The Impact of Fatherhood on Treatment Response for Men with Co-Occurring Alcohol Dependence and Intimate Partner Violence

Abstract

Objective

The role of fathers in the lives of children has gained increasing attention over the last several decades, however studies that specifically examine the parenting role among men who are alcohol dependent and have co-occurring intimate partner violence (IPV) have been limited. This brief report is intended to highlight the need to develop and focus interventions for men with co-occurring substance abuse and IPV with an emphasis on their roles as fathers.

Method

Sixty-nine men who participated in a randomized comparison study of a coordinated substance abuse and domestic violence treatment program (SADV) and Twelve Step Facilitation (TSF), provided information about whether they were fathers. Analysis of covariance was used to assess the impact of fatherhood on the outcomes of intimate partner violence and alcohol use during the 12-weeks of treatment.

Results

There was a significant interaction between type of treatment (SADV vs. TSF) and fatherhood. SADV resulted in significantly less IPV and use of alcohol over the 12 weeks of treatment than TSF for men without children. There were no significant differences between SADV and TSF for men who were fathers.

Conclusion

Results indicate a need to further explore the role of fatherhood for men with co-occurring substance abuse and IPV and development of specialized treatments that may improve treatment outcomes for fathers.

Children’s attachments to their fathers and the role of fathers in child development have gained increasing attention over the last few decades (e.g. Lamb, 2004; Pruett, 1983; Hawkins & Dollahite, 1997; Hobson, 2002). Involvement of fathers and the nature of the father-child relationship as it relates to development have been the primary focus of recent research (Marsiglio et al., 2000). Despite the vast literature on the parenting of mothers and the development of interventions for mothers with various psychiatric conditions, substance abuse disorders, and personality styles, fathers have only recently begun to receive the same attention in the intervention literature (Marsiglio et al., 2000). This is particularly true in the area of fathers who abuse substances or perpetrate intimate partner violence (IPV). There is significant evidence to demonstrate the deleterious effects of witnessing IPV or parental substance abuse on children (see Edleson, 1999; Peled, Jaffe, & Edleson, 1995; Sternberg et al., 1993; Kendall-Tackett, 2004; Johnson & Leff, 1999; West & Prinz, 1987). In addition, the comorbidity of substance abuse and intimate partner violence is generally thought to be between 40% and 50% (e.g., see Easton, Swan, & Sinha, 2000). Two recent meta-analytic reviews found significant effects for the association of alcohol and drug use with intimate partner violence (Foran & O’Leary, 2008; Moore et al., 2008). Yet, little is currently known about the number of children living with fathers with co-morbid substance abuse and violence problems or how their roles as fathers might impact treatment.

The limited literature suggests that a large percentage of fathers who perpetrate domestic violence have an awareness that their violence negatively impacts their children (Rothman, Mandel, & Silverman, 2007) and report increased parenting stress, lower feelings of parental competence, and no one to turn to for advise or questions related to parenting (Baker, Perilla, & Norris, 2001). Still, the paternal role is only minimally acknowledged and integrated into treatment programs for substance abuse or domestic violence.

Whether men who are fathers respond differently to court-mandated treatment programs following domestic violence or substance abuse charges is unknown. This secondary analysis of a randomized study of treatment efficacy was conducted to provide preliminary evidence of the impact of fatherhood on treatment response for men with co-morbid alcohol dependence and IPV.

Methods

Participants

Participants were recruited for enrollment in a pilot treatment study of an innovative, manualized integrated domestic-violence and substance abuse treatment (SADV; Easton, Mandel, Hunkele, Nich, Rounsaville, & Carroll, 2007) from individuals seeking court mandated treatment at an outpatient substance abuse treatment facility affiliated with Yale University in New Haven, CT. They were referred from probation or family relations following an arrest. Subjects were males, 18 years or older who met DSM-IV criteria for Alcohol Dependence and were arrested for domestic violence within the past year. Individuals were excluded if they expressed serious suicidal or homicidal ideation, reported psychotic or manic symptoms in the past month, were in need of medically supervised alcohol detoxification, could not read at the fifth grade level, were currently receiving substance abuse and/or anger management treatment elsewhere or were unwilling to provide consent to contact a female collateral informant.

Sixty-nine of the 75 randomized participants in the longitudinal study completed baseline and follow-up interviews needed for the present analysis. Men were randomized to either SADV treatment (Easton et al., 2007) or Twelve Step Facilitation (TSF; Nowinski, Baker, & Carroll, 2002) using urn randomization to balance the two groups on amount of alcohol use, number of domestic violence arrests, age, education, and race. Study treatments were delivered to participants in weekly, 90-minute group therapy sessions by two master’s level therapists for 12 weeks. SADV is a manualized cognitive behavioral therapy approach modified from that used in the Project MATCH manual that focuses on substance abuse, interpersonal violence, and the relationship between the two in each session. TSF focuses solely on substance abuse and methods of recovery (for further details see (Easton et al., 2007). Following treatment, men were interviewed again by trained research assistants.

The participants had a mean age of 38 (SD = 9.4) years. Forty-nine percent were Caucasian, 33% were African American, and 10% were Hispanic. Seventy percent of the sample was employed, and 63% reported living alone. Most had reported obtaining at least a high school diploma (78%) and 70% reported being employed full-time. The sample had a mean of 2 (SD = 2.0) previous domestic violence arrests. The participants reported using alcohol a mean of 6.0 days (SD = 6.3), marijuana a mean of 1.0 days (SD= 4.4), cocaine a mean of 0.2 days (SD = 2.2) of the past 28 days. All met criteria for current alcohol dependence; 22% met criteria for lifetime marijuana dependence and 20% met lifetime criteria for cocaine dependence.

Measures

For purposes of these analyses, demographic questions about current living situation, how many children, and the custody status of men’s children under the age of 18 at the outset of treatment were utilized. In addition, the Structured Clinical Interview for the DSM-IV (SCID-IV) (First et al., 1995) was used to assess current and lifetime substance dependence diagnosis.

Violence

The Conflict Tactics Scale Revised (CTS2; Straus, Hamby, Boney-McCoy, & Sugarman, 1996) was administered pre and post-treatment as a measure of intimate partner violence episodes. Participants were asked to report use of physical violence toward their partner in the last 30 days. Examples of violence items included: twisting partner’s arm, partner received a cut/bruise in fight, forcing sex, pushing/shoving, hitting/holding down, punching, slamming, grabbing, slapping, partner feeling pain, using a weapon, partner needing medical attention, and/or partner having a sprain or broken bone. The frequency in the last 30 days was recorded on a likert type scale (0=never, 1=once, 2=twice, 3= three to five times, 4= 6–10 times, 5=11–20 times, 6=more than 20 times). A sum of all physical and severe violence items was calculated for the month prior to treatment (baseline) and for the 30 days preceding post-treatment interviews (post-treatment) (Straus et al., 1996).

Collateral Reports of Violence

Female collateral informants were contacted by phone at baseline and post-treatment to complete the CTS2. Data was gathered to assess consistency in reporting between the male participants and his female partner. Collateral data were successfully obtained from female partners of 55% of the male participants. Fifty-eight percent were consistent with the male participants’ reports of violence. Twenty-two percent of the male participants reported violence when the female collateral informant did not, while 20% of the female participants reported violence when their male partner did not. If there was a discrepancy, any endorsement of physical violence by either the male or female participant was counted as an incident of physical violence. Thus the CTS2 score is an aggregate report of both male and female collateral informant report.

To safeguard female collateral informants, information from the male participant and the female collateral informant was kept confidential from the other. The limits of confidentiality and voluntary nature of the study were fully reviewed with each female collateral informant as part of the consent process. Staff members were supervised weekly and investigators, who were all licensed professionals, were kept aware of male participant and collateral informant progress during the course of treatment. All interviews took place at a licensed/JCAHO accredited treatment facility. No serious adverse events, adverse events or mandated reporting occurred during the course of the study. For further details see (Easton, et al., 2007).

Alcohol use

Participants’ reports of the number of days they used alcohol in the last month on the Addiction Severity Index (McLellan, Kushner, Metzger, et al., 1992) was used to assess alcohol use in the month prior to treatment. A weekly timeline followback calendar (Fals-Stewart, O'Farrell, Freitas, McFarlin, & Rutigliano, 2000) was used to assess use of alcohol and substances during the 12 weeks of treatment resulting in a total number of days used alcohol during treatment score.

Statistical Analysis

Analysis of Covariance (ANCOVA) was used to assess differences in use of physical violence and number of days of alcohol use during treatment associated with treatment group (TSF vs. SADV) and fatherhood (dichotomous yes/no variable). Baseline use of violence in the last month (CTS2) and number of days alcohol was used 28 days prior to treatment served as baseline scores and were entered as covariates. ANCOVA is preferable to examination of change scores when there are baseline differences between groups (Vickers & Altman, 2001).

Results

Sixty-eight percent of the men had a least one biological child with 20% of those children of preschool age. Thirty percent of the men reported having custody of a child under the age of 18. There was a significant difference at baseline between men with and without children on severity of violence (F = 5.74, p=.02), with fathers reporting lower levels of violence in the year prior to treatment than non-fathers, M (SD) = 1.83 (4.65) and 8.45 (17.92) respectively. There was no significant difference in frequency of alcohol use in the month prior to treatment (F = 0.87, p = .36), for men with or without children, M (SD) = 6.57 (7.09) and 5.09 (5.24) respectively.

ANCOVA results revealed a significant main effect for treatment group (SADV vs. TSF) on physical violence (F=5.27, p<.05) and alcohol use (F=5.28, p<.05) during the course of treatment. There was no significant main effect for fatherhood, but there was a significant interaction between treatment group and fatherhood status on both use of violence (F=6.54, p<.05) and drinking (F=8.17, p<.01) over the 12 weeks of treatment. Post-hoc comparisons revealed non-fathers who received the SADV treatment were significantly less likely to perpetrate violence or use alcohol during 12 weeks of treatment than non-fathers receiving TSF (see Table 1). There were no significant differences in violence or alcohol use between the groups for men who were fathers.

Table 1.

Pre and Post Treatment Means and Least Square Means for Intimate Partner Violence (IPV) and Use of Alcohol by Fatherhood and Treatment Group

Outcome	N	TSF			N	SADV
		Pre Tx	Post Tx	Ls		PreTx	Post Tx	Ls
Fatherhood		M (se)	M (se)	M(se)		M(se)	M(se)	M(se)
IPV
Fathers	21	2.29(.82)	.10(.06)	0.74(0.63)	19	1.35(.39)	.42(.18)	0.95(0.66)
Non-fathers	8	3.89(1.08)	3.13(1.15)	3.12(1.01)	11	11.62(2.88)	1.27(.55)	−0.87(0.93)*
Use of Alcohol
Fathers	26	7.16(.92)	7.50(1.21)	6.19(1.61)	22	6.17(.80)	6.82(1.07)	7.28(1.73)
Non-fathers	9	7.11(.90)	15.56(2.27)	15.16(2.70)	12	3.69(.43)	1.75(.29)	4.06(2.38)*

Note: TSF=Twelve Step Facilitation; SADV=Substance Abuse and Domestic Violence Treatment; Ls=Least Squared

^*p < .05 for pairwise comparisons with Tukey’s LSD.

Discussion

A sample of men seeking substance abuse treatment with histories of intimate partner violence had high rates of continued cohabitation with their partners and thirty percent had custody of their children. Men who were fathers also reported perpetrating less physical intimate partner violence but similar percentage of days drinking in the month prior to treatment. SADV was significantly more effective than TSF in reducing violence and alcohol use over the 12 weeks of treatment for men who were not fathers. There were no significant differences between treatment groups for men who were fathers. These data suggest a need to explore whether fathers with co-morbid alcohol dependence and IPV may benefit from intervention that more specifically targets their parental role. Given the baseline differences in violence between fathers and non-fathers, it is also possible that men with children are less violent as presence of children serves as a protective factor. Another possibility is that fathers report less or minimize their self-report of violence for fear of child protective service involvement. Further studies are needed to understand the interaction effects of fatherhood found in this small sample. Regardless of the reason for the differences, focus on the parenting role may increase fathers’ motivation for change, success as parents, and improve outcomes for men, women and their children.

Despite the significant overlap of substance abuse and IPV, coordinated treatment approaches for the two conditions have not been the norm. Schumacher, Fals-Stewart, and Leonard (2003) reported that men with a recent history of intimate partner violence rarely are referred by substance abuse counselors to specialized treatment for domestic violence. In addition, studies of batterer intervention groups, typically mandated by the courts, have shown limited efficacy in reducing violence or substance abuse (Babcock, Green, and Robie, 2004). The lack of efficacy for interventions currently mandated by domestic violence courts is alarming, and highlights the need for further research exploring what kind of treatment works best for which types of men with various substance abuse and violence issues.

A few studies have shown substance abuse treatment and Behavioral Couple Therapy to have positive effects on reducing both substance use and violence (Schumacher et al. 2003; Fals-Stewart & O’Farrell, 2003; O’Farrell, Fals-Stewart, Murphy & Murphy, 2003; O’Farrell, Muphy, Stephan, Fals-Stewart, & Murphy, 2004) and others have found some utility to group couple’s interventions for male batterers (Dunford, 2000) especially for couples dealing with issues of co-morbid substance abuse and intimate partner violence (Brannen and Rubin, 1996; Fals-Stewart, Kashdan, O’Farrell, & Birchler, 2003). The feasibility of family or couples’ treatments for IPV has been minimized, often thought to be unethical, and likely to put women and children at increased risk for violence. In fact, a large number of state statutes prohibit the inclusion of partners and children in mandated treatment programs (Healy et al., 1998; Lipchick, Sirles, & Kubicki, 1997; Austin & Dankwort, 1999). Although there are severe cases of IPV, when this would be appropriate, state laws that prohibit the inclusion of partners and children may be contraindicated given the high percentage of victims who remain in their relationships or have continued contact due to shared children as the present data suggest. Development of approaches that would increase positive co-parenting in these families is warranted.

This preliminary study supports the need for further exploration of the role of fatherhood and its relationship to treatment engagement and outcome for men with co-occurring IPV and substance abuse. Researchers and clinicians have begun to write about the need for interventions for battering and substance-abusing men that address their roles as partners and fathers (Mathews, 1995; Pence et al., 1991; Peled, 2000; McMahon & Rounsaville, 2002; Stover, Meadows & Kaufman, 2009), yet there are no controlled studies that have been conducted to evaluate the efficacy of integrated treatment approaches on reducing violence in families and improving parent-child relationships for battering men and their children. Group treatments for male batterers or substance abusers typically designate some of the sessions to psychoeducation and self-disclosure about the impact of domestic violence or substance abuse on children. However, there are no published research studies that review the use of parenting treatment or a father-child component of batterers’ or substance abuse treatment. Given that dyadic work with mothers and children has been found beneficial for children exposed to IPV (Lieberman et al., 2007) and maltreatment (Toth et al., 2002) implementation and evaluation of such dyadic treatment with fathers bears study. Further attention to the role of men as fathers and the development of appropriate treatments to assist men with co-morbid substance abuse and intimate partner violence to be better parents is needed and will likely have significant impact on the functioning of the entire family. Additional studies with larger samples to replicate and further examine demographic and family contexts that may interact with treatment approaches are needed. Small sample size and limited numbers of participants without children restricted analysis and did not allow for a more nuanced examination of the context of fatherhood (e.g. living with the child, amount of contact, nature of the relationship, age of the child, etc.) and its relationship to treatment efficacy. Small numbers of men with custody of children in each of the treatment conditions did not allow the fatherhood status variable to be broken down further. This is a limitation of this study and further exploration of these issues in future research is warranted. Additionally, studies should focus on better understanding the father-child relationship and whether IPV incidents happened in the presence of children as it is well documented that witnessing IPV has significant implications for children’s symptoms and adjustment (Cummings & Davies, 1994).