Twayne.

Cancio, R. (2015). The pain of combat for Mexican- American Veterans: A cohort analysis of Vietnam and Iraq/Afghanistan Veterans. Open Access Theses, 555. Retrieved from https://scholarlyrepository.miami.edu/oa_ theses/555

Cancio, R. (2017). Structural pathways between race/ethni- city, substance use, military service, and male-perpetrated intimate partner violence in pre-9/11 military families. Journal of Interpersonal Violence, 088626051772314. doi: 10.1177/0886260517723142

Cancio, R. (2018). Pain and masculinity: A cohort compari- son between Mexican American Vietnam and post–9/11 combat Veterans. Men and Masculinities, 1097184X18761779, 1097184X1876177.

Cancio, R., & Altal, D. (2019). Comparing post-Gulf War and post-9/11 era of service among veterans: Intimate partner violence and substance use by race and ethnicity. Journal of Ethnicity in Substance Abuse, 1–27. doi:10. 1080/15332640.2019.1571978

Choenni, V., Hammink, A., & van de Mheen, D. (2017). Association between substance use and the perpetration of family violence in industrialized countries: A system- atic review. Trauma. Violence, & Abuse, 18(1), 37–50. doi:10.1177/1524838015589253

Coker, A. L., Davis, K. E., Arias, I., Desai, S., Sanderson, M., Brandt, H. M., & Smith, P. H. (2002). Physical and

mental health effects of intimate partner violence for men and women. American Journal of Preventive Medicine, 23(4), 260–268. doi:10.1016/S0749- 3797(02)00514-7

De Bruijn, D. M., & De Graaf, I. M. (2016). The role of substance use in same-day intimate partner violence: A review of the literature. Aggression and Violent Behavior, 27, 142–151. doi:10.1016/j.avb.2016.02.010

Fontana, A., & Rosenheck, R. (2010). War zone veterans returning to treatment: Effects of social functioning and psychopathology. The Journal of Nervous and Mental Disease, 198(10), 699–707. doi:10.1097/NMD. 0b013e3181f4ac88

Gewirtz, A. H., Polusny, M. A., DeGarmo, D. S., Khaylis, A., & Erbes, C. R. (2010). Posttraumatic stress symptoms among National guard soldiers deployed to Iraq: Associations with parenting behaviors and couple adjust- ment. Journal of Consulting and Clinical Psychology, 78(5), 599–610. doi:10.1037/a0020571

Gonzalez, O., Novaco, R., Reger, M., & Gahm, G. (2016). Anger intensification with combat-related PTSD and depression comorbidity. Psychological Trauma: Theory, Research, Practice, and Policy, 8(1), 9–16. doi:10.1037/ tra0000042

Grossman, D. (1995). On killing: The psychological cost of learning to kill in war and society. New York, NY: Little, Brown.

Harris, K. M., Halpern, C. T., Whitsel, E., Hussey, J., Tabor, J., Entzel, P., & Udry, J. R. (2009). The National longitu- dinal study of adolescent to adult health: Research design [WWW document]. Retrieved from http://www.cpc.unc. edu/projects/addhealth/design.

Hart, J. D., & Andrew, J. (2013). Practical implications of current intimate partner violence research for victim advo- cates and service providers. Washington D.C.: U.S. Department of Justice.

Heinz, A. J., Makin-Byrd, K., Blonigen, D. M., Reilly, P., & Timko, C. (2015). Aggressive behavior among military veterans in substance use disorder treatment: The roles of posttraumatic stress and impulsivity. Journal of Substance Abuse Treatment, 50, 59–66. doi:10.1016/j.jsat.2014.10.014

Heyman, R. E., Taft, C. T., Howard, J. M., Macdonald, A., & Collins, P. S. (2012). Intimate partner violence. In D. K. Snyder & C. M. Monson (Eds.), Couple based inter- ventions for military and veteran families: A practitioner’s guide (pp. 122–146). New York, NY: Guilford Press.

Hoyle, R. H., & Kenny, D. A. (1999). Sample size, reliability, and tests of statistical mediation. Statistical strategies for small sample research, 1, 195–222.

Hoyle, R. H. (Ed.). (1999). Statistical strategies for small sample research. Sage.

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55.

Hutchinson, S. R. (1998). The stability of post hoc model modifications in confirmatory factor analysis models. The Journal of Experimental Education, 66(4), 361–380.

Johnson, W., Giordano, L., Manning, P., & Longmore, C. (2015). The age–IPV curve: Changes in the perpetration of intimate partner violence during adolescence and

250 R. CANCIO

young adulthood. Journal of Youth and Adolescence, 44(3), 708–726. doi:10.1007/s10964-014-0158-z

Judd, L. L., Schettler, P. J., Coryell, W., Akiskal, H. S., & Fiedorowicz, J. G. (2013). Overt irritability/anger in uni- polar major depressive episodes: Past and current charac- teristics and implications for long-term course. JAMA Psychiatry, 70(11), 1171–1180. doi:10.1001/jamapsychia- try.2013.1957

Kline, R. B. (2011). Convergence of structural equation modeling and multilevel modeling.

Marsh, H. W., & Hau, K. T. (1999). Confirmatory factor analysis: Strategies for small sample sizes. Statistical strat- egies for small sample research, 1, 251–284.

Marsh, H. W., Wen, Z., & Hau, K. T. (2004). Structural equation models of latent interactions: Evaluation of alternative estimation strategies and indicator construc- tion. Psychological Methods, 9(3), 275. Retrieved from https://search.ebscohost.com/login.aspx?direct=true&db= eric&AN=EJ685065&site=eds-live&scope=site.

Marshall, A. D., Panuzio, J., & Taft, C. (2005). Intimate partner violence among military veterans and active duty servicemen. Clinical Psychology Review, 25(7), 862–876. doi:10.1016/j.cpr.2005.05.009

McCarroll, J. E., Ursano, R. J., Newby, J. H., Liu, X., Fullerton, C. S., Norwood, A. E., & Osuch, E. A. (2003). Domestic violence and deployment in U.S. army soldiers. The Journal of Nervous and Mental Disease, 191(1), 3–9. doi:10.1097/00005053-200301000-00002

Moore, T., Stuart, G., Meehan, J., Rhatigan, D., Hellmuth, J., & Keen, S. (2008). Drug abuse and aggression between intimate partners: A meta-analytic review. Clinical Psychology Review, 28(2), 247–274. doi:10.1016/j.cpr.2007. 05.003

Muth�en, L. K., & Muth�en, B. O. (2002). How to use a Monte Carlo study to decide on sample size and deter- mine power. Structural Equation Modeling: A Multidisciplinary Journal, 9(4), 599–620. doi:10.1207/ S15328007SEM0904_8

Novaco, R., & Chemtob, C. (2015). Violence associated with combat-related posttraumatic stress disorder: The import- ance of anger. Psychological Trauma: Theory, Research, Practice, and Policy, 7(5), 485–492. doi:10.1037/ tra0000067

Rosen, L. N., Kaminski, R. J., Parmley, A. M., Knudson, K. H., & Fancher, P. (2003). The effects of peer group cli- mate on intimate partner violence among married U.S. Army soldiers. Violence against Women, 9(9), 1045–1071. doi:10.1177/1077801203255504

Rosen, L. N., Knudson, K. H., Brannen, S. J., Fancher, P., Killgore, T. E., & Barasich, G. G. (2002). Intimate partner violence among U.S. Army soldiers in Alaska: A compari- son of reported rates and survey results. Military Medicine, 167(8), 688–691. doi:10.1093/milmed/167.8.688

Seal, K. H., Metzler, T. J., Gima, K. S., Bertenthal, D., Maguen, S., & Marmar, C. R. (2009). Trends and risk factors for mental health diagnoses among Iraq and Afghanistan veterans using department of Veterans affairs health care, 2002–2008. American Journal of Public Health, 99(9), 1651–1658. doi:10.2105/AJPH.2008.150284

Shorey, R. C., Stuart, G. L., Mcnulty, J. K., & Moore, T. M. (2014). Acute alcohol use temporally increases the odds of male perpetrated dating violence: A 90-day diary ana- lysis. Addictive Behaviors, 39(1), 365–368. doi:10.1016/j. addbeh.2013.10.025

Smith, P. H., Homish, G. G., Leonard, K. E., & Cornelius, J. R. (2012). Intimate partner violence and specific sub- stance use disorders: Findings from the national epide- miologic survey on alcohol and related conditions. Psychology of Addictive Behaviors, 26(2), 236–245. doi:10. 1037/a0024855

Tabachnick, B. G., & Fidell, L. S. (2001). Principal compo- nents and factor analysis. Using multivariate statistics, 4, 582–633.

Taft, C. T., Watkins, L. E., Stafford, J., Street, A. E., & Monson, C. M. (2011). Posttraumatic stress disorder and intimate relationship programs: A meta-analysis. Journal of Consulting and Clinical Psychology, 79(1), 22–33. doi: 10.1037/a0022196

The Women’s Justice Project. (2011). Military calls 2006–2011. Retrieved from http://www.bwjp.org/assets/ documents/pdfs/ndvh_military_calls_2006_2014.pdf

Tinsley, H. E., & Tinsley, D. J. (1987). Uses of factor analysis in counseling psychology research. Journal of Counseling Psychology, 34(4), 414.

Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38(1), 1–10.

Ullman, J. (2006). Structural equation modeling: Reviewing the basics and moving forward. Journal of Personality Assessment, 87(1), 35–50. doi:10.1207/s15327752jpa8701_03

United States Department of Defense. (2015). 2015 Demographics profile of the military community. Arlington, VA: The Pentagon.

Weissbecker, I., Sephton, S. E., Martin, M. B., & Simpson, D. M. (2008). Psychological and physiological correlates of stress in children exposed to disaster: Current research and recommendations for intervention. Children Youth and Environments, 18(1), 30–70. Retrieved from https:// www.jstor.org/stable/10.7721/chilyoutenvi.18.1.0030

Wolf, E. J., Harrington, K. M., Clark, S. L., & Miller, M. W. (2013). Sample size requirements for structural equation models: An evaluation of power, bias, and solution pro- priety. Educational and Psychological Measurement, 73(6), 913–934. doi:10.1177/0013164413495237

SUBSTANCE USE & MISUSE 251

Copyright of Substance Use & Misuse is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.