Social Processes Explaining the Benefits of Al-Anon Participation

Abstract

This study examined social processes of support, goal direction, provision of role models, and involvement in rewarding activities to explain benefits of participating in Al-Anon, a 12-step mutual-help program for people concerned about another person’s substance use. Newcomers to Al-Anon were studied at baseline and 6 months later, when they were identified as having sustained attendance or dropped out. Among newcomers, and among established Al-Anon members (“oldtimers”), we also used number of Al-Anon meetings attended during follow-up to indicate extent of participation. Social processes significantly mediated between newcomers’ attendance status as sustained (versus dropped out) and outcomes of Al-Anon in the domains of life context (e.g., better quality of life, better able to handle problems due to the drinker), improved positive symptoms (more self-esteem, more hopeful), and decreased negative symptoms (e.g., less abuse, less depressed). Social processes also significantly mediated between newcomers’ number of meetings and outcomes. However, among oldtimers, Al-Anon attendance was not associated with outcomes, so the potential mediating role of social processes could not be examined, but social processes were associated with outcomes. Findings add to the growing body of work identifying mechanisms by which 12-step groups are effective, by showing that bonding, goal direction, and access to peers in recovery and rewarding pursuits help to explain associations between sustained Al-Anon participation among newcomers and improvements on key concerns of Al-Anon attendees. Al-Anon is free of charge and widely available, making it a potentially cost-effective public health resource in helping to alleviate negative consequences of concern about another’s addiction.

Introduction

Mutual-help groups are a key component of the system of care for individuals with substance use disorders and their Concerned Others (COs). Participation in these groups improves the likelihood of achieving and maintaining positive outcomes and reduces the need for further professional care (Krentzman et al., 2010; Magura, McKean, Kosten, & Tonigan, 2013; Ye & Kaskutas, 2009). To better understand the association between mutual-help group participation and improved outcomes, more information is needed about its underlying social processes or “active ingredients.” Accordingly, this study identified social processes that may help explain the benefits of Al-Anon participation. Al-Anon, formally known as Al-Anon Family Groups, is a 12-step mutual-help program for COs, and is the most widely used form of help by COs in the US (Miller, Meyers, & Tonigan, 1999; O’Farrell & Clements, 2012).

Social Processes in Mutual-Help Groups

In a series of papers, Moos (2007, 2008a,b) posited some social processes that may help explain the benefits of mutual-help group participation. These processes include support and bonding; goal direction and structure; the provision of norms and role models; and engagement in rewarding activities. This framework posits that social processes mediate between 12-step group participation and outcomes (Figure 1). To examine social processes as mediators, we utilized data collected from a sample of newcomers to Al-Anon who had either discontinued or sustained meeting attendance at a 6-month follow-up. We also utilized longitudinal data from a sample of established Al-Anon members; for convenience, we refer to this sample using the 12-step term of “oldtimers.”

Figure 1.

Social processes mediate between Al-Anon attendance and outcomes.

Support and bonding, goal direction and structure

In Moos’ framework, support and bonding reflect interpersonal relationships embodying group cohesion. Goal orientation reflects areas in which a group encourages personal growth. Structure covers the extent to which a group conveys clear expectations for individual behavior. Generally, mutual-help groups provide support by emphasizing bonding among members, community building, and members’ common welfare. In addition, these groups have a purpose provided by the Twelve Steps and Twelve Traditions, and so tend to be goal-directed and well-organized. An assessment of these social processes in four Alcoholics Anonymous (AA) groups found moderate to high emphasis on cohesion, goal direction (e.g., independence, self-discovery), and organization (Montgomery, Miller, & Tonigan, 1993).

Norms and role models, rewarding activities

In Twelve-Step Facilitation therapy, the importance of norms and role models is seen in the focus on accepting abstinence as the norm; identifying with individuals, such as sponsors, in recovery; and interacting with role models in abstinence-oriented social networks (Nowinski, 2006). Al-Anon also provides norms and role models because participation involves exposure to members who have created good lives despite having a loved one who is drinking excessively (Barber & Gilbertson, 1997). Mutual-help groups such as Al-Anon are also known for engaging members in rewarding social pursuits such as meetings, parties, and community activities. The 12-step orientation of groups such as Al-Anon highlights the value of lifestyle changes involving participation in substance-free social activities and in actively helping others.

Research findings

In general, research on mutual-help’s social processes has focused on AA participation and increased social support among individuals who received addiction treatment. For example, young adults who had been attending 12-step groups for at least three years valued the fellowship’s connectedness, support, and opportunities to learn (Dadich, 2010). A study of male inpatients found that 12-step group involvement (e.g., attended meetings, read 12-step literature) was associated with more support (more close friends) and role models (more friends who abstained from alcohol and drugs) (Humphreys & Noke, 1997); these findings support Path A in Figure 1. Subsequent studies extended these findings to show that more 12-step group involvement was associated with reduced substance use one year later (Path C in Figure 1), and that this association was mediated by both support and abstinent role models; these findings support Figure 1’s mediational model (Bond, Kaskutas, & Weisner, 2003; Humphreys, Mankowski, Moos, & Finney, 1999; Kelly, Hoeppner, Stout, & Pagano, 2012; Laudet, Cleland, Magura, Vogel, & Knight, 2004).

Present Study

Although some research findings are consistent with the model in Figure 1, there has been little examination of potentially beneficial social processes other than support. In addition, research findings have been based largely on treated substance use disorder patients participating in AA. It is not known whether similar social processes are associated with Al-Anon’s benefits, which differ from those of AA’s emphasis on abstinence. In particular, benefits of Al-Anon involve skills such as learning how to handle problems due to the drinker, achieving a better overall quality of life and well-being, and feeling more hopeful (Timko, Laudet, & Moos, 2014b). The aim of this study was to examine social processes or therapeutic ingredients that may underlie the benefits of Al-Anon participation using Moos’ model of the active social processes of mutual-help groups: support and bonding; goal direction and structure; the provision of norms and role models; and involvement in rewarding activities.

We approached this aim in three ways. First, we utilized data from newcomers to Al-Anon; these newcomers were studied at baseline and 6 months later, when they were identified as having sustained Al-Anon attendance or having dropped out of Al-Anon. We used sustained attendance versus dropped out as the independent variable to examine social processes explaining outcomes (Figure 1). Second, among the same newcomers, we examined number of Al-Anon meetings during the 6-month follow-up as the independent variable. The purpose was to examine the extent to which identified mediational processes were robust across indicators of mutual-help group participation. Third, we used data collected from Al-Anon oldtimers at baseline and follow-up to examine number of Al-Anon meetings during the 6-month follow-up as the independent variable. The intent was to examine the extent to which social processes that explain positive outcomes of mutual-help group participation among people new to the group also explain those of ongoing members. Social processes may have a stronger role in explaining outcomes among newcomers than among established 12-step group members.

Methods

Sample

At baseline, we received 631 surveys from newcomers (N=365, 58.7%) and oldtimers (N=266, 41.3%). Conforming to Al-Anon convention (http://www.al-anon.alateen.org/what-can-i-expect), newcomers were defined at baseline as having attended ≤6, and oldtimers as having attended >6, Al-Anon meetings (lifetime). At baseline, newcomers and oldtimers had attended an average of 3.3 (SD=1.7) and 62.8 (SD=152.0) Al-Anon meetings, respectively. Of 305 newcomers who agreed at baseline to be contacted about the 6-month follow-up, 253 (83%) completed the 6-month survey. Of 215 oldtimers at baseline who agreed at the time of the baseline survey to be contacted about the 6-month follow-up, 184 (86%) completed the follow-up survey.

At baseline, participants were mainly women (84.3%, N=532), white (92.4%, N=583), married or partnered (60.3%, N=381), and employed (61.6%, N=389), with a mean age of 48.6 years (13.3) and a mean of 14.9 years of education (SD=2.1). At baseline, compared to oldtimers, newcomers were somewhat younger and more likely to be employed. Otherwise, Al-Anon attendance status as newcomer or oldtimer was not associated with these baseline characteristics. In addition, these sample characteristics are similar to those reported in Al-Anon’s internal survey of long-term (mean=12 years) members (http://www.al-anon.org/membershipsurvey), in which 86% of respondents were female, 91% white, 58% married, and 56% employed.

2.2 Procedure

2.2.1 Recruitment

Recruitment procedures are fully described by Timko, Cronkite, Kaskutas, Laudet, Roth, and Moos (2013). To acquire the sample, Al-Anon Family Groups World Service Office (WSO) mailed (but did not pay for) a random sample of 4,500 Al-Anon groups. Of 979 groups (22%) responding back to the researchers, 853 (87%) gave permission to be contacted. Of the 853 groups, 672 (85.7%) agreed to participate. Surveys were received from 54% (N=360) of groups that agreed to participate. Individual survey respondents (N=631; mean number per group=1.9; SD=1.2) were offered a $25 gift card.

2.2.2 Follow-up

After 6 months, participants who had agreed to follow-up participation at baseline were mailed a copy of the follow-up survey and contacted by email to let them know the follow-up survey had been mailed to them. As at baseline, participants returned their survey and payment information separately to protect their confidentiality, and received $25 as compensation. Participants who, at baseline, agreed to follow-up and returned completed surveys, agreed to follow-up but did not return surveys, or did not agree to follow-up, did not differ on demographic characteristics (gender, race, marital status, employment status; p>.20).

2.2.3 Surveys

The baseline and 6-month survey items were quite similar and drawn mainly from the Health and Daily Living Form (HDL; Moos, Cronkite, & Finney, 1992), which has demonstrated strong psychometric characteristics in family studies of alcohol use and other mental health disorders (Brennan, Schutte, & Moos, 2010; Timko, Cronkite, Swindle, Robinson, Sutkowi, & Moos, 2009). The follow-up survey’s measures of social processes (Moos & Moos, 2005) and of Al-Anon outcomes were used.

Social processes

On the follow-up survey, respondents described whether, in the past 6 months, they had or had not experienced aspects of Al-Anon: (1) Bonding (sum of 5 items, e.g., You felt a sense of group belonging), (2) Goal direction and structure (sum of 7 items, e.g., The group helped you work toward your important goals, The group meetings tended to be clear and well-organized), (3) Provision of norms and role models (6 items, e.g., You met members who are role models), and (4) Involvement in rewarding activities (6 items, e.g., You shared activities with Al-Anon members outside of meetings); Cronbach’s alphas on these indices, calculated for newcomers and oldtimers separately, ranged from .68 to .87. Because these four social processes were highly intercorrelated (average r=.63 for newcomers, .66 for oldtimers) they were combined into one composite (for newcomers, alpha=.86, M=13.4, SD=6.4, range=0–24; for oldtimers, alpha=.87, M=17.1, SD=5.8, range=0–24).

2.2.4 Al-Anon outcomes

The follow-up survey asked the respondent, for each of 18 potential positive outcomes of Al-Anon attendance, whether or not it was a benefit of attending Al-Anon. Specifically, for each of 18 potential gains, the participant checked “yes, a gain” or “no, not a gain.” These gains involved better life context (N=7), improved positive symptoms (N=5), and lessened negative symptoms (N=6). Table 1 lists 14 of the 18 outcomes; the remainder are described in the Results section.

Table 1.

Regressions Examining Social Processes as a Mediator Between Newcomers’ Al-Anon Attendance (Sustained vs. Dropout) and Outcomes

Outcome	Predictor: Regression 1			Predictors: Regression 3						z
	Attendance			Attendance			Social Processes
	B	OR	CI	B	OR	CI	B	OR	CI
Better life context:
Quality of life, well-being	1.14	3.14**	1.46–6.74	0.38	1.46	0.59–3.64	0.15	1.16***	1.09–1.23	4.63***
Home, neighborhood	0.57	1.77*	1.03–3.04	0.24	1.27	0.69–2.33	0.07	1.07**	1.02–1.12	3.39***
Financial situation	0.64	1.90*	1.07–3.36	0.26	1.30	0.70–2.45	0.07	1.07*	1.01–1.12	3.07***
Do better at work or school	0.60	1.82*	1.05–3.14	0.13	1.14	0.62–2.12	0.07	1.08**	1.03–1.13	3.42***
Physical health	0.68	1.97*	1.14–3.41	0.04	1.05	0.55–2.00	0.13	1.14***	1.08–1.20	4.55***
Handle problems due drinker	1.71	5.52**	2.05–14.82	0.96	2.62	0.79–8.67	0.16	1.18***	1.10–1.26	4.71***
Improved positive symptoms:
Involved in what’s important	1.31	3.71*	1.74–7.92	0.54	1.72	0.71–4.23	0.14	1.15***	1.08–1.22	4.98***
Self-esteem, self-confidence	1.39	4.02***	1.98–8.12	0.64	1.90	0.84–4.35	0.15	1.16***	1.09–1.22	5.09***
Hope	1.29	3.62**	1.69–7.72	0.56	1.75	0.72–4.29	0.14	1.15***	1.08–1.22	4.97***
Lessened negative symptoms:
Verbal, physical abuse	0.63	1.88*	1.08–3.25	0.12	1.13	0.59–2.17	0.13	1.14***	1.08–1.20	4.71***
Depressed, moody	1.19	3.28***	1.75–6.17	0.52	1.68	0.81–3.52	0.16	1.17***	1.10–1.24	5.13***
Angry	1.61	4.99***	2.36–10.53	0.81	2.26	0.94–5.47	0.18	1.20***	1.13–1.28	4.70***
Stress, tension	1.45	4.24***	2.05–8.76	0.81	2.24	0.97–5.22	0.16	1.17***	1.10–1.24	4.36***
Confusion on how to cope	1.68	5.38***	2.38–12.12	0.84	2.33	0.88–6.17	0.18	1.21***	1.13–1.29	4.75***

^*p<.05;

^**p<.01;

^***p<.001

Note: B=regression coefficient, OR=odds ratio; CI=confidence interval; z=Sobel test statistic

2.3 Data Analysis

Newcomers: Dropout vs sustained attendance

Dropout from Al-Anon was defined as not having attended any Al-Anon meetings after Month 5 of follow-up; retention was defined as having attended at least one meeting after Month 5. A total of 23 of the 253 newcomers who returned follow-up surveys did not provide information on their meeting attendance in Month 6, leaving 230 respondents for analyses. Using the definitions, 133 (57.8%) newcomers had dropped out, and 97 (42.2%) had sustained Al-Anon attendance. The 97 with sustained attendance attended a mean of 23.5 meetings (SD=19.5) over the follow-up period, compared to drop-outs’ mean of 6.4 (SD=10.5) (t=−8.47, p<.001).

Newcomers and oldtimers: Number of Al-Anon meetings

Taken together, the 253 newcomers responding to the follow-up survey reported attending an average of 13.4 meetings (SD=16.9) over the 6-month follow-up period. The 184 oldtimers responding to the follow-up survey reported attending an average of 30.9 (SD=31.4) meetings over the follow-up period.

Mediation

To examine mediation (Kenny, 2014; Mackinnon & Dwyer, 1993), first, Al-Anon attendance among newcomers as sustained or dropped out was entered as the independent variable in a regression to predict each of the Al-Anon outcomes (Regression 1; Path C in Figure 1). Secondly, attendance status was entered as the independent variable in a regression to predict the combined social process score (Regression 2; Path A in Figure 1). Thirdly, attendance status and the combined social process score were entered together in a regression to predict the outcome (Regression 3). If the coefficient for the social process score (the potential mediator) was significant and the coefficient for attendance status on the outcome was reduced, then we concluded that a mediating effect was supported, and evaluated whether the indirect effect of attendance on the outcome via social processes was significant using the Sobel test (MacKinnon, Warsi, & Dwyer, 1995; Preacher, 2015). Also to examine mediation, number of Al-Anon meetings attended (reported at the 6-month follow-up) among newcomers, and then among oldtimers, was used instead of the sustained/dropout indicator, followed by Sobel tests when appropriate.

Results

Newcomers: Dropout vs. sustained attendance

Table 1 lists Al-Anon outcomes for which social processes were significant mediators of attendance status (dropout or sustained). For example, sustained attendance increased the odds of having a better quality of life by 3.14 times (Regression 1, Path C in Figure 1). In Regression 2 (Path A), which is not tabled, and is constant for each outcome examined, sustained attendance was associated with higher scores on the social processes composite (B=5.23, p<.001). As shown in Table 1, in Regression 3, attendance status and social processes were entered together to predict better quality of life; Al-Anon attendance was no longer a significant predictor, whereas the social processes score was a significant predictor, indicating mediation. The indirect effect of sustained attendance in Al-Anon among newcomers on better quality of life and well-being via social processes was significantly different from zero according to the Sobel test. The same pattern held for each of the other life context and symptom outcomes listed in Table 1.

Table 1 does not display the three outcomes (the life context factor of helping the drinker, improved meeting of responsibilities, and lessened loneliness and isolation) for which the association with attendance status was not significant (p>.05). Thus, for these three outcomes, there was no association for social processes to potentially mediate.

One additional outcome, having a more satisfying spiritual life, showed a different pattern from all other outcomes. In this case sustained attendance predicted more spiritual satisfaction (B=1.86, OR=6.41, p<.001, CI=2.96–13.88); when attendance status and social processes were both entered to predict spiritual satisfaction, sustained attendance held as a significant predictor of the outcome (B=1.05, OR=2.86, p<.05, CI=1.10–7.47), as did social processes (B=0.19, OR=1.21, p<.001, CI=1.14–1.29). The coefficient for attendance was reduced, indicating partial mediation.

Newcomers: Number of meetings

Table 2 shows results for mediation when the independent variable was newcomers’ number of Al-Anon meetings attended between baseline and the 6-month follow-up. For example, attending more meetings increased the odds of reporting a better quality of life by 1.07 times (Regression 1). In Regression 2 (not tabled; constant for each outcome), more meetings was a significant positive predictor of the social processes composite (B=.163, p<.001). As shown in Table 2, in Regression 3, when number of meetings attended and social processes were entered together to predict better quality of life, number of meetings attended was no longer a significant predictor, whereas the social processes composite was a significant predictor, indicating mediation. The indirect effect of more Al-Anon meeting attendance among newcomers on better quality of life and well-being via social processes was significantly different from zero according to the Sobel test. The same pattern held for the life context factors of improved physical health and handling problems due to the drinker, and for all of the symptom factors listed in Table 2, except verbal and physical abuse.

Table 2.

Regressions Examining Social Processes as a Mediator Between Newcomers’ Al-Anon Attendance (Number of Meetings) and Outcomes

Outcome	Predictor: Regression 1			Predictors: Regression 3						z
	Attendance			Attendance			Social Processes
	B	OR	CI	B	OR	CI	B	OR	CI
Better life context:
Quality of life, well-being	0.07	1.07**	1.03–1.11	0.02	1.02	.98–1.06	0.13	1.14***	1.07–1.22	4.46***
Home, neighborhood	0.01	1.01	0.99–1.02	0.00	1.00	.98–1.01	0.08	1.08***	1.03–1.14	N/A
Financial situation	0.01	1.01	1.00–1.02	0.00	1.00	.98–1.01	0.07	1.07*	1.02–1.13	N/A
Do better at work or school	0.01	1.01	1.00–1.03	0.00	1.00	.98–1.01	0.08	1.09**	1.03–1.14	N/A
Physical health	0.02	1.02*	1.01–1.04	−0.01	.99	.97–1.01	0.14	1.15***	1.08–1.21	5.08***
Handle problems due drinker	0.11	1.11**	1.05–1.17	0.05	1.05	.98–1.12	0.15	1.16***	1.07–1.25	5.00***
Improved positive symptoms:
Involved in what’s important	0.08	1.08***	1.04–1.12	0.03	1.03	1.00–1.08	0.13	1.13***	1.06–1.21	5.71***
Self-esteem, self-confidence	0.08	1.08***	1.04–1.12	0.03	1.03	.99–1.07	0.13	1.14***	1.07–1.21	5.88***
Hope	0.06	1.06***	1.02–1.10	0.02	1.02	.98–1.05	0.14	1.14***	1.07–1.22	5.71***
Lessened negative symptoms:
Verbal, physical abuse	0.02	1.02	1.00–1.04	0.00	1.00	.98–1.01	0.14	1.15***	1.08–1.21	N/A
Depressed, moody	0.07	1.07***	1.04–1.10	0.03	1.03	.97–1.06	0.14	1.15***	1.08–1.22	5.94***
Angry	0.09	1.09***	1.05–1.13	0.03	1.03	.99–1.07	0.18	1.19***	1.11–1.28	4.65***
Stress, tension	0.07	1.07***	1.03–1.10	0.02	1.02	.99–1.06	0.15	1.17***	1.09–1.25	4.83***
Confusion on how to cope	0.10	1.11***	1.06–1.16	0.04	1.04	.99–1.09	0.17	1.19***	1.10–1.28	5.31***

^*p<.05;

^**p<.01;

^***p<.001

Note: B=regression coefficient, OR=odds ratio; CI=confidence interval; z=Sobel test statistic; N/A=not applicable

Again, the associations between number of meetings and the three outcomes of helping the drinker, improved meeting of responsibilities, and lessened loneliness and isolation were not significant (p>.05). Thus, as before, for these three outcomes there was no association for social processes to potentially mediate. As before, having a more satisfying spiritual life showed a different pattern from all other outcomes. More meeting attendance predicted more spiritual satisfaction (B=.15, OR=1.17, p<.001, CI=1.10–1.23); when number of meetings and social processes were both entered to predict spiritual satisfaction, more meetings held as a significant predictor of the outcome (B=.10, OR=1.11, p<.01, CI=1.03–1.18; partial mediation), as did social processes (B=0.15, OR=1.16, p<.001, CI=1.08–1.25).

Oldtimers: Number of meetings

Among Al-Anon oldtimers, there were no significant associations (p>.05) between number of meetings attended and any of the outcomes (Regression 1). However, number of meetings was significantly positively associated with the social processes composite (B=.07, p<.001). When number of meetings and social processes were entered together to predict outcomes (Regression 3), social processes were significantly positively associated with life context factors (better quality of life; home and neighborhood; financial situation; physical health; handling of problems due to the drinker); improved positive symptoms (more involved in what’s important in life; more self-esteem); and lessened negative symptoms (verbal, physical abuse; depressed and moody; stressed and tense; confused about how to cope); Bs ranged from .08 to .14, ORs ranged from 1.08 to 1.15; p<.05.

Discussion

Social processes indexed by a composite of bonding, goal direction, provision of role models, and involvement in rewarding activities (Moos, 2007, 2008a,b) significantly mediated between Al-Anon participation and outcomes among newcomers. Specifically, sustained attendance over the initial 6-month period, in contrast to dropping out, was associated with better outcomes in the domains of life context (e.g., better quality of life and well-being, better able to handle problems due to the drinker), improved positive symptoms (e.g., more involvement in what’s important in life and more self-esteem), and decreased negative symptoms (e.g., less verbal and physical abuse, and less confusion about how to cope with life’s problems). Social processes also significantly mediated between newcomers’ Al-Anon participation as reflected in number of meetings attended over the 6-month period and outcomes, although less robustly than when participation was operationalized as sustained attendance versus dropped out. These findings support the model presented in Figure 1. In contrast to findings for newcomers, among “oldtimers,” there was no significant association between Al-Anon attendance and outcomes and thus no potential for a mediating role of social processes, although social processes were significantly associated with outcomes.

Social Processes and Outcomes among Newcomers

Our finding that Al-Anon newcomers’ sustained attendance was associated with an array of better outcomes within the domains of life context, improved positive symptoms, and decreased negative symptoms is consistent with studies that conceptualize and model recovery as multidimensional, including physical and psychological health and other aspects of daily functioning (Garner, Scott, Dennis, & Funk, 2014; White, Weingartner, Levine, Evans, & Lamb, 2013). In addition, our finding that social processes helped to explain associations between sustained attendance and better outcomes is consistent with studies of individuals in treatment for substance use disorders showing that the benefits of 12-step group participation are due to such social processes as obtaining information about recovery, being exposed to healthy people with positive attitudes and goals, and a sponsor’s provision of direction and commitment (Brown, Tracy, Jun, Park, & Min, 2014). In addition to bonding and goal direction, tangible help from peers (e.g., child care, transportation to appointments) was identified as important to recovery among individuals in treatment and 12-step groups (Tracy, Munson, Peterson, & Floersch, 2010).

Our findings on the mediational role of social processes are also consistent with studies of peer support in mental health treatment; that is, experiential sharing, building community and relationships, and skill building and goal setting help patients feel hopeful and boost their self-esteem (Jacobson, Trojanowski, & Dewa, 2012). There also is evidence that 12-step groups can hinder recovery when bonding is absent within the fellowship and group members are not actively working the 12-step program; for example, sponsors do not follow through with promises or are not honest about their experiences (Brown et al., 2014; Tracy et al., 2010).

One mechanism by which 12-step group participation explains positive outcomes is helping others (Zemore & Pagano, 2008), captured within Moos’ (2007, 2008a,b) model as an aspect of being involved in rewarding activities. The practice of giving back, such as to other newcomers or to sponsors, has been linked to improved self-esteem (Brown et al., 2014; Johansen, Brendryen, Darnell, & Wennesland, 2013; Kubicek, Morgan, & Morrison, 2002). Relationships that focus on how to achieve personal goals in practical ways, in addition to mutual bonding and support, may be most helpful to Al-Anon attendees seeking recovery, because they contribute to both self-efficacy and self-worth (Johansen et al., 2013).

Sustained Al-Anon attendance and Al-Anon’s social processes were independently associated with the outcome of having a more satisfying spiritual life. Krentzman, Cranford, and Robinson (2013) found that mutual-help group participation was a predictor of daily spiritual experiences, but noted that it did not fully mediate between participation and outcomes. Because spirituality is a multidimensional construct, comprising, for example, religious practices, connecting to God, experiencing forgiveness, and having meaning or purpose in life (Kelly, Stout, Magill, Tonigan, & Pagano, 2011; Krentzman et al., 2013), further research is needed to understand what Al-Anon attendees mean by having a satisfying spiritual life and how much different aspects are fostered by Al-Anon participation and social processes.

We also found that Al-Anon participation was not significantly associated with three of the outcomes. In hindsight, the lack of associations fits within Al-Anon’s framework. Sustained attendance was not associated with the life context factor of learning how to help the drinker, perhaps because Al-Anon discourages trying to change the drinker, and advises detachment from the drinker and focus on the self by obtaining help for distress and related psychological difficulties (Al-Anon Family Groups, 1995). Sustained attendance also was not associated with improved meeting of responsibilities, possibly due to Al-Anon newcomers often having a heightened sense of responsibility, especially for the caretaking of others (Young & Timko, in press); indeed, a favorite saying learned in Al-Anon is, “I am not responsible for other people’s happiness,” which implies a need to let go of, rather than take on, responsibility. Finally, sustained attendance was not associated with lessened loneliness and isolation among newcomers, perhaps because new Al-Anon participation over a six-month period is not long enough to overcome these feelings.

Number of Meetings and Outcomes among Newcomers and Oldtimers

Among newcomers, number of meetings attended over the initial six-month period was not as robustly associated with outcomes as was operationalizing Al-Anon participation as sustained or dropped out. These results echo those of long-term, observational studies demonstrating that participation in treatment or mutual-help groups over a longer duration predicts better outcomes more strongly than intensity or amount of participation (Moos & Moos, 2007). Sustained attendance may give newcomers greater opportunity to become more involved in 12-step practices, which also predicts better outcomes even when meeting attendance is considered (Kaskutas, Ammon, & Weisner, 2004; Timko & DeBenedetti, 2007).

Among oldtimers, the number of Al-Anon meetings attended over the six-month study period was not associated with outcomes, but number of meetings was positively associated with Al-Anon social processes, which were associated with better outcomes. Similarly, individuals in recovery who reported particular social experiences as part of mutual-help (e.g., identifying as a member of a specific group) had higher levels of wisdom (operationalized in part as having positive self-esteem, being kind, and demonstrating concern for others); however, meeting attendance was not related to wisdom (DiGangi, Majer, Mendoza, Droege, Jason, & Contreras, 2014). Possibly, as duration of Al-Anon participation lengthens, aspects of membership taking place outside of meetings, such as working on the steps, or serving as a group representative, gain in importance in the maintenance of positive outcomes. Indeed, long-term studies of AA have found that meeting attendance is not synonymous with engagement with AA, such as feeling like a member of the fellowship (Kaskutas, Ammon, Delucchi, Room, Bond, & Weisner, 2005).

Limitations

A limitation of this study was that the same individual reported Al-Anon participation, social processes, and outcomes. In addition, participation, social processes, and outcomes were assessed at follow-up, precluding conclusions about causality among these constructs; however, participation and social processes were measured with respect to the preceding 6 months, whereas outcomes were measured with respect to the present. Because recovery is a dynamic process involving mutual-help group participation, social processes, and outcomes reciprocally influencing each other over time, future studies should utilize multiple follow-ups with multiple reporters of Al-Anon attendees’ life context and negative and positive symptoms.

Another limitation was that participants self-selected into dropout vs. sustained attendance groups (and number of meetings attended). Therefore, it is possible that factors not examined in this study influenced how long people stayed in Al-Anon, and also had an impact on the measures of process and outcome. Indeed, a previous report on newcomers in this sample found that, at baseline, individuals who later dropped out of Al-Anon reported less severe problems than individuals who continued to attend, but were more often concerned about their drinker’s psychological health (Timko, Laudet, & Moos, 2014a). In this regard, we note the debate about whether observed correlations between AA participation and better outcomes reflects AA’s effectiveness or are an artifact of self-selection, which has spurred progressively more sophisticated and rigorous statistical analyses. The major result is that AA participation has a genuine benefit that is not attributable to self-selection bias (Humphreys, Blodgett, & Wagner, 2014); future research on Al-Anon may find a similar result.

Although participants were from every state in the US and had demographic characteristics similar to those in Al-Anon’s internal membership survey, the representativeness of our sample of all Al-Anon attendees is unknown. A final limitation is that we conducted analyses for multiple outcomes such that our findings require replication.

Conclusions

Despite these limitations, this study is important in identifying potential “active ingredients” or mechanisms that may help to explain the benefits of participation in Al-Anon and other 12-step groups. These mechanisms representing bonding, goal direction, and access to role models and rewarding activities should be examined in more detail in longitudinal studies measuring social processes and outcomes using different methods and at different time points. This study adds to the growing body of work identifying mechanisms by which 12-step groups are effective (Kelly, Magill, & Stout, 2009). It suggests that bonding, goal direction, and exposure to sponsors and peers in recovery help to explain associations between Al-Anon participation among newcomers and improvements on key concerns of Al-Anon attendees (Timko et al., 2013), such as improved quality of life and learning how to handle problems due to the drinkers in their lives. Al-Anon is free of charge and widely available, making it a potentially cost-effective public health resource to help alleviate the burden of concern about someone else’s addiction, by means of the social processes it offers.