A Categorical Assessment of 12-Step Involvement in Relation to Recovery Resources

Abstract

The present study examined 12-step involvement categorically, independent of 12-step meeting attendance, and its relation to recovery resources among 100 members of Alcoholics Anonymous and Narcotics Anonymous residing in recovery homes. Participants who were actively involved in 12-step activities reported significantly higher levels of recovery resources compared with those who were less involved, whereas meeting attendance was significantly and negatively related to self-efficacy for abstinence and meaning in life. Findings suggest that categorical involvement in 12-step activities equip recovering alcoholics/addicts with resources for their ongoing recovery. Implications for future research are discussed.

KEYWORDS:

• 12-step involvement
• meaning
• recovery resources
• self-efficacy
• Oxford House

Vaillant (1995) concluded that relatively short-term substance abuse treatment effects are often not sustained during longer periods and that the most effective interventions might be those that engage clients (Moos, 1994) and promote naturally occurring healing processes. Non-treatment factors like participation in 12-step fellowships such as Alcoholics Anonymous (AA) and Narcotics Anonymous (NA) might be some of the best prognosticators of future recovery status (Humphreys et al., 2004; Moos & Moos, 2004; Owen et al., 2003). Research studies involving recovering addicts/alcoholics have shown participation in 12-step fellowships to be a significant predictor of decreased stress in early recovery (Laudet & White, 2008) and reduced alcohol/drug use at 2-year (McKellar, Stewart, & Humphreys, 2003), 5-year (Ritsher, McKellar, Finney, Otilingam, & Moos, 2002) and 16-year follow-up (Moos & Moos, 2007).

Participation has been assessed in several ways, including measuring 12-step meeting attendance rates in terms of frequency and duration (Moos & Moos, 2004, 2007; Ritsher et al., 2002) and consistency (Weiss et al., 2005). However, participation in 12-step fellowships might be best considered a construct that involves meeting attendance in addition to one's involvement in such fellowships—in other words, one's participation in other 12-step-related activities/behaviors (Tonigan, Connors, & Miller, 1998). A multifaceted approach to measuring involvement in 12-step activities, such as having a sponsor, reading 12-step literature, involvement in service, and “working the 12 steps,” has been applied during the past decade.

For example, the Alcoholics Anonymous Involvement scale (Tonigan et al., 1998) and Alcoholics Anonymous Affiliation Scale (AAAS; Humphreys, Kaskutas, & Weisner, 1998) are two important and widely used instruments in assessing 12-step involvement. These and similar measures (Allen, 2000) are helping researchers investigate the effects of 12-step fellowships that evidence suggests empower recovering individuals both as a social resource for abstinence (Moos & Moos, 2007) and as a catalyst of protective resources for recovery (Moos & Moos, 2004). Recovery is a concept that involves more than abstinence, and involvement in 12-step activities is likely to increase one's resources for ongoing recovery (Laudet & White, 2008). However, the assessment of 12-step involvement has not been consistent across investigations.

On the one hand, a number of studies have examined 12-step meeting attendance as an integral part of 12-step involvement (e.g., Humphreys et al., 1998; Mankowski, Humphreys, & Moos, 2001; McKellar et al., 2003; Moos & Moos, 2007; Morgenstern, Labouvie, McCrady, Kahler, & Frey, 1997; Staines et al., 2003; Tonigan et al., 1998; Zenmore, Kaskutas, & Ammon, 2004). On the other hand, fewer investigations have examined involvement apart from 12-step meeting attendance. For instance, increased and consistent involvement in 12-step activities within 3 months time, independent of meeting attendance, was associated with decreased cocaine use at 6-month follow-up (Weiss et al., 2005).

Montgomery, Miller, and Tonigan (1995) found 12-step involvement in terms of one's extent of working the 12 steps of AA significantly predicted increased rates of meaning in life in addition to abstinence from alcohol at 31 weeks of follow-up, even when controlling for meeting attendance. However, when controlling for 12-step involvement, increased meaning in life was found to be associated with increased odds of no heavy drinking at 6-month follow-up (Robinson, Cranford, Webb, & Brower, 2007). Taken together, findings from these investigations suggest that involvement in 12-step fellowships, independent of 12-step meeting attendance, empowers recovering individuals with abstinence while also providing increased life meaning that might serve as a protective resource for ongoing recovery.

In addition, a lack of 12-step involvement may increase the risk of relapse. Timko and DeBenedetti (2007) found increased involvement in 12-step practices (independent of meeting attendance) was associated with an 83% likelihood of abstinence at both 6- and 12-month follow-up intervals, whereas the likelihood of abstinence was only 26% among persons who had no such involvement. Furthermore, specific 12-step activities, such as having a sponsor and being of service, might be key components of 12-step involvement (Witbrodt & Kaskutas, 2005). Recovering individuals who are actively involved in a number of 12-step activities are likely to have better outcomes related to abstinence, such as self-efficacy for abstinence.

Self-efficacy for abstinence is a construct based on Bandura's (1997, pp. 357–368) cognitive behavior self-efficacy theory, and it is regarded as a crucial resource toward relapse prevention (Marlatt & Gordon, 1985) that can help substance abusers cope with high-risk situations (Annis & Davis, 1991). In an 8-year outcome study, Moos and Moos (2004) found continued 12-step attendance was related to increased abstinence and self-efficacy for abstinence. In addition, Majer, Jason, and Olson (2004) found increased self-efficacy for abstinence was associated with increased abstinence and that efficacy expectations stabilize at 6 months into abstinence among two independent samples of recovering addicts and alcoholics. Furthermore, longitudinal data have shown that both increased self-efficacy for abstinence within 12 months of abstinence and increased duration of AA attendance significantly predicted less alcohol consumption at 16 years of follow-up (Moos & Moos, 2007). However, it is not clear whether increased 12-step involvement in activities apart from meeting attendance predicts increases in self-efficacy for abstinence and continuous abstinence.

In summary, involvement in 12-step activities has been assessed as a multifaceted construct that includes 12-step meeting attendance as a component. Alternatively, 12-step involvement that is independent of 12-step meeting attendance has been demonstrated to promote abstinence—perhaps the most valuable but not necessarily the only resource for recovery. In addition, little is known about whether 12-step involvement apart from meeting attendance engenders additional resources for recovery such as meaning in life and self-efficacy for abstinence for both recovering addicts and alcoholics. Specific activities, such as having a sponsor, working the 12 steps, and being of service, are some activities that have been used to indicate 12-step involvement.

However, nearly every investigation involving 12-step involvement cited in this report used a summary-score approach in assessing participation across various 12-step activities whereby involvement in any specific 12-step activity could have been artificially reduced or inflated by averaging effects. In addition, abstinence outcomes in a number of studies consisted of examining alcohol and drug use independently but not total continuous abstinence. Only one study (Timko & DeBenedetti, 2007) found more overall 12-step involvement was related to better outcomes, with two specific activities (having a sponsor, having had a spiritual awakening) associated with abstinence from both alcohol and drugs. There is a need to determine whether categorical involvement in specific 12-step activities leads to increases in recovery resources including continuous abstinence, as findings would have implications for treatment providers serving recovering addicts/alcoholics.

The present study examined 12-step involvement among a sample of both recovering addicts and alcoholics residing in Oxford Houses, which are self-run, communal-living settings quite similar to the recovery home modality of care except that they are nonprofessional and democratically operated (Jason, Davis, Ferrari, & Bishop, 2001). We hypothesized that high levels of recovery resources (continuous abstinence, self-efficacy for abstinence, meaning in life, and length of stay in an Oxford House) would be observed among participants who were actively involved (as opposed to less involved) in 12-step activities independent of their 12-step attendance.

METHODS

Participants

A total of 100 participants (57 men, 43 women) with a mean age of 37.5 years (SD = 8.6) were involved in the present study. In terms of racial background, 69% were African American, 28% were Anglo-American (i.e., “White”), 2% were Latino, and 1% was multiracial. The majority identified themselves as being single (60%) high school graduates (87%) who were employed full time (79%). They also identified themselves as members of NA (72%), AA (23%), or both (2%). Their most frequent drugs formerly used were cocaine (79%), alcohol (64%), cannabis (28%), opioids/heroin (17%), sedatives/hypnotics (4%), and amphetamines (3%). Intravenous drug use was reported by 17% of all participants. The average length of stay in an Oxford House reported by participants was 255 days (SD = 376 days), and the average rate of continuous abstinence from both alcohol and drugs reported was 422 days (SD = 560 days). In addition, participants reported attending an average of 4.4 (SD = 2.0) AA and/or NA meetings each week.

Procedures

The investigation was proposed to and approved by an institutional review board. Participants for the present study were recruited through two methods that yielded a total sample of 100 participants (also described in the parent study, Majer, Jason, Ferrari, Venable, & Olson, 2002). The first method involved recruiting Oxford House residents from the Mid-Atlantic region of the United States. Ninety questionnaires were originally mailed in March 1999 to the cofounder of Oxford House (J. Paul Molloy) after the primary investigator discussed the overview of the study and how to handle self-report surveys. Mr. Molloy agreed to disseminate and return by mail sets of questionnaires throughout Oxford Houses located in North Carolina and Virginia with the assistance of two Oxford House representatives from this region. These representatives recruited participants at monthly Oxford House business meetings. At the regional business meetings, representatives gave an overview of the study and disbursed packets of surveys to interested persons. Forty-five participants (23 men, 22 women), from a pool of a possible 90 living in the Mid-Atlantic region, agreed to participate in the study. Participants completed their measures in small groups within their homes within 1 week of receiving survey packets when their representatives collected each House's surveys and mailed them to the primary investigator using this method (50% return rate).

The remainder of the present sample (n = 34 men, 21 women) was obtained with the assistance of an Oxford House representative by recruiting participants from a pool of a possible 120 residents (45.8% return rate) from the Midwestern region (Northern Illinois) in a manner similar to the Mid-Atlantic region. Residents from both regions completed their questionnaires in approximately 90 minutes in small groups. All participants were given questionnaires and a consent form and were engaged in a process of informed consent during a mandatory weekly Oxford House meeting or during a monthly business meeting from a regional representative. Oxford House representatives informed all potential participants that their involvement was voluntary and that their responses to items would be confidential and would not impact their status as a resident. Those who participated in the study signed their consent forms and received an educational feedback form (describing the study and providing contact information for the primary investigator) after being engaged in a process of informed consent.

Measures

Twelve-Step Involvement. Participants completed a Basic Information Survey (27 items) and a Treatment Involvement Survey (21 items) developed as a way of gathering relevant sociodemographic and substance abuse information. Items included surveying participants’ continuous days of abstinence from both alcohol and drugs, drugs of choice, route of drug administration, length of stay in an Oxford House, number of AA/NA meetings attended per week, and whether they identified more closely with one 12-step fellowship (i.e., AA or NA). Self-reports on Oxford House residents’ abstinence rates have been verified with collateral informants in one national longitudinal investigation (Jason, Davis, Ferrari, & Anderson, 2007) that found very high rates of consistent reporting—greater than 97% for abstinence from both alcohol and drugs—between residents and informants.

In addition, four questions with a dichotomous response range (yes/no) were given to measure participants’ active involvement with 12-step activities: “I have a sponsor who knows something about ‘the 12 steps;’” “I utilize a network of AA/NA members outside of meetings;” “I am involved with service (e.g., setting up meeting, coffee maker, secretary, etc.);” and “I have a home group that I attend regularly.” The primary author designed these questions as a result of his clinical and research experiences with recovering addicts and alcoholics, and they are very similar to those listed in 12-step involvement measures. Participants who positively endorsed all four 12-step involvement items (n = 42) were considered to have been “actively involved,” whereas participants who did not (n = 53) were considered to have been “less involved.” The internal consistency for 12-step involvement in the present study was adequate for a research study (Cronbach's alpha = .65).

An independent samples t-test was conducted to determine whether reported 12-step involvement substantially differed between these groups, with a total 12-step involvement scale score (calculated by adding the number of positively endorsed items; ranging from 0 to 4) as the dependent variable. After controlling for violation of homogeneity of variance (Levene's test for equality of variance; F(93) = 59.29, p < .001), actively involved participants (n = 42) reported significantly more total 12-step involvement than less involved (n = 53) participants, M = 4.00 vs. 2.08; SE = 0.00 vs. 0.13; t(52) = 15.28, p < .001, CI = 1.67, 2.18.

Abstinence Self-Efficacy. Annis and Graham's (1988) Situational Confidence Questionnaire-39 (SCQ-39) was completed as an assessment of participants’ confidence in resisting the urge to use drugs or alcohol across 39 hypothetical situations. The SCQ-39 is rooted in Bandura's (1997) cognitive-behavioral self-efficacy theory, and it is based on antecedents of substance abuse relapse (Annis & Graham). Previous research indicates that the SCQ-39 was appropriate for the present study, because it has been used among people with different addiction typologies (Ross, Filstead, Parrella, & Rossi, 1994). Because the SCQ-39 consists of eight highly reliable subscales (.81 to .97; Annis & Graham), a total confidence score was used in the present study by collapsing the subscale scores and averaging these scores on a scale that ranges from 0% (not at all confident) to 100% (very confident). This total score calculation has been effectively used in previous studies that compared short- and long-term abstinence groups (Greenfield et al., 2000; Majer et al., 2004; Miller, Ross, Emmerson, & Todt, 1989). The SCQ-39 had a Cronbach's alpha of .99 with the present sample.

Life Purpose. Hablas and Hutzell's (1982) Life Purpose Questionnaire (LPQ) was administered to measure the extent of participants’ sense of meaning in life. Responses (disagree-agree) to the 20 items of the LPQ are scored on a dichotomous response range (0 to 1). Positively endorsed items are summed, with higher scores reflecting a greater sense of meaning in life. The LPQ has good internal consistency (Cronbach's alpha = .82; Hutzell, 1989), and excellent (.90) 1-week test-retest correlation (Hablas & Hutzell). The LPQ had a Cronbach's alpha of .81 in the present sample.

Data Analysis

Multivariate analysis of covariance (MANCOVA) was conducted to test for differences in recovery resources (i.e., number of continuous days abstinent, self-efficacy for abstinence, meaning in life, and number of days living in an Oxford House) while controlling for 12-step attendance rates (weekly average) in relation to one factor: involvement in 12-step activities. This method was used to control for potential shared variance among covariate and dependent variables in addition to control for Type I error that might have resulted through repeated one-way analysis of variance (ANOVA) testing.

RESULTS

There were no significant differences observed between participants in terms of their region (Mid-Atlantic vs. Midwestern) on the following variables: sex, ethnicity, marital status, parental status, number of children, employment, age, education, income, drug of choice, time in Oxford House, continuous abstinence, weekly AA/NA meeting attendance, and 12-step involvement. Therefore, we collapsed across regional samples for all further analyses. In addition, there were no significant differences in the proportion of participants who were actively involved or less involved in relation to their identification with AA, NA, or both AA/NA fellowships, X ² (2, N = 95) = 1.79, p < .41.

Multivariate Analysis of Recovery Resources

Differences in levels of recovery resources were compared among participants in relation to their active involvement in 12-step activities. A one-way MANCOVA was employed (actively involved in 12-step activities, less involved in 12-step activities), covarying 12-step meeting attendance, with continuous days abstinent, levels of self-efficacy for abstinence and meaning in life, and days living in an Oxford House as dependent variables.

Main Effects

Results from the MANCOVA test demonstrated a significant main effect for 12-step involvement, Wilks’ λ (4, 82) = 4.43, p < .01, ηp² = .18. Follow-up ANOVA tests revealed actively involved participants (n = 38) reported significantly higher levels of abstinence, M = 506.22 vs. 271.26; SE = 69.12, 60.05; F(1, 85) = 6.43, p < .01, ηp² = .07, and self-efficacy for abstinence, M = 82.46 vs.70.21; SE = 3.77, 3.28; F(1, 85) = 5.87, p < .02, ηp² = .07, compared with less involved participants (n = 50). In addition, actively involved participants reported significantly higher levels of meaning in life, M = 14.90 vs.12.40; SE = 0.60, 0.52; F(1, 85) = 9.58, p < .01, ηp² = .10, and days in Oxford Houses, M = 316.58 vs.164.60; SE = 51.10, 44.39; F(1, 85) = 4.92, p < .03, ηp² = .06, compared with less involved participants.

Covariate Effects

Results from the MANCOVA test also demonstrated a significant covariate effect for 12-step meeting attendance, Wilks’ λ (4, 82) = 3.76, p < .01, ηp² = .16. Follow-up ANOVA tests revealed a significant relationship for 12-step meeting attendance with self-efficacy for abstinence, F(1, 85) = 7.00, p < .01, ηp² = .08, and meaning in life, F(1, 85) = 8.26, p < .01, ηp² = .09. Parameter estimates of the model revealed 12-step meeting attendance was a significant negative predictor of self-efficacy for abstinence, B = −3.24, t(87) = −2.65, p < .01, CI = −5.68, −0.81, and meaning in life, B = −0.56, t(87) = −2.87, p < .01, CI = −0.95, −0.17. However, 12-step meeting attendance was not significantly related to continuous days abstinent or days living in an Oxford House.

DISCUSSION

Twelve-step involvement, independent of 12-step meeting attendance, was significantly related to increases in recovery resources. Significantly more continuous days abstinent from both alcohol and drugs were reported by participants who were actively involved in 12-step activities than those who were less involved, a finding that is consistent with previous investigations (Montgomery et al., 1995; Timko & DeBenedetti, 2007; Witbrodt & Kaskutas, 2005). Being actively involved in 12-step activities in the present study meant that participants reported they utilized a social support network of recovering others outside the context of AA/NA meetings, had a sponsor who was knowledgeable about the 12 steps, were involved in service, and regularly attended a home group. These activities are considered to be some of the critical aspects of 12-step recovery (Laudet & White, 2008).

Twelve-step involvement was measured as a categorical variable consisting of participation in several specific 12-step activities in the present study, two of which (having a sponsor, being of service) have been identified as key components of 12-step involvement (Witbrodt & Kaskutas, 2005). Twelve-step involvement measured categorically, as opposed to an averaged summary score (dimensional) approach, can help us understand more explicitly the multifaceted aspects of 12-step involvement related to outcomes. Timko and DeBenedetti (2007) recently examined twelve-step activities with the AA Affiliation Scale (AAAS; Humphreys et al., 1998) and found overall involvement, measured as a combination of twelve-step activities, was related to better abstinence outcomes. However, twelve-step involvement measured categorically, as opposed to an averaged summary score (dimensional) approach, can help us understand more explicitly the multifaceted aspects of twelve-step involvement related to outcomes. In addition, we have learned anecdotally from recovering persons that involvement in 12-step activities such as those specified in the present study are regarded as “suggestions” or “the basics” of recovery that should not be taken selectively. Future investigations are needed to determine what combination of 12-step activities is necessary for ongoing recovery, as results would inform treatment providers in helping clients fully engage in such 12-step activities.

Twelve-step involvement in the present study also was related to increases in two intrapersonal recovery resources: self-efficacy for abstinence and meaning in life. Actively involved 12-step members reported significantly higher levels of self-efficacy for abstinence than those who were less involved. This finding is consistent with previous investigations that examined self-efficacy for abstinence where attendance was an integral part of 12-step participation (Moos & Moos, 2004, 2007) and involvement (Morgenstern et al., 1997). In addition, the present study extends previous research in that active involvement in 12-step activities, separate from meeting attendance, was related to significant increases in self-efficacy for abstinence, a recovery resource that was positively and significantly related to total continuous days abstinent in the parent study (Majer et al., 2002). Together, these findings have implications for further examining the role of 12-step involvement as a catalyst for resources such as self-efficacy for abstinence that might combat stress and sustain recovery across various abstinence time points (Laudet & White, 2008).

In addition, the relationship between active involvement in 12-step activities and increased meaning in life is consistent with previous investigations on 12-step participation (Majer, 1992; Montgomery et al., 1995). Findings in the present study extend previous research in that active involvement consisted of a number of 12-step activities beyond working the 12 steps of AA (Montgomery et al.) among a community as opposed to a clinical sample (Majer). Meaning in life is an important resource for recovery because the inability to find meaning might contribute to one's active addiction (Frankl, 1986). Moreover, research evidence suggests that meaning in life is an important recovery resource that protects against relapse (Robinson et al., 2007) while promoting sustained quality recovery (Laudet & White, 2008).

Furthermore, active 12-step involvement was related to increased length of stay in an Oxford House. Oxford House residents typically attend 12-step meetings (Nealon-Woods, Ferrari, & Jason, 1995) for their ongoing recovery. In addition, Oxford Houses operate loosely on traditions adapted from AA (Oxford House, Inc., 2002), so it is likely that residents who choose to remain in their houses for extended periods are more involved in their 12-step fellowships and related activities. This claim can only be verified by a longitudinal, not cross-sectional, design. Nonetheless, findings in the present study suggest that recovering addicts/alcoholics who are invested in the 12-step model of recovery would do well living in Oxford Houses.

Although active involvement in 12-step activities was related to increases in recovery resources, meeting attendance was found to be a significant covariate. The significant negative relationship between 12-step meeting attendance and self-efficacy for abstinence and meaning in life is not surprising for a couple reasons. First, newcomers to AA and NA lack recovery resources and typically attend more AA/NA meetings early in their abstinence when they begin to attribute meaning to their experiences (Galanter, 2007). In addition, research evidence suggests that recovering addicts/alcoholics later develop recovery resources such as meaning (Robinson et al., 2007) and self-efficacy for abstinence (Majer et al., 2004) within 6 months of their ongoing abstinence. Second, although meeting attendance may decrease with time, duration of involvement in 12-step activities which has been significantly related to increases in recovery resources (Moos & Moos, 2004, 2007) may not. The non-significant relationship between meeting attendance and continuous days abstinent implies that frequent meeting attendance does not necessarily facilitate good outcomes, consistent with previous research (Moos & Moos, 2004). Future investigations are needed to clarify the benefits of 12-step fellowships by evaluating measures of both 12-step involvement (i.e., categorical vs. dimensional approaches) and meeting attendance (e.g., duration vs. frequency) with time to better understand the dynamics of 12-step involvement in relation to recovery resources.

There are a few limitations of the study. One is that it relied on the use of self-reported data at a single time point when repeated administrations might have provided more meaningful information about 12-step involvement and recovery resources. In addition, participants were living in self-run, democratically operated recovery homes (Oxford Houses), so the effect of involvement in specific 12-step activities may not easily generalize among those involved in professional treatment or those who simply attend AA and NA. Furthermore, the present study viewed abstinence and time in an Oxford House as outcome variables, and future studies are needed to determine whether such variables moderate the relationship between 12-step involvement/attendance and other recovery resources. Finally, there might have been some selection bias in the recruitment effort. However, if any selection bias did occur in the present study, our findings suggest that the effects were relatively equal between groups.

The present study investigated involvement in 12-step activities, independent of 12-step meeting attendance, in relation to resources for ongoing recovery. Twelve-step involvement was measured as a categorical variable to prevent averaging effects from a dimensional assessment that might have artificially inflated or reduced participation in specific 12-step activities. Findings showed significant relationships between active 12-step involvement and increases in resources that sustain ongoing recovery, with treatment implications that addiction clinicians should target and encourage clients’ simultaneous involvement in a number of 12-step activities. The use of follow-up assessment intervals could provide data on changes throughout time and should be considered in future investigations. Comparison groups should be included in future investigations to better understand the effects of active 12-step involvement, while combinations of specific 12-step activities should be evaluated in addition to meeting attendance in terms of frequency and duration. With comparison groups consisting of patients from professionally run treatment modalities for alcohol/drug dependence and members of the 12-step community, we can better understand the empowering effects of both involvement in 12-step activities and meeting attendance. Overall, findings from the present study suggest that categorical involvement in 12-step activities is likely to empower recovering addicts/alcoholics with resources to sustain their ongoing recovery.