Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Apr 4.
Published in final edited form as: Drug Alcohol Depend. 2014 Jul 30;143:181–188. doi: 10.1016/j.drugalcdep.2014.07.023

Do young people benefit from AA as much, and in the same ways, as adult aged 30+? A moderated multiple mediation analysis

Bettina B Hoeppner 1, Susanne S Hoeppner 1, John F Kelly 1
PMCID: PMC10071823  NIHMSID: NIHMS1882752  PMID: 25150401

Abstract

Aims:

Research has shown that participation in Alcoholics Anonymous (AA) confers significant recovery benefit to adults suffering from alcohol use disorder (AUD). Concerns persist, however, that AA may not work as well for younger adults, who tend to have shorter addiction histories, different social circumstances, and less spiritual/religious interest than adults.

Design:

Secondary data analysis of Project MATCH, using a prospective, moderated multiple mediation analysis to test and compare six previously identified mechanisms of change in younger adults (n=266) vs. adults aged 30+ (n=1,460).

Setting:

Nine clinical sites within the United States.

Participants:

Treatment-seeking adults (n=1,726) suffering from AUD who participated in 12 weeks of outpatient treatment and completed follow-ups at 3-, 9- and 15-months.

Measurements:

AA attendance during treatment; mediators at 9 months; and outcomes [percentage of days abstinent (PDA) and drinks per drinking day (DDD)] at 15 months.

Findings:

AA attendance was associated with improved drinking outcomes in both younger adults (PDA: F(1,247)=8.55, p<0.01; DDD: F(1,247)=15.93, p<0.01) and adults aged 30+ (PDA: F(1,1311)=86.58, p<0.01; DDD: F(1,1311)=11.96, p<0.01). Only two of the six hypothesized pathways (i.e., decreases in pro-drinking social networks, self-efficacy in social situations) appeared to work in younger adults.

Conclusion:

Unidentified mechanisms of behavior change that are mobilized by AA participation appear to be at work in young people. Once identified, these mechanisms may shed new light on how exactly AA confers similar benefits for young people and, more broadly, may enhance our understanding of recovery-related change for young adults that could yield novel intervention targets.

Keywords: young adults, alcohol, alcohol use disorder (AUD), Alcoholics Anonymous, mediation, mechanism of change

1. Introduction

Individuals with severe alcohol and other drug addiction problems typically require ongoing monitoring and management over the long-term to facilitate stable and enduring recovery (Dennis and Scott, 2012; Hser et al., 2011; Kelly and White, 2011; White, 2008). A large number of studies have shown that participation in the international mutual-help organization Alcoholics Anonymous (AA) confers significant recovery benefit both during and following treatment (Humphreys, 1999; Kaskutas et al., 2005; Kelly et al., 2006; Kissin et al., 2003; Moos and Moos, 2004; Tonigan et al., 2003; Witbrodt et al., 2012). Concerns have persisted, however, regarding the fit of AA for certain subgroups of individuals (Humphreys, 2004; Kelly, 2003), including women, those with psychiatric comorbidities, atheists and agnostics, and young people (Kelly and Myers, 2007; Sussman, 2010).

AA’s 2011 membership survey shows that individuals under the age of 30 are in a minority in AA, comprising only 13% of members (Alcoholics Anonymous, 2011). This relative scarcity of younger people may present barriers to engagement and derived benefit for younger adults in an organization where the majority of members may have had more severe addiction problems (e.g., greater addiction severity, withdrawal symptoms, medical sequelae; Brown, 1993; Stewart and Brown, 1995) and are at a different life-stage (Hser et al., 2007). Compared to adults in their 40s, 50s, and 60s, those under the age of 30 are likely to be more often exposed to alcohol and drug cues in social contexts since the prevalence of alcohol and drug use in the general population is highest during the developmental phase of young adulthood (SAMHSA, 2011). Younger people also typically face different psychosocial stressors such as more transient living situations, sexual and romantic challenges, and financial stressors (SAMHSA, 2010). Of specific relevance to AA is also the fact that younger people tend to be less interested in spiritual and religious ideology and related practices (Pew Forum, 2008), making the overt spiritual focus of AA less attractive. Young people also more frequently use an illicit primary drug (Brown, 1993; Stewart and Brown, 1995), which may mean younger people may be less likely to engage in, and benefit from, an organization such as AA where the primary focus is on alcohol and complete lifelong abstinence.

Despite these challenges, there is emerging evidence that young people can and do benefit from organizations such as AA (Alford et al., 1991; Chi et al., 2012; Chi et al., 2009; Hsieh et al., 1998; Kelly et al., 2010a; Kelly and Urbanoski, 2012; Kennedy and Minami, 1993). Most of this evidence, however, is based on adolescents, and little is known about young adults and whether they benefit as much as their older counterparts. Similarly, little is known about the mechanisms with which AA attendance may confer benefit. Current evidence on young adults has shown that changes in impulsivity in young adults, but not adults, mediates the effect of AA attendance on drinking outcomes (Blonigen et al., 2011), and that adaptive social network changes, an important mechanism in adults, is related to substance use outcomes in young adults, but is not impacted by AA (Kelly et al., 2014). These findings suggest that AA attendance may confer benefit for young adults in different ways than for adults.

The goal of this study is to examine whether younger adults benefit from AA attendance as much, and in the same ways, as adults aged 30+, using the uniquely large clinical Project MATCH data (Project MATCH Research Group, 1993). Based on prior research on adults (e.g., Kelly et al., 2012; Kelly and Hoeppner, 2013; Kelly et al., 2010b), we examine six different mediators which have been shown to explain a large proportion of the alcohol recovery benefits that individuals derive from AA participation. These mechanisms include the ability to cope in high risk social contexts and when encountering negative affect, such as depression, anger, boredom, anxiety; spiritual/religious practices; depression symptoms; and two aspects of an individual’s social network: the number of pro-abstainers, and the number of pro-drinkers.

2. Methods

2.1. Participants

Participants were adults (n=1,726) suffering from alcohol use disorder (AUD) who participated in 12 weeks of outpatient treatment. Project MATCH inclusion criteria were: current DSM-III-R AUD diagnosis; alcohol as principal drug of misuse; drinking during 3 months prior to study; 18 or older; minimum sixth grade reading level. Exclusion criteria were: current DSM-III-R diagnosis of dependence on sedative-hypnotics, stimulants, cocaine or opiates; intravenous drug use in prior 6 months; danger to self/others; probation ⁄parole requirements that might interfere with participation; risk of residential instability; inability to identify at least one “locator” person to assist tracking; psychosis/organic impairment; involvement in alternative treatment other than MATCH (i.e., > 6 hours, except for self-help groups). We compared younger adults (n=266), and adults aged 30+ (n=1,460). While a dichotomization at age 25 would have been preferable given current definitions of emerging adulthood (Arnett, 2000), only n=80 participants were between 18–24, an insufficient sample size to conduct the necessary analyses (Kenny, 2014; Kline, 2005). Age 30, however, is consistent with older definitions of young adulthood (Levinson, 1978), AA’s own age group designation in membership surveys (Alcoholics Anonymous, 2011), and is of relevance regarding health insurance coverage.

2.2. Procedure

Subjects were randomly assigned to 1 of 3 individually-delivered, psychosocial interventions: cognitive behavioral therapy (CBT: Kadden et al., 1992), motivational enhancement therapy (MET: Milller et al., 1992), and 12-step facilitation therapy (TSF: Nowinski et al., 1992). Participants were reassessed at 3, 6, 9, 12, and 15 months following study intake, with follow-up rates over 90%. More complete details can be found elsewhere, including the psychometric properties of the measures used (Project MATCH Research Group, 1997). This study focused on baseline, 3-, 9-, and 15-month follow-ups, because only these time points contained the necessary variables needed for our lagged model.

2.3. Measures

2.3.1. Alcohol Use.

Alcohol consumption was assessed using the Form 90 (Miller and Del Boca, 1994), which combines an interview procedure with calendar-based and drinking pattern estimation methods. Two drinking outcomes were based on past 90 days: percent days abstinent (PDA) and number of drinks per drinking day (DDD).

2.3.2. Alcoholics Anonymous Attendance.

AA attendance was also assessed using Form 90, which captured the number of AA meetings attended during the past 90 days at intake and 3, 9, and 15 months. For analyses, we calculated the proportion of days attending AA by dividing the number of days attended by the total number of days in the period. Descriptively, because this variable was not normally distributed, we provide two indices characterizing AA attendance in Table 1 (i.e., % ever attended, if attended, % of days).

Table 1.

Project MATCH sample description at intake

Younger Adults (<30 years) (n=266) Adults aged 30+ (≥30 years) (n=1,460) t / χ2
mean / % SD mean / % SD
Demographics
 Age 25.8 2.7 42.9 9.8 56.2 **
 Gender (% male) 74.1 76.0 0.5
 Race (%) 66.4 **
  White 73.7 83.4
  Hispanic 5.6 10.6
  Black 20.7 6.0
 Marital Status (% married / cohabiting) 19.6 37.7 32.5 **
 Employment Status (% full-time) 45.5 50.3 2.1
Intake Clinical Descriptors
 Percent of days abstinent (PDA) † 37.2 [12.2, 65.6] 18.9 [2.2, 52.2] −6.0 **
 Drinks per drinking day (DDD) † ‡ 12.7 [9.0, 18.0] 14.0 [9.3, 21.4] 3.5 **
 # of prior Alcohol Treatments (max=4) ‡ 0.8 1.1 1.1 1.4 4.2 **
AA attendance
 (BL) Ever attended AA in past 3-months 26.0 26.6 0.1
 (BL) If attended, % of days † ‡ 3.3 [2.1, 6.6] 6.6 [2.3, 18.8] 4.1 **
 (3-mo) Ever attended AA in past 3-months 52.3 64.1 12.9 **
 (3-mo) If attended, % of days † 14.8 [5.5, 38.0] 27.0 [8.8, 51.2] 3.0 **
Intake levels of mediators
 Self-Efficacy (NA) 3.0 1.0 2.8 1.1 −1.9
 Self-Efficacy (Soc) 2.8 1.0 2.9 1.1 1.4
 Religiousness ‡ 22.8 8.1 24.2 9.2 2.3 *
 Depression † 8.0 [4.0, 15.0] 8.0 [4.0, 14.0] −0.1
 SocNet: pro-abst. ‡ 2.9 1.0 2.9 1.2 0.3
 SocNet: pro-drk. 1.3 1.2 1.1 1.2 −2.7 **
Open in a new tab

Note:

*

p < 0.05,

**

p < 0.01;

†

non-normally distributed, thus t-tests were run on transformed values, and medians and corresponding interquartile ranges are presented instead of means;

‡

Satterthwaite t-test for unequal variances was used

2.3.3. Self-efficacy.

The Alcohol Abstinence Self-Efficacy Scale (DiClemente et al., 1994) is a 20-item scale that assesses self-efficacy using four subscales, two of which were included in this study (Negative Affect: α=0.88; Social / Positive: α=0.82), as they have been shown to be mediators of the effect of AA attendance on alcohol outcomes (Owen et al., 2003).

2.3.4. Spiritual/Religious practices.

Spirituality/religiousness was assessed with the religious background and behavior instrument (RBB; Connors, 1996). As in previous research (Kelly et al., 2011b), we focused on RBB questions pertaining to past 90-day religious and spiritual practices (i.e.,: ”thought about God”, “prayed”, “meditated”, “attended worship services”, “read or studied scriptures ⁄ holy writings”, and “had direct experiences of God”), rated on 8-point Likert-scale (“never” to “more than once a day”).

2.3.5. Depression.

The Beck Depression Inventory (BDI: Beck et al., 1961) was used, a well-established, psychometrically sound (Beck et al., 1988) 21-item measure that assesses past-week depression symptom severity, with higher values indicating greater severity.

2.3.6. Social Networks.

The Important People and Activities Instrument (IPA: Clifford and Longabaugh, 1991) was used, where we used “pro-drinking” and “pro-abstinence” to characterize patients’ social networks, based on previous research (Kelly et al., 2011a). In the IPA, patients are asked to name the four most important people of the past 6 months, and to rate how each reacted to their abstinence or drinking. A person was coded as “pro-abstinence” if s/he either encouraged abstinence or discouraged drinking, or both. A person was coded as “pro-drinking” if s/he either encouraged drinking or discouraged abstinence, or both. The number of each type of network member was summed.

2.3.7. Baseline Characteristics.

Demographic information and the number of prior alcohol treatments were recorded. Gender, marital status, and employment status were coded as binary variables; race was coded as a 3-level categorical variable. The number of prior alcohol treatments was capped at four.

2.4. Analytic Strategy

2.4.1. Data Preparation.

The dependent variables (i.e., PDA & DDD) and the independent AA attendance variable were transformed prior to analyses (PDA⁄ arcsine transformed; DDD⁄ square root transformed, and AA attendance ⁄ log transformed).

2.4.2. Effect of AA on Drinking Outcomes.

In order to test if the previously reported effect of AA on drinking outcomes (i.e., PDA, DDD) (Kelly and Hoeppner, 2013) was impacted by age, we fit multivariable regression models, where PDA and DDD (assessed at 15-month) were the dependent variables, respectively, and AA attendance (assessed at 3-month), age group and the interaction between age group and AA attendance were the predictors. We interpreted a significant interaction effect as differences in the relationship between AA and drinking outcomes.

2.4.3. Mediational Analyses.

To avoid temporal confounding (Kazdin and Nock, 2003), we employed a fully lagged mediational design. We examined AA attendance during Project MATCH treatment (months 0–3), mediators at 9-month follow-up, and alcohol outcomes (PDA; DDD) at 15-month follow-up (Figure 1). We controlled for demographic variables, MATCH treatment assignment and treatment program site, prior alcohol treatment, and the baseline levels of the outcome and mediator variables.

Figure 1.

Figure 1.

Open in a new tab

Path diagram of the mediational model fit for both age groups. Baseline covariates are included to help rule out other causes of change in the mediators and outcomes (i.e., other than AA).

We used a structural equation modeling (SEM) multi-group approach using SAS 9.3 PROC CALIS to test for age group differences in the mediational paths in the previously tested multiple mediator model (Kelly et al., 2012). To test for overall model differences between the two age groups, we first fit a model in which we constrained all parameter estimates to be equal for both age groups, and then compared it to a partially constrained model in which parameter estimates of the mediational paths could differ (i.e., paths demarked by thicker lines in Figure 1). Based on this second model, we conducted planned comparisons of the mediational paths, and tested individual mediation pathways separately for younger adults and adults aged 30+ using the product-of-coefficients approach (Sobel, 1982, 1986), and constructed 95% confidence intervals using the Monte Carlo Method for Assessing Mediation (MCMAM: MacKinnon et al., 2004), as implemented by the interactive tool created by Selig and Preacher (2008). We fit these models for both alcohol outcomes (i.e., PDA and DDD).

2.4.4. Missing Data.

Missing data were observed for 3.6% at 3-month, 9.5–13.6% at 9-month, and 9.1% at 15-month. To address missing data, we used the maximum likelihood estimation approach (Schafer and Graham, 2002), where we first estimated the variance-covariance matrix using all available data points (using the iterative expectation-maximization (EM) algorithm [SAS 9.3 Proc MI]), and then used this matrix as the input data for fitting the structural equation models.

3. Results

3.1. Age Group Differences in Sample Characteristics

The two age groups differed on numerous demographic and clinical dimensions at baseline (Table 1), with less clinical severity in the younger age group (i.e., higher PDA, fewer DDD and prior alcohol treatments), lower religiousness, and a stronger pro-drinking social network. At baseline, the proportions of ever attenders of AA meetings was similar between the age groups (26.0% vs. 26.6%), but among those who did attend, younger adults attended AA meetings less frequently (3% vs. 7% of days). By the 3-month follow-up, AA attendance had increased overall, though fewer younger adults than adults aged 30+ had attended any AA meetings (52% vs. 64%), and younger adults who attended did so less frequently (15% vs. 27% of days).

3.2. Effect of AA attendance on Drinking Outcomes

In both multivariable regression models, the effects of AA attendance on drinking outcome were significant (PDA: F(1,1558)= 35.63, p<0.0001; DDD: F(1,1558)=20.93, p<0.0001), while the interaction effects between age group and AA attendance were non-significant (PDA: F(1,1558)=2.71, p=0.10; DDD: F(1,1558)=3.13, p=0.08), indicating that the effect of AA attendance on drinking outcomes did not differ between the two age groups. At the 15-month follow-up, fewer younger adults than adults age 30+ were completely abstinent from alcohol (30.4% vs. 39.0%, χ2(1)=6.7, p<0.01), but among those who were not completely abstinent, younger adults were drinking less often (median PDA=82.1% [interquartile range: 55.6–94.4%] vs. 69.7% [27.8–92.1%]; t(319)=−5.4, p<0.01). There were no age group differences regarding DDD (median=6.9 [4.6–11.5] vs. 7.4 [4.4–12.9], t(313)=1.9, p=0.07). Post-hoc simple linear regression analyses indicated that in both younger adults (PDA: F(1,247)=8.55, p<0.01; DDD: F(1,247)=15.93, p<0.01) and adults aged 30+ (PDA: F(1,1311)=86.58, p<0.01; DDD: F(1,1311)=11.96, p<0.01), the effect of AA participation on drinking outcomes was significant. Effect sizes were larger for PDA in adults aged 30+ (R2=0.03 and 0.06 for younger adults and adults aged 30+, respectively) and for DDD in younger adults (R2=0.06 and 0.01).

3.3. Age Differences in Mediational Relationships

Fitting mediation models with different parameters for the two age groups (i.e., the partially constrained model) significantly improved model fit for both PDA and DDD (PDA: χ2(161)=214.1, p < 0.01; DDD: χ2(161)=230.2, p < 0.001). Examination of the specific mediational paths (Table 2) indicated that most paths were not significantly different between age groups, but that isolated significant differences existed. First, the remaining direct effect of AA on DDD, but not PDA, was significantly stronger (χ2(1)=7.4, p<0.05) in younger adults (β=−0.22) than adults aged 30+ (β=−0.03) (see Table 3 for standardized path estimates), indicating that the mediational paths explained the effect of AA on DDD less well in younger adults. Second, for the effect of AA attendance on mediators, only the effect on the pro-abstinence social network differed between age groups (χ2(1)=6.1, p<0.05), with younger adults showing a weaker effect of AA attendance on these networks than adults aged 30+ years (β=0.05 vs. β=0.21). Finally, for the effects of the mediators on drinking outcomes, differences were found for self-efficacy in social situations (PDA: (χ2(1)=4.1, p < 0.05 and χ2(1)=0.9) and religiousness (PDA: χ2(1)=7.4, p< 0.05; DDD: χ2(1)=4.4, p<0.05). Here, the effect of self-efficacy in social situations on PDA was less strong in younger adults (β=0.18) than adults aged 30+ (β=0.25). Similarly, religiousness had no significant effect on drinking outcomes in younger adults while it had a positive effect for adults aged 30+, for both PDA (β=−0.04 vs. β=0.11) and DDD (β=0.03 vs. β=−0.10).

Table 2.

Age differences in the mediational paths

Percent of Days Abstinent (PDA)
 Drinks per Drinking Day (DDD)
b(A) - b(YA) df χ 2 b(A) - b(YA) df χ 2
Direct effect: AA attendance predicting alcohol outcomes
 AA attendance → PDA/DDD −0.04 1 0.0 0.19 1 7.4 **
Mediational path: AA attendance predicting mediators
 AA attendance → Self-Efficacy (NA) 0.05 1 0.4 0.05 1 0.4
 AA attendance → Self-Efficacy (Soc) −0.04 1 0.5 −0.04 1 0.5
 AA attendance → Religiousness 0.02 1 0.3 0.02 1 0.3
 AA attendance → Depression −0.10 1 2.3 −0.10 1 2.3
 AA attendance → SocNet: pro-abst. 0.16 1 6.1 * 0.16 1 6.1 *
 AA attendance → SocNet: pro-drk. 0.06 1 1.7 0.06 1 1.7
Mediational path: Mediators predicting alcohol outcomes
 Self-Efficacy (NA) → PDA/DDD −0.02 1 0.0 −0.08 1 2.2
 Self-Efficacy (Soc) → PDA/DDD 0.07 1 4.1 * −0.04 1 0.9
 Religiousness → PDA/DDD 0.15 1 7.4 * −0.13 1 4.4 *
 Depression → PDA/DDD 0.06 1 0.5 0.00 1 0.2
 SocNet: pro-abst. → PDA/DDD 0.02 1 0.6 0.08 1 1.3
 SocNet: pro-drk. → PDA/DDD 0.04 1 0.6 −0.03 1 0.0
Open in a new tab

Note:

*

p < 0.05,

**

p < 0.01

b(A) - b(YA) values are based on differences of standardized parameter estimates, while the chi-squre test of the difference is based on the unstandardized parameters. The direction of the difference can be interpreted by comparing the standardized path estimates in Table 5.

Table 3.

Standardized path parameter estimates of the multiple mediator model

Type of Path Percent of Days Abstinent
 Drinks per Drinking Day
Younger Adults (adj. n=246) Adults aged 30+ (adj. n=1,298) Younger Adults (adj. n=246) Adults aged 30+ (adj. n=1,298)
 Path β SE t β SE t β SE t β SE t
Direct effect: AA attendance predicting alcohol outcomes
 AA attendance → PDA/DDD 0.16 0.06 2.51 * 0.12 0.03 4.06 * −0.22 0.06 −3.46 * −0.03 0.03 −0.98
Mediational path: AA attendance predicting mediators
 AA attendance → Self-Efficacy (NA) 0.02 0.06 0.25 0.06 0.03 1.94 0.02 0.06 0.25 0.06 0.03 1.94
 AA attendance → Self-Efficacy (Soc) 0.19 0.06 3.01 * 0.14 0.03 4.63 * 0.19 0.06 3.01 * 0.14 0.03 4.63 *
 AA attendance → Religiousness 0.16 0.05 3.47 * 0.18 0.02 8.77 * 0.16 0.05 3.47 * 0.18 0.02 8.77 *
 AA attendance → Depression 0.02 0.06 0.26 −0.08 0.03 −2.86 * 0.02 0.06 0.26 −0.08 0.03 −2.86 *
 AA attendance → SocNet: pro-abst. 0.05 0.06 0.84 0.21 0.03 7.09 * 0.05 0.06 0.84 0.21 0.03 7.09 *
 AA attendance → SocNet: pro-drk. −0.21 0.06 −3.38 * −0.14 0.03 −4.46 * −0.21 0.06 −3.38 * −0.14 0.03 −4.46 *
Mediational path: Mediators predicting alcohol outcomes
 Self-Efficacy (NA) → PDA/DDD 0.09 0.06 1.56 0.07 0.02 2.85 * −0.06 0.06 −0.96 −0.13 0.03 −4.95 *
 Self-Efficacy (Soc) → PDA/DDD 0.18 0.06 3.19 * 0.25 0.02 10.30 * −0.16 0.06 −2.72 * −0.20 0.03 −7.65 *
 Religiousness → PDA/DDD −0.04 0.06 −0.69 0.11 0.04 3.09 * 0.03 0.06 0.45 −0.10 0.04 −2.66 *
 Depression → PDA/DDD −0.10 0.06 −1.72 −0.04 0.03 −1.57 0.14 0.06 2.36 * 0.14 0.03 5.02 *
 SocNet: pro-abst. → PDA/DDD 0.13 0.06 2.45 * 0.15 0.03 6.03 * −0.14 0.06 −2.48 * −0.07 0.03 −2.35 *
 SocNet: pro-drk. → PDA/DDD −0.27 0.05 −4.84 * −0.22 0.02 −9.31 * 0.16 0.06 2.71 * 0.13 0.03 4.87 *
Open in a new tab

Note: “adj. n” refers to the average number of participants who had data for the 8 structural variables; not shown but included in model are baseline covariates (see Figure 1).

*

p < 0.05

3.4. Multiple Mediation Effects per Age Group

Mediation through the multiple mediators was statistically significant for both outcomes in both younger adults (PDA: t=3.0, p < 0.01; DDD: t=−2.4, p < 0.05) and adults aged 30+ (PDA: t=8.4, p < 0.01; DDD: t=−6.8, p < 0.01). Indirect effects (i.e., mediational pathways) explained more of the drinking outcomes variance in adults aged 30+ than in younger adults (Table 4) for both PDA and DDD.

Table 4.

Partioning of the standardized effect of AA on drinking outcome: Total, direct and indirect effects

Drinking Outcome Total Direct Indirect
 Sample R2 EST EST % EST %
Percent of Days Abstinent
 Younger adults 0.27 0.25 ** 0.16 * 64.3% 0.09 ** 35.7%
 Adults aged 30+ 0.29 0.25 ** 0.12 ** 48.9% 0.13 ** 51.1%
Drinks per Drinking Day
 Younger adults 0.20 −0.29 ** −0.22 ** 77.7% −0.06 * 22.3%
 Adults aged 30+ 0.17 −0.13 ** −0.03 24.4% −0.10 ** 75.6%
Open in a new tab

Note:

*

p < 0.05

**

p < 0.01

3.5. Specific Mediational Effects

In younger adults, only two of the six mediators exhibited complete mediational pathways (Table 5), for both PDA and DDD, whereas four (for PDA) to five (for DDD) of the six complete mediational pathways were significant in adults aged 30+. These two mediators were self-efficacy in social situations and pro-drinking social networks. While significant in both age groups, they had a greater impact and relative importance (based on the percent of the total indirect effect explained) in younger adults than adults aged 30+, more so for the number of pro-drinking social network members (PDA: 52.4% vs. 24.8% of the mediational effect; DDD: 42.3% vs. 18.2%) than self-efficacy in social situations (PDA: 31.5% vs. 28.0%; DDD: 38.0% vs. 29.1%). Note, however, that the differences in individual path parameters (Table 2) were not necessarily significantly different.

Table 5.

Test of mediation for specific indirect (i.e., mediated) effects using the Product of Coefficients approach

Alcohol Outcome Mediator Younger Adults (adj. n=246)
 Adults aged 30+ (adj. n=1,298)
EST SE 95% CI %† EST SE 95% CI %†
Percent of Days Abstinent (PDA)
 Self-Efficacy (NA) 0.000 0.001 [−0.0012, 0.0017] 1.3% 0.001 0.000 [0.0000, 0.0013] 3.3%
 Self-Efficacy (Soc) 0.003 0.002 [0.0008, 0.0071] * 31.5% 0.004 0.001 [0.0025, 0.0067] * 28.0%
 Religiousness −0.001 0.001 [−0.0031, 0.0013] 6.6% 0.003 0.001 [0.0009, 0.0043] * 15.8%
 Depression 0.000 0.001 [−0.0017, 0.0012] 1.4% 0.000 0.000 [−0.0001, 0.0012] 2.7%
 SocNet: pro-abst. 0.001 0.001 [−0.0010, 0.0030] 6.8% 0.004 0.001 [0.0025, 0.0059] * 25.4%
 SocNet: pro-drk. 0.006 0.002 [0.0020, 0.0104] * 52.4% 0.004 0.001 [0.0021, 0.0060] * 24.8%
Drinks per Drinking Day (DDD)
 Self-Efficacy (NA) 0.000 0.001 [−0.0053, 0.0039] 1.1% −0.003 0.002 [−0.0076, 0.0000] 8.0%
 Self-Efficacy (Soc) −0.012 0.006 [−0.0258, −0.0020] * 38.0% −0.013 0.003 [−0.0193, −0.0068] * 29.1%
 Religiousness 0.002 0.004 [−0.0063, 0.0110] 6.0% −0.008 0.003 [−0.0149, −0.0022] * 19.0%
 Depression 0.001 0.003 [−0.0061, 0.0087] 2.7% −0.005 0.002 [−0.0097, −0.0015] * 11.8%
 SocNet: pro-abst. −0.003 0.004 [−0.0125, 0.0042] 9.8% −0.006 0.003 [−0.0116, −0.0009] * 13.8%
 SocNet: pro-drk. −0.013 0.006 [−0.0277, −0.0025] * 42.3% −0.008 0.002 [−0.0132, −0.0036] * 18.2%
Open in a new tab

Note: “adj. n” refers to the average number of participants who had data for the 8 structural variables;

†

percentage of total indirect effects attributable to specific mediators

*

p < 0.05 based on 95% confidence intervals using the Monte Carlo Method for Assessing Mediation (MCMAM) (MacKinnon, Lockwood, & Williams, 2004)

Mediational pathways significant in adults aged 30+ but not in younger adults were religiousness, the number of pro-abstinence social network members and depression (DDD only). Notably, AA attendance also significantly increased religiousness in younger adults, yet religiousness had no significant impact on drinking outcomes in younger adults. Conversely, the number of pro-abstinence social network members was significantly related to drinking outcomes in both age groups, but AA attendance only had a beneficial effect on the pro-abstinence social network in adults aged 30+, and not in younger adults. A similar effect was found for depression in regards to DDD, where reductions in depression was linked to salutary effects in both age groups, but was only positively impacted by AA in adults aged 30+.

4. Discussion

In this secondary data analysis of the Project MATCH data, we found that younger adults do indeed benefit from AA; that is, their AA attendance was associated with subsequent better drinking outcomes, just as it was in adults aged 30+.

The mechanisms by which these salutary effects were achieved appeared to be similar in the two age groups in many respects, but overall were less accounted for in younger adults: fewer mediational pathways were significant in younger adults, and a larger direct effect remained after accounting for mediational pathways. Thus, while the previously identified pathways are quite good at explaining the effect of AA attendance on drinking outcomes in adults aged 30+ (i.e., 51–76% of the total effect of AA participation, for PDA and DDD, respectively), these pathways explain a much more modest percent of the effect in younger adults (i.e., 22–36%, for DDD and PDA, respectively). Of note, these effects were observed despite very similar levels of AA attendance frequency. This finding suggests that there may be currently unidentified pathways through which AA attendance helps younger people achieve improved drinking outcomes that, if they could be identified, might provide novel treatment targets for younger adults seeking to recover from AUD.

There are several possibilities for such pathways. One possibility is that AA attendance improves recovery motivation in younger adults, as suggested by findings in adolescents (Kelly et al., 2000). Typically, younger age groups have lower motivation than older age groups to remain abstinent: they have fewer negative associations with drinking (Gadon et al., 2004), while experiencing quite a few positive consequences of drinking (Park, 2004), where higher positive drinking associations have been shown to predict drinking behavior (Houben and Wiers, 2008). By attending AA meetings, younger adults are vividly exposed to the toll drinking can take on one’s life through the life stories shared by their older peers, thereby strengthening the negative associations with alcohol, and maintaining if not increasing the motivation to remain abstinent. Other possibilities include increased feelings of hope, cohesiveness, and belonging through AA participation (Labbe et al., under review), or perhaps an improved sense of agency or taking control of one’s disorder by taking the explicit step of going to AA meetings. Further research investigating such understudied pathways is necessary to enhance our understanding of AUD recovery in this age group.

Identified pathways in younger adults, which were also present in adults aged 30+, involved increases in self-efficacy in high-risk social situations, and decreases in pro-drinking social networks. Interestingly, the converse, increases in pro-abstinence social networks, which was an important mechanism in adults aged 30+, was not a functional mechanism in younger adults, because AA attendance was much less strongly associated with increases in pro-abstinence social network members in younger adults. This lack of an effect is in line with another recent finding on young adults (Kelly et al., 2014), yet nevertheless unfortunate, as the effect of pro-abstinence social networks on drinking outcomes was similarly strong in both age groups. Thus, it appears that AA may be less well suited to help younger adults make new social connections with persons supporting abstinence, perhaps because there are fewer young adults at 12-step meetings with whom to form new social contacts (Kelly et al., 2014). Note, for example, that in Project MATCH younger adults were less likely to have an AA sponsor than adults older than 30 years of age (20% vs. 30% by the 15-month follow-up). This lower rate suggests that it may be beneficial to facilitate young people’s AA meetings, perhaps in close proximity to treatment programs. The benefits of such an approach, however, appear mixed. Findings regarding age composition of AA meetings in adolescents (Kelly et al., 2005) and young adults (Labbe et al., 2013) suggest that while AA meetings specific to young people do foster engagement, they may not promote long-term sobriety. Thus, young adults may need to find new social contacts outside of AA.

Among the mechanisms that helped adults aged 30+, but not younger adults, was spiritual/religious practices. Specifically, contrary to our expectations, our findings indicate that in this sample, AA attendance was associated with increases in these practices very similarly in younger adults and adults aged 30+. This finding is in line with research on young adults who are both long term AA members and successful abstainers who overwhelmingly (92%)self-report experiencing a “spiritual awakening” due to AA (Galanter et al., 2012). Such increases, however, were not associated with improvements in drinking behaviors in younger adults, neither in frequency (PDA) nor intensity (DDD). Together, these findings suggest that spiritual changes may be co-occurring but are unlikely to be causally linked to positive drinking outcomes in younger adults.

Depression was another incomplete path in younger adults. That is, reductions in depression were linked to lower DDD (but not in PDA) in both age groups, in line with epidemiological evidence (Graham et al., 2007), but AA attendance was not related to reductions in depression in younger adults. Curiously, this absence of lowered depression scores occurred despite concurrent increases in religiousness, which have been linked with lower depression scores in this age group (Galanter et al., 2012). Regardless, depression as well does not appear to be a functional mechanism by which AA confers benefit in younger adults.

Finally, the observed effect sizes for the effects of AA on drinking outcomes, though not statistically significant, looked quite different between the age groups. These effect sizes suggest that young people may benefit from AA less in terms of abstinence, and more in terms of reducing their intensity of drinking. Further research is necessary to clarify this point.

4.1. Strengths and Limitations

Strengths of this project include the use of a uniquely large, geographically diverse sample, which allowed us to examine younger adults, who are typically underrepresented in AA, and thus difficult to study. Still, our age cut-off at age 30 rather than 24 was a necessary limitation imposed by sample size consideration, and even so statistical power to detect age group differences was somewhat limited, as was power to detect younger adult specific effects. Nevertheless, our use of state-of-the-art mediational modeling enabled us to compare standardized effect estimates and effect sizes to better understand processes of change in younger adults vis-à-vis adults aged 30+. Another limitation is our selection of mediators, which we limited to previously identified mechanisms that, moreover, would be applicable to diverse treatment modalities, not just AA. Thus, the examined mediators by no means represent an exhaustive list of possible mediators, and indeed, the adding and removing of mediators may well impact the relative importance rating of these mechanisms. Finally, it should be noted that causal inferences are merely suggested and not tested in this design, as all relationships are correlational rather than experimentally controlled. Indeed, the temporal resolution (i.e., observations several months apart) are not ideal to examine mechanisms of change, and underlie the observed low effect sizes. Also of note is the fact that the Project MATCH data were collected three decades ago (1990–1997), and thus, the social environment of the study participants was undoubtedly quite different than it is today (e.g., pre smartphone, online social networking, etc.).

4.2. Conclusion

Our findings indicate that AA attendance is significantly, positively, and similarly strongly related to positive drinking outcomes in younger adults as it is in adults aged 30+. The mechanisms by which this positive effect is brought about, however, are less well explained by previously identified pathways. This finding suggests that other mechanisms are at work in younger adults that, once identified, could provide novel treatment targets for younger adults seeking to recover from AUD.

5. References

  1. Anonymous Alcoholics, 2011. 2011 Membership Survey. Alcoholics Anonymous World Services, Inc., New York, NY. [Google Scholar]
  2. Alford GS, Koehler RA, Leonard J, 1991. Alcoholics Anonymous-Narcotics Anonymous model inpatient treatment of chemically dependent adolescents: a 2-year outcome study. J Stud Alcohol 52, 118–126. [DOI] [PubMed] [Google Scholar]
  3. Arnett JJ, 2000. Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist 55, 469–480. [PubMed] [Google Scholar]
  4. Beck AT, Steer RA, Garbin MG, 1988. Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review 8, 77–100. [Google Scholar]
  5. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J, 1961. An inventory for measuring depression. Archives of General Psychiatry 4, 561–571. [DOI] [PubMed] [Google Scholar]
  6. Blonigen DM, Timko C, Finney JW, Moos BS, Moos RH, 2011. Alcoholics Anonymous attendance, decreases in impulsivity and drinking and psychosocial outcomes over 16 years: moderated-mediation from a developmental perspective. Addiction 106, 2167–2177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brown SA, 1993. Recovery patterns in adolescent substance abuse. In: Baer JS, Marlatt GA, McMahon RJ (Eds.), Addictive behaviors across the life span: Prevention, treatment, and policy issues. Sage Publications, Inc, Thousand Oaks, CA, US. pp. 161–183. [Google Scholar]
  8. Chi FW, Campbell CI, Sterling S, Weisner C, 2012. Twelve-Step attendance trajectories over 7 years among adolescents entering substance use treatment in an integrated health plan. Addiction 107, 933–942. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chi FW, Kaskutas LA, Sterling S, Campbell CI, Weisner C, 2009. Twelve-Step affiliation and 3-year substance use outcomes among adolescents: social support and religious service attendance as potential mediators. Addiction 104, 927–939. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Clifford PR, Longabaugh R, 1991. Manual for the Administration of the Important People and Activities Instrument: Adapted for use by Project MATCH for NIAAA 5R01AA06698–05. Center for Alcohol and Addiction Studies, Brown University, Providence. [Google Scholar]
  11. Connors GJ, Tonigan JS, and Miller WR, 1996. Measure of religious background and behavior for use in behavior change research. Psychology of Addictive Behaviors 10, 90–96. [Google Scholar]
  12. Dennis ML, Scott CK, 2012. Four-year outcomes from the Early Re-Intervention (ERI) experiment using Recovery Management Checkups (RMCs). Drug Alcohol Depend 121, 10–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. DiClemente CC, Carbonari JP, Montgomery RPG, Hughes SO, 1994. The Alcohol Abstinence Self-Efficacy scale. Journal of Studies on Alcohol 55, 141–148. [DOI] [PubMed] [Google Scholar]
  14. Gadon L, Bruce G, McConnochie F, Jones BT, 2004. Negative alcohol consumption outcome associations in young and mature adult social drinkers: a route to drinking restraint? Addict Behav 29, 1373–1387. [DOI] [PubMed] [Google Scholar]
  15. Galanter M, Dermatis H, Santucci C, 2012. Young People in Alcoholics Anonymous: the role of spiritual orientation and AA member affiliation. J Addict Dis 31, 173–182. [DOI] [PubMed] [Google Scholar]
  16. Graham K, Massak A, Demers A, Rehm J, 2007. Does the association between alcohol consumption and depression depend on how they are measured? Alcohol Clin Exp Res 31, 78–88. [DOI] [PubMed] [Google Scholar]
  17. Houben K, Wiers RW, 2008. Implicitly positive about alcohol? Implicit positive associations predict drinking behavior. Addict Behav 33, 979–986. [DOI] [PubMed] [Google Scholar]
  18. Hser YI, Evans E, Huang D, Messina N, 2011. Long-term outcomes among drug-dependent mothers treated in women-only versus mixed-gender programs. J Subst Abuse Treat 41, 115–123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hser YI, Longshore D, Anglin MD, 2007. The life course perspective on drug use: a conceptual framework for understanding drug use trajectories. Eval Rev 31, 515–547. [DOI] [PubMed] [Google Scholar]
  20. Hsieh S, Hoffmann NG, Hollister CD, 1998. The relationship between pre-, during-, post-treatment factors, and adolescent substance abuse behaviors. Addict Behav 23, 477–488. [DOI] [PubMed] [Google Scholar]
  21. Humphreys K, 1999. Professional interventions that facilitate 12-step self-help group involvement. Alcohol Res Health 23, 93–98. [PMC free article] [PubMed] [Google Scholar]
  22. Humphreys K, 2004. Circles of recovery: Self-help organizations for addictions. Cambridge University Press, Cambridge, UK. [Google Scholar]
  23. Kadden RM, Carroll KM, Donovan D, Cooney NL, Monti PM, Abrams D, Litt MD, Hester R, 1992. Cognitive-behavioral coping skills therapy manual: A clinical research guide for therapists treating individuals with alcohol abuse and dependence. Project MATCH Monograph Series, 4 (DHHS Publication No. (ADM) 92–1895). [Google Scholar]
  24. Kaskutas LA, Ammon L, Delucchi K, Room R, Bond J, Weisner C, 2005. Alcoholics anonymous careers: patterns of AA involvement five years after treatment entry. Alcohol Clin Exp Res 29, 1983–1990. [DOI] [PubMed] [Google Scholar]
  25. Kazdin AE, Nock MK, 2003. Delineating mechanisms of change in child and adolescent therapy: methodological issues and research recommendations. Journal of Child Psychology and Psychiatry 44, 1116–1129. [DOI] [PubMed] [Google Scholar]
  26. Kelly JF, 2003. Self-help for substance use disorders: history, effectiveness, knowledge gaps, and research opportunities. Clin Psychol Rev 23, 639–663. [DOI] [PubMed] [Google Scholar]
  27. Kelly JF, Dow SJ, Yeterian JD, Kahler CW, 2010a. Can 12-step group participation strengthen and extend the benefits of adolescent addiction treatment? A prospective analysis. Drug Alcohol Depend 110, 117–125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kelly JF, Hoeppner B, Stout RL, Pagano M, 2012. Determining the relative importance of the mechanisms of behavior change within Alcoholics Anonymous: a multiple mediator analysis. Addiction 107, 289–299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kelly JF, Hoeppner BB, 2013. Does Alcoholics Anonymous work differently for men and women? A moderated multiple-mediation analysis in a large clinical sample. Drug Alcohol Depend 130, 186–193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kelly JF, Myers MG, 2007. Adolescents’ participation in Alcoholics Anonymous and Narcotics Anonymous: Review, implications and future directions. Journal of Psychoactive Drugs 39, 259–269. [DOI] [PubMed] [Google Scholar]
  31. Kelly JF, Myers MG, Brown SA, 2000. A multivariate process model of adolescent 12-step attendance and substance use outcome following inpatient treatment. Psychol Addict Behav 14, 376–389. [PMC free article] [PubMed] [Google Scholar]
  32. Kelly JF, Myers MG, Brown SA, 2005. The Effects of Age Composition of 12-Step Groups on Adolescent 12-Step Participation and Substance Use Outcome. J Child Adolesc Subst Abuse 15, 63–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kelly JF, Stout RL, Greene MC, Slaymaker V, 2014. Young adults, social networks, and addiction recovery: post treatment changes in social ties and their role as a mediator of 12-step participation. PloS one 9, e100121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kelly JF, Stout RL, Magill M, Tonigan JS, 2011a. The role of Alcoholics Anonymous in mobilizing adaptive social network changes: a prospective lagged mediational analysis. Drug Alcohol Depend 114, 119–126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kelly JF, Stout RL, Magill M, Tonigan JS, Pagano ME, 2010b. Mechanisms of behavior change in alcoholics anonymous: does Alcoholics Anonymous lead to better alcohol use outcomes by reducing depression symptoms? Addiction 105, 626–636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kelly JF, Stout RL, Magill M, Tonigan JS, Pagano ME, 2011b. Spirituality in recovery: a lagged mediational analysis of alcoholics anonymous’ principal theoretical mechanism of behavior change. Alcohol Clin Exp Res 35, 454–463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kelly JF, Stout RL, Zywiak W, Schneider R, 2006. A 3-year study of addiction mutual-help group participation following intensive outpatient treatment. Alcoholism: Clinical and Experimental Research 30, 1381 – 1392. [DOI] [PubMed] [Google Scholar]
  38. Kelly JF, Urbanoski K, 2012. Youth recovery contexts: the incremental effects of 12-step attendance and involvement on adolescent outpatient outcomes. Alcohol Clin Exp Res 36, 1219–1229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Kelly JF, White WL (Eds.), 2011. Addiction Recovery Management. Springer, New York. [Google Scholar]
  40. Kennedy BP, Minami M, 1993. The Beech Hill Hospital/Outward Bound Adolescent Chemical Dependency Treatment Program. J Subst Abuse Treat 10, 395–406. [DOI] [PubMed] [Google Scholar]
  41. Kenny DA, 2014. Measuring Model Fit. [Google Scholar]
  42. Kissin W, McLeod C, McKay J, 2003. The longitudinal relationship between self-help group attendance and course of recovery. Evaluation and Program Planning 26, 311–323. [Google Scholar]
  43. Kline RB, 2005. Principles and practice of structural equation modeling. Guilford Press, New York, NY. [Google Scholar]
  44. Labbe AK, Greene C, Bergman BG, Hoeppner B, Kelly JF, 2013. The importance of age composition of 12-step meetings as a moderating factor in the relation between young adults’ 12-step participation and abstinence. Drug Alcohol Depend 133, 541–547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Labbe AK, Slaymaker V, Kelly JF, under review. Toward enhancing Twelve-Step Facilitation among young people: A systematic qualitative investigation of young adults’ 12-step experiences. Substance Abuse. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Levinson DJ, 1978. The Seasons of a Man’s Life. Ballantine Books, Toronto, Canada. [Google Scholar]
  47. MacKinnon DP, Lockwood CM, Williams J, 2004. Confidence Limits for the Indirect Effect: Distribution of the Product and Resampling Methods. Multivariate Behavioral Research 39, 99–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Miller WR, Del Boca FK, 1994. Measurement of drinking behavior using the Form 90 family of instruments. Journal of Studies on Alcohol SUPPL 12, 112–118. [DOI] [PubMed] [Google Scholar]
  49. Milller WR, Zweben A, DiClemente CC, Rychtarik RG, 1992. Motivational Enhancement Therapy Manual: A Clinical Research Guide for Therapists Treating Individuals with Alcohol Abuse and Dependence. Rockville, MD. [Google Scholar]
  50. Moos RH, Moos BS, 2004. Long-term influence of duration and frequency of participation in alcoholics anonymous on individuals with alcohol use disorders. J Consult Clin Psychol 72, 81–90. [DOI] [PubMed] [Google Scholar]
  51. Nowinski J, Baker S, Carroll KM, 1992. Twelve-Step Facilitation Therapy Manual: A Clinical Research Guide for Therapists Treating Individuals with Alcohol Abuse and Dependence. Rockville, MD. [Google Scholar]
  52. Owen PL, Slaymaker V, Tonigan JS, McCrady BS, Epstein EE, Kaskutas LA, Humphreys K, Miller WR, 2003. Participation in alcoholics anonymous: intended and unintended change mechanisms. Alcohol Clin Exp Res 27, 524–532. [DOI] [PubMed] [Google Scholar]
  53. Park CL, 2004. Positive and negative consequences of alcohol consumption in college students. Addict Behav 29, 311–321. [DOI] [PubMed] [Google Scholar]
  54. Forum Pew, 2008. US Religious Landscape Survey: Religious affiliation: Diverse and dynamic. The Pew Forum on Religion and Public Life, Pew Research Center, Washington, DC. [Google Scholar]
  55. Preacher KJ, Hayes AF, 2008. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods 40, 879–891. [DOI] [PubMed] [Google Scholar]
  56. Project MATCH Research Group, 1993. Project MATCH (Matching Alcoholism Treatment to Client Heterogeneity): rationale and methods for a multisite clinical trial matching patients to alcoholism treatment. Alcohol Clin Exp Res 17, 1130–1145. [DOI] [PubMed] [Google Scholar]
  57. Project MATCH Research Group, 1997. Matching Alcoholism Treatments to Client Heterogeneity: Project MATCH posttreatment drinking outcomes. J Stud Alcohol 58, 7–29. [PubMed] [Google Scholar]
  58. Schafer JL, Graham JW, 2002. Missing data: Our view of the state of the art. Psychological Methods 7, 147–177. [PubMed] [Google Scholar]
  59. Sobel ME, 1982. Asymptotic confidence intervals for indirect effects in structural equations models. In: Leinhart S (Ed.), Sociological Methodology 1982. Jossey-Bass, San Francisco, CA. pp. 290–312. [Google Scholar]
  60. Sobel ME, 1986. Some new results on indirect effects and their standard errors in covariance structure models. In: Tuma N (Ed.), Sociological Methodology 1986. American Sociological Association, Washington, DC. pp. 159–186. [Google Scholar]
  61. Stewart DG, Brown SA, 1995. Withdrawal and dependency symptoms among adolescent alcohol and drug abusers. Addiction 90, 627–635. [DOI] [PubMed] [Google Scholar]
  62. Substance Abuse and Mental Health Services Administration, 2010. Results from the 2009 National Survey on Drug Use and Health: Volume I. Summary of National Findings (NSDUH Series H-38A, HHS Publication No. SMA 10–4586Findings). Office of Applied Studies, Rockville, MD. [Google Scholar]
  63. Substance Abuse and Mental Health Services Administration, 2011. Results from the 2010 National Survey on Drug Use and Health: Summary of National Findings. In: Services U.S.D.o.H.a.H. (Ed.). Substance Abuse and Mental Health Services Administration, Rockville, MD. [Google Scholar]
  64. Sussman S, 2010. A review of Alcoholics Anonymous/Narcotics Anonymous programs for teens. Evaluation & the Health Professions 33, 26–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Tonigan JS, Connors GJ, Miller WR, 2003. Participation and involvement in Alcoholics Anonymous. Treatment Matching in Alcoholism, 184–204. [Google Scholar]
  66. White WL, 2008. Recovery Management and Recovery-Oriented Systems of Care: Scientific Rationale and Promising Practices. Northeast Addiction Technology Transfer Center, Great Lakes Addiction Technology Transfer Center, Philadelphia Department of Behavioral Health/Mental Retardation Services. [Google Scholar]
  67. Witbrodt J, Mertens J, Kaskutas LA, Bond J, Chi F, Weisner C, 2012. Do 12-step meeting attendance trajectories over 9 years predict abstinence? J Subst Abuse Treat 43, 30–43. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES