Abstinence Social Support: The Impact of Children in Oxford House

Abstract

The present study compared the characteristics of individuals living with (42 men, 52 women) and without children (561 men, 241 women) residing in an communal-iving recovery program called Oxford Houses.. Results indicated that men living with children and women living without children had more general social support, compared to men living without children and women living with children. Additionally, women and residents of adult-only houses reported having more drug users in their social networks. However, men and women living with and without children reported similar levels of social support for abstinence. It is suggested that that men in recovery who take care of their children are in situations more advantageous to sustained recovery and have more resources compared to recovering women with children. Women in substance abuse recovery and taking care of children may require additional resources and assistance compared to men.

Substance-related disorders are most prevalent during young adulthood, and approximately half of substance-abusing individuals who seek treatment are parents (McMahon, Winkel, Luthar, & Rounsaville, 2005; Meier, Donmall, & McElduff, 2004; Stewart, Gossop, & Trakada, 2007). However, researchers generally neglected the experiences of substance-using parents and their children (Suchman & Luthar, 2002). Research on parenting among men who abuse substances has been particularly rare (McMahon et al.).

The presence of children is often thought to add an extra burden to individuals in recovery, and recovering women in particular. Substance related disorders are more prevalent among men; thus, more men than women in recovery report being parents (McMahon et al., 2005). However, a greater proportion of women seeking substance abuse services are parents, and women in recovery are more likely to have custody of their children (McMahon et al.; Meier et al., 2004; Stewart et al., 2007). Nonetheless, these women are less likely then men to be supported throughout treatment (Reed, 1985). In many cases, alternative child care is too expensive or simply not available to women, especially those with lower socio-economic status, and few substance abuse treatment programs provide child care options or allow children into the program (Nelson-Zlupko, Kauffman, & More, 1995). Furthermore, addicted women are frequently discouraged by family members from seeking treatment due to the concern that treatment could interfere with caring for their family (Nelson-Zlupko et al.).

Studies indicated that children may provide added motivation to stop using and may serve as sources of support for abstinence (Luthar, D’Avanzo, & Hites, 2003; McMahon, Winkel, Suchman, & Luthar, 2001). Christensen (1999) found that children of alcoholic parents frequently provide specific social support for abstinence through attempting, generally unsuccessfully, to convince their parents to quit drinking. Koski-Jännes (1991) suggested that the relationship between having children and reduced drinking results from the social support provided by the children A child’s response to his or her parent’s sobriety may provide more meaningful social support than that of the spouse or other adults (Koski-Jännes). Finally, children were cited by parents in recovery as the number one relationship that helped them decide to enter treatment (Mays, Beckman, Oranchak, & Harper, 1994).

In addition, parents often receive outside social pressures discouraging substance use. For example, only 3% of Ontarian adults in 1992 felt that it was acceptable for parents to drink enough to be slightly intoxicated in front of their children, and 53% felt that consuming alcohol in front of children is never appropriate (Ferris, Templeton, & Wong, 1994). Gullestad (1984) found that blue-collar mothers in Bergen who were young and generally divorced received great pressure from their neighbors to abstain. Additionally, the recent push to avoid drinking during pregnancy may provide additional pressure for abstinence among parents, especially mothers (Room, 1996).

In recent years, substance abuse treatment programs addressed the needs of recovering mothers by including children in both residential and outpatient interventions (Dawe, Harnett, Rendalls, & Staiger, 2003; Knight, Hood, Logan, & Chatham, 1999; Wexler, Cuadrado, & Stevens, 1998; Wobie, Eyler, Conlon, & Clarke, 1997; Worley, Conners, Crone, Williams, & Bokony, 2005). These programs provided a variety of services including parenting education and child-focused interventions. A number of studies indicate that including children in their mothers’ treatment leads to better retention and outcome among women in recovery (Conners, Grant, Crone, & Whiteside-Mansell, 2006: Hughes, Coletti, Neri, Urmann, Stahl, Sicilian, & Anthony, 1995; Stevens & Patton, 1998; Szuster, Rich, Chung, & Bisconer, 1996).

In addition, having children may relate to greater success in substance abuse treatment. A Finnish study (Koski-Jännes, 1991) found that compared to women without children, those who had children tended to have more days abstinent before treatment, stay in treatment longer, and have greater treatment compliance. In fact, living with children was the strongest predictor of recovery in their study, even stronger than the number of children, or having a partner (Koski-Jännes).

An example of a supportive community-based recovery home for individuals dealing with substance abuse problems is Oxford House (OH; see Jason, Ferrari, Davis, & Olson, 2006). A low cost, self-run, democratic recovery home model, Oxford House has grown since 1975 to over 1,200 homes across the USA, 30 in Canada, and eight in Australia. Regarding the operation and maintenance of Oxford Houses, no professional staff is involved, enabling residents to create their own rules for communal governance. Residents live together in a democratic, single-sex home and provide each other with a supportive abstinent mutual-support network. The residents are allowed to stay indefinitely, provided that they pay rent, abstain from alcohol and drug use, and avoid disruptive behavior. Failure to comply with these guidelines is grounds for expulsion from the House.

Regarding parenting practices, Oxford Houses are single-sex dwellings, and a number of Houses allow mothers and fathers to live with their minor children (Oxford House, 2003). Fathers or more typically, mothers live in an Oxford House with one or more of their children along with other residents who have no children in the House. For example, in Northern Illinois, eight women on average live in a House, and up to four of those women may have children with them. This arrangement allows for greater financial stability and prevents overcrowding (Paul Molloy, personal communication, December, 2006). The other House residents are expected to take an active role in helping take care of the children in the House. Oxford House also generally requires mothers to take some form of parenting class outside of the Houses.

Several research studies have focused specifically on parents and children within Oxford Houses. Multiple studies demonstrated that the children in Oxford House have a positive effect on the recovery of the adult women residents (d’Arlach, Olson, Jason, & Ferrari, 2006; Kim, Davis, Jason, & Ferrari, 2006). This positive effect was identical for both mothers and non-mothers, possibly because having children present leads to increased responsibility among all house residents, aiding in recovery (d’Arlach et al.). These women also reported experiencing a high sense of satisfaction with the Houses and stated that residents provided one another with support for abstinence and parenting. Finally, while there are several Oxford Houses for fathers with children, the experience of these male residents has not been studied.

Social support plays a significant role in the effectiveness of Oxford House (Groh, Jason, Davis, Olson, & Ferrari, 2007; Jason, Davis, Ferrari, & Anderson, 2007). Regarding general social support, Oxford House residents rated "fellowship with similar peers" the most important aspect of living in an Oxford House (Jason, Ferrari, Dvorchak, Groessl, & Molloy, 1997). Oxford House also provides residents with abstinent-specific social support networks consisting of other residents in recovery (Flynn, Alvarez, Jason, Olson, Ferrari, & Davis, 2006). Longer lengths of stay in Oxford House related to less support for substance use (Davis & Jason, 2005) and increased support for abstinence (Majer, Jason, Ferrari, Venable, & Olson, 2002). Furthermore, Oxford House residents whose social networks provided less support for substance use were more likely to remain abstinent (Jason et al., 2007). Other studies report that women who have supportive relationships are more likely to complete substance abuse treatment and have better outcomes (Coughey, Feighan, Cheney, & Klein, 1998; Knight, Hood, Logan, & Chatham, 1999). Finally, abstinence-specific social support has been found to be a particularly strong predictor of long-term abstinence following treatment (Longabaugh, Wirtz, Beattie, Noel, & Stout, 1995; Longabaugh, Wirtz, Zweben, & Stout, 1999).

The current study explored the experiences of different types of general and substance use-specific support among men and women living with and without children in Oxford House. Based on research indicating that mothers have better outcomes when living with their children during substance abuse treatment, we hypothesized that both men and women living in Oxford Houses with children would experience more general social support and more support for remaining abstinent than men and women living without children in Oxford House.

Methods

Procedure

Analysis of data provided by Oxford House, Inc. using a geographic information systems (GIS) program indicated that Houses clustered in the states of Washington, Oregon, Pennsylvania, New Jersey, North Carolina, Illinois, and Texas. Participants were recruited either by research staff who visited 170 Oxford Houses in these states or at the 2001 Oxford House World Convention. After explaining the study to participants and securing informed consent, research assistants administered the study’s measures in a group format. Research assistants were available to answer questions while participants completed the paper and pencil measures (see Jason et al., 2007). The current investigation reports baseline data for the longitudinal study which collected data every four months for a period of approximately on year (Jason et al.).

Participants

Participants included 42 men and 52 women living in Oxford Houses with children and 561 men and 241 women living in Houses without children. Participant demographic characteristics are presented in Table 1. Among participants residing in Houses with children, 52 (55%) were from Washington or Oregon, 24 (26%) resided in Houses in Pennsylvania or New Jersey, 12 (13%) lived in the Midwest or Texas, and 6 (6%) resided in North Carolina. Among participants living in Houses with no children, 280 (34%) lived in North Carolina, 190 (24%) resided in Washington or Oregon, 168 (21%) lived in the Northeast, and 165 (21%) lived in the Midwest or Texas.

Table 1.

Mean Demographics for Men and Women Living With and Without Children

Descriptor Variables	Children	No Children
Age:
Males	40.85	39.23
Females	34.50	36.89
Monthly Income:
Males	1267.62	1078.61
Females	538.79	792.22
Education (years):
Males	12.78	12.75
Females	12.26	12.39
Length of Stay in OH (years):
Males	1.86	.925
Females	.723	.718

In terms of ethnicity, 62 (66%) participants living in Houses with children were European Americans, whereas 18 (19%) were African Americans, 3 (3%) were Hispanics/Latinos, and 11 (12%) represented other ethnicities including Asian Americans, American Indians, and biracial or multiracial individuals. Among participants living in Houses without children, 462 (58%) were European Americans, 287 (36%) were African Americans, 28 (3%) were Hispanics/Latinos, and 26 (3%) represented other ethnicities. Approximately 50% of participants in both types of houses were never married, 45% divorced, separated or widowed, and 5% married.

ANOVAS were run to test differences in age, years of education, monthly income, and time in Oxford House at baseline. A significant main effect for gender, F(1,890) = 18.35, p < .001, and a significant interaction between gender and house type, F(1,890) = 3.87, p < .05, was found for participants’ age. A significant main effect for gender, F(1,868) = 3.91, p <. 05, was found for years of education. ANCOVAS, controlling for years of education, indicated a significant main effect for gender, F(1,842) = 25.13, p < .001, and a significant interaction between gender and house type, F(1,842) = 4.31, p < .05, for monthly income. A significant main effect for gender, F(1,884) = 24.02, p < .001, and house type, F(1,884) = 11.48, p < .001, was found for time in Oxford House. The interaction between gender and house type also was significant, F(1,884) = 11.20, p < .001, for time in Oxford House at baseline.

Measures

Participants completed the Addiction Severity Index-Lite (ASI; McLellan, Kushner, Metzger, Peters, Smith, Grissom, Pettinati et al., 1992), a valid and reliable measure of lifetime and recent of substance use and related medical, psychological, family, employment, and legal problems. The ASI collected demographic and treatment history data and provided seven valid and reliable composite scores (i.e., drug, alcohol, medical, psychological, family, legal, and employment) based on reports of problems during the 30 days prior to scale administration. In the current study, the ASI was used to collect demographic data.

Finally, a modified version of the Important People and Activities Inventory (IPA; Clifford & Longabaugh, 1991) with the activities items omitted (i.e., the IP) was administered to assess social support variables. Additionally, the modified version of this measure administered for the current study included question regarding alcohol as well as illicit drug use. The IP requires respondents to identify important members in their networks with whom they have had frequent contact within the past 6 months. For each person the participant lists in his/her network, the scale examines the type of relationship (e.g., spouse, parent, or friend), the duration of relationship in years, and the frequency of contact. In addition, the participant reports how often the network member drinks, how much the network member drinks on a maximum drinking day, and the network member’s overall drinking status (i.e., heavy, moderate, light, abstainer, or recovering). The IP was used in several studies including Project MATCH, and demonstrated good test-retest reliability and construct validity (Beattie, Longabaugh, Elliott, Stout, Fava, & Noel, 1993; Longabaugh et al., 1995; Longabaugh, Wirtz, Zweben, & Stout, 1998). The current study scored the IP according to the three factors derived by Groh and his colleagues: general social support, drinking behaviors of network members, and support for drinking from network members (Groh, Olson, Jason, Ferrari, & Davis, in press). Higher scores on the IP indicate greater support for substance use.

Results

Differences in general social support between men and women living in Oxford Houses with and without children were examined using an ANOVA. Distributions for drinking behaviors of network members and support for substance use among network members were skewed. Therefore, these scores were dichotomized into “high” and “low” based on a median split and analyzed using chi-square analyses.

In terms of general social support, a significant interaction was found between house type and gender, F(1, 896) = 9.89, p < .001, such that men living in Houses with children and women living in Houses without children had higher levels of general social support compared to men living without children and women living with children. Chi-squares found no significant association between gender or house type and drinking behaviors of network members. However, a significant association was found between the illicit drug use behaviors of network members and both gender, X² (1, 871) = 12.25, p < .001, and house type, X² (1, 871) = 3.87, p < .05, such that women and residents of Houses without children had more individuals who used illicit drugs in their social networks. No significant association was found between gender or house type and support for drinking or illicit drug use.

Discussion

This study explored demographic and social support differences among men and women living in Oxford Houses with and without children at the baseline of a longitudinal, national study of Oxford House residents. Results indicated a number of important between group differences that may have implications for recovery. For instance, men were significantly older than women. Although women with children comprised the youngest group of participants, men with children were the oldest group. Men had significantly more years of education than women along with higher incomes, even after controlling for years of education. Furthermore, women with children had the lowest whereas men with children had the highest incomes of all groups. These differences are important because age, education, and employment status have been found to predict recovery outcomes (Brewer, Catalano, Haggerty, Gainey, & Fleming, 1998; Scott-Lennox, Rose, Bohlig, & Lennox, 2000). Significant gender and house type differences were also found in terms of time in Oxford House: men living in Oxford Houses with children had stays that were over twice as long as those of the other groups. All of the findings indicate a number of advantages for men compared to women living in Oxford Houses with children.

Analysis of baseline social support data indicated that men living in Oxford Houses with children and women living in Houses without children had more general social support that men living without children and women living with children. This is consistent with past research suggesting that women are less likely than men to be supported throughout treatment (Reed, 1985), and that women are often discouraged by from seeking treatment because it might interfere with child care (Nelson-Zlupko et al., 1995). However, no significant group differences were found in terms of abstinence-specific social support for either alcohol or illicit drug use. In terms of the composition of social support networks, no group differences were found regarding the alcohol use habits of friends and family members. However, results indicated that women and residents of Oxford Houses without children had more illicit drug users in their social networks. Thus, having children may help protect against maintaining relationships with negative drug-using friends and family members. However, these findings are partially in contrast to research suggesting that compared to men, recovering women with children might have more abstinence support in their social networks (Room, 1996), which can include avoiding friends and family who use drugs.

In summary, the results of this investigation suggests that men in recovery who take care of their children are in situations more conducive to sustained recovery and have more resources (i.e., they are older, more educated, have higher incomes, and have longer lengths of stay in OH), including positive social supports (i.e., they receive more general support and have fewer drug users in their social networks), whereas women in recovery who take care of their children have the least of these types of resources. It is possible that recovering women who have children are frequently forced to take care of their children because no one else is available to take on these child care duties; on the other hand, recovering men may be more likely to take care of their children when they are doing well in recovery and elect to take on this responsibility. Therefore, it is suggested that recovering women taking care of children may require additional resources compared to recovering men taking care of children. It may be important for the Oxford House organization to provide more assistance when opening homes for women with children. In addition, clinicians and treatment providers working with women in recovery with children may want to address these additional struggles and help these individuals develop supportive social networks.