Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 30.
Published in final edited form as: Alcohol Clin Exp Res (Hoboken). 2026 May;50(5):e70284. doi: 10.1111/acer.70284

Predictors and trajectories of negative and positive interpersonal dynamics in remission from alcohol use disorder

Samuel F Acuff 1,*, Samuel N Meisel 2, Emily A Hennessy 1, Kyla L Belisario 3, Molly Garber 3, James MacKillop 3, John F Kelly 1
PMCID: PMC13416704  NIHMSID: NIHMS2192136  PMID: 42101822

Abstract

Background

Research has established social contextual factors as a critical determinant of recovery from alcohol use disorder (AUD), yet limited research exists investigating quality of interpersonal dynamics, especially during the transition to remission from AUD. Positive interpersonal dynamics (friendship, emotional support) may reinforce recovery, whereas negative interpersonal dynamics (hostility, rejection) may undermine it. Identifying determinants of interpersonal dynamic trajectories, including remission, may guide the development of personalized interventions.

Method

Naturalistic prospective study of individuals with moderate-severe AUD initiating a new recovery attempt (N=501, Mage=41.42, 57.5% female) who were assessed at baseline, and 1.5-, 3-, 6-, 9-, and 12-months later on measures of recovery capital and negative and positive interpersonal dynamics (NIH toolkit scales) and other measures assessing demographic, clinical, and psychological process variables (e.g., abstinence self-efficacy). Latent growth models examined trajectories and time-invariant predictors of positive and negative interpersonal dynamics.

Results

Negative interpersonal dynamics declined at the first follow-up and flattened over the remainder of the year. These results differed by remission status, with the greatest reduction in negative interpersonal dynamics among those who achieved abstinent remission at 12-month follow-up. Positive interpersonal dynamics did not change significantly throughout the study period. In covariate models, baseline recovery capital predicted positive and negative interpersonal dynamic intercepts and slopes and there were differences in significant predictors across remission groups.

Conclusions

Among those achieving AUD remission (abstinent or non-abstinent) at 12-months, early recovery is marked by reductions in negative interpersonal dynamics, in contrast to those who do not achieve remission. Regardless of outcome, little change was observed in positive interpersonal dynamics. The predictive utility of recovery capital highlights a modifiable outcome that could help improve the quality of interpersonal dynamics in recovery, although confirmation of this relationship is warranted.

Keywords: Alcohol, remission, social, interpersonal dynamics, recovery capital

Introduction

Each year approximately 30 million Americans meet criteria for alcohol use disorder (AUD), making it the most prevalent type of substance use disorder (SAMHSA, 2022). Although most enter sustained remission at some point in their lives, for many it takes decades (Lopez-Quintero et al., 2011). Initial and early recovery periods, defined as the first three months and first year, respectively (Hagman et al., 2022), are particularly risky periods for return to use (Andersson et al., 2019; Brecht & Herbeck, 2014; Kelly et al., 2019; Maisto et al., 2018; Nguyen et al., 2020), as people in recovery must contend with cravings, symptoms of withdrawal or post-acute withdrawal, anhedonia, and low social support. As a result, people with AUD make several recovery attempts on average before successful recovery is sustained (Kelly et al., 2019). Given the gravity of the public health crisis, identification of novel, malleable and enduring factors modulating recovery success could inform intervention targets that may improve care for patients with AUD.

Studies suggest that social context is a critical determinant of substance use broadly (Acuff et al., 2021; Bullers et al., 2001; Fairbairn & Sayette, 2014; Strickland & Acuff, 2023) and recovery from alcohol use disorder more specifically (Kelly et al., 2012, 2014; Meisel et al., 2023). The social identity model of recovery postulates that, in recovery, identity shifts from being defined by the norms and values of substance using social groups to being defined by the norms and values of recovery groups which causes, or is caused by, changes in social network composition (Best et al., 2016). Studies examining the influence of social dynamics on outcomes during the recovery stage of AUD have generally demonstrated that sustained recovery is marked by shifts away from substance-using to substance-free peers (Eddie & Kelly, 2017; Kelly et al., 2012; Martinelli et al., 2021; Stout et al., 2012), who can serve as buffers against return to use. Further, some of the effects of popular recovery support services (i.e., alcoholics anonymous) on sustained reductions in drinking have been shown to be partially mediated by changes in social network composition such that alcohol using peers are replaced over time by peers who do not use alcohol or who themselves are in, or supportive of, recovery (Kelly et al., 2008, 2012).

Less research has investigated shifts in the patterns and quality of relationships (i.e., interpersonal dynamics) during the early recovery period. Recovery from AUD often requires foregoing immediate reward (e.g., alcohol consumption with friends one feels comfortable with) and facing challenges and engaging in new activities, the rewards of which may be significantly delayed and uncertain. Interpersonal dynamics toward the end of active AUD may be largely negative, as harm toward others accumulates and pressure mounts from loved ones to make a change (Roessler et al., 2018). People in early recovery may simultaneously experience less criticism or hostility from loved ones that wanted them to make a change, while also experiencing less positive social connection because of loss in close friendships while support builds over time from recovery-supportive peers. However, little research has documented these kinds of changes in interpersonal dynamics. If, on average, changes are positive (increases in positive, and decreases in negative, interpersonal dynamics), interpersonal stress would likely decrease as this would be perceived as largely supportive. On the other hand, worsening negative relationship dynamics (decreases or no improvements in positive interpersonal dynamics) may increase interpersonal stress, creating a barrier inhibiting sustained change (Leach & Kranzler, 2013). Greater understanding of such dynamics and their correlates, as well as when they may change during the early phases of recovery could help inform the nature and timing of interventions that could be developed to improve them (Strickland & Acuff, 2023). These trajectories may differ according to whether the individual attempting change in their alcohol use is able to do so (e.g., whether they can successfully achieve AUD remission or not), presumably with greater remission status associated with greater positive interpersonal changes.

In addition to potentially supporting recovery (Harrison et al., 2017), changes in interpersonal dynamics represent a quality-of-life improvement for people in recovery and may lead to greater physical health, mental health, and longevity (Holt-Lunstad et al., 2010; Vila, 2021). Identifying predictors of changes in interpersonal dynamics in early recovery may also guide the nature and timing of additional support for those with particularly high negative or low positive interpersonal dynamics over time. Demographic factors such as age, sex, and income may partially determine interpersonal dynamics in AUD remission, though most research has been conducted in general populations (Akyunus et al., 2021). Broadly, people experience less conflict and greater satisfaction with their social relationships as they age, suggesting that interpersonal dynamics may be positively associated with age (Luong et al., 2011); however, AUD complicates and impacts this association, given that chronic AUD across the lifespan often deteriorates relationships and increases rates of conflict. Some data suggest that, in general populations, women are more connected socially than men (Shin & Park, 2023), and that women exhibit greater levels of qualities associated with positive interpersonal dynamics, such as responsiveness and warmth (Hopwood et al., 2020; Lin et al., 2019). Regarding income, though some research suggests that higher income is associated with more time spent in isolation (Bianchi & Vohs, 2016), other findings demonstrate that greater income buffers against the positive association between financial stress and negative interpersonal events at the day-level (Sturgeon et al., 2014).

Factors that are strong predictors of substance-focused recovery outcomes, such as severity of AUD, motivation to change, self-efficacy beliefs, recovery capital, and social network recovery support, may also predict interpersonal functioning in the early recovery period (Anderson et al., 2021; Hennessy, 2017; Kelly et al., 2019; Kelly & Greene, 2014). For example, those with higher levels of AUD severity at the start of a remission attempt would likely have more conflict and isolation, and for this reason higher negative and lower positive interpersonal dynamics. On the other hand, those with greater motivation to change may already be aligned in their remission goal with their immediate family network or isolated from substance using friends or partners. Alternatively, those with greater motivation to change may have significant recent negative consequences which may be interpersonal in nature. This may suggest poorer interpersonal dynamics at baseline that increase over the course of remission. Further still, those with greater abstinence self-efficacy may have lower AUD severity and greater recovery-related resources, which in turn may be associated with more positive interpersonal dynamics over time. Greater recovery capital, which is the sum total of all resources that aid in one’s recovery journey (Cloud & Granfield, 2008), is likely to buffer against interpersonal conflict and increase positive interpersonal interactions. Finally, greater recovery support in one’s social network is likely to be associated with more positive interpersonal dynamics over time, as successful remission requires shifting social connection from substance-oriented to recovery-oriented social networks.

The current pre-registered analyses extend the literature in two specific ways. First, we characterize trajectories of important negative (hostility, rejection) and positive (friendship, emotional support) domains of interpersonal dynamics during the first year after individuals with AUD started an AUD recovery attempt. Given the limited literature, while we expected change, we did not pre-register specific hypotheses regarding the characterization of the growth trajectories for negative and positive relational dynamics. In a deviation from the pre-registered aims, we also characterized differences in interpersonal dynamic scores based on whether remission (abstinent or non-abstinent) was achieved at 12 months. Second, we examined baseline, time-invariant predictors of trajectories of interpersonal dynamics. We hypothesized that male sex, higher baseline AUD severity, lower motivation for recovery, lower recovery capital, lower perceived income, lower coping skills, and lower baseline social support for recovery would be associated with shallower decline (i.e., less improvement) in negative interpersonal dynamics over time, and similarly that male sex, higher baseline AUD severity, and lower reported income would be associated with less improvements in positive interpersonal dynamics over time. We deviated from the pre-registered protocol to examine differences in baseline predictors associations with interpersonal dynamic trajectories split by remission status at 12-months. On both occasions, deviations were due to the realization that it would be important to observe whether interpersonal dynamic trajectories differed by 12-month remission status. Due to the large number of pre-planned analyses, this manuscript reports Aims 1-2 of the preregistration (https://osf.io/vgmf4/?view_only=286f623e8d55433d894193e4ccc65aba).

Methods

Participants

Participants were adults (N=501; 21-65 years of age) participating in or planning to participate in treatment or who self-identify as making a new AUD recovery attempt. To be eligible for the study, participants had to meet the following criteria: 1) DSM-5 diagnosis of current AUD (American Psychiatric Association, 2013), 2) high-risk drinking or greater per NIAAA (i.e., >14/7 standard drinks/week for males/females, respectively) in the 30 days pre-treatment or pre-recovery attempt, 3) alcohol as a “drug of choice” if other substances were used, 4) between 21-65 years old, and 4) participating in, or going to participate in, outpatient treatment at the time of enrollment, or making a new recovery attempt, that is, a serious effort to abstain from drinking or to drink without problems, in the past 90 days or the next 14 days. Individuals were excluded from the study based on the following criteria: 1) current psychosis, 2) active suicidality, 3) unstable domicile and contact information, and 4) below functional literacy (i.e., cannot read at level to complete questionnaires).

Procedure

Participants were recruited between May 2019 and September 2021 from outpatient treatment sites as well as community settings in Boston, MA, United States and from Hamilton, ON, Canada. Study visits were conducted by research staff at baseline and 4-6 weeks post baseline, in addition to 3-, 6-, and 12-months post-baseline. Assessments consisted of staff-administered and self-administered surveys via REDCap (Harris et al., 2019), and a computerized task to assess impulsivity. In-person assessments were counter-balanced where the participant switched back and forth between completing self-administered surveys and staff-administered surveys (i.e., interviews). After the declaration of emergency due to the COVID-19 pandemic, all assessments were conducted remotely.

Measures

Positive Interpersonal Dynamics

Past month positive interpersonal dynamics were measured with the Emotional Support (8-items) and Friendship (8-items) self-report scales from the adult National Institute of Health (NIH) Toolbox for the Assessment of Neurological and Behavioral Function (Cyranowski et al., 2013). The Emotional Support survey captured the perception that people in one’s social network are available to listen to one’s problems with empathy, caring, and understanding (e.g., “I have someone who understands my problems.”). The Friendship scale captured the perception of availability of friends in one’s social network (e.g., “I can find a friend when I need one.”). All items were rated on a 5-point scale from 1 (Never) to 5 (Always). Internal consistency in the current sample for the friendship (αrange = .93 - .95; ωrange = .93 - .95) and emotional support (αrange = .97 - .98; ωrange = .97 - .98) subscales was excellent across all assessments.

Negative Interpersonal Dynamics

Past month positive interpersonal dynamics were measured with the Perceived Hostility (8-items) and Perceived Rejection (8-items) self-report scales from the adult NIH Toolbox for the Assessment of Neurological and Behavioral Function (Cyranowski et al., 2013). These surveys captured the extent to which an individual perceives his/her daily social interactions as hostile (e.g., “People argue with me.”) or rejecting (e.g., “People don’t listen when I ask for help.”). All items were rated on a 5-point scale from 1 (Never) to 5 (Always). Internal consistency in the current sample for the perceived hostility (αrange = .93 - .95; ωrange = .93 - .95) and perceived rejection (αrange = .95 - .96; ωrange = .95 - .96) subscales was excellent across all assessments.

Time-Invariant Baseline Predictors

Demographic Predictors.

Sex assigned at birth (male or female) and perceived current household income (Not enough to pay some bills no matter how hard you try; Enough to pay bills but have had to cut back; Enough to pay bills without cutting back but no “extras”; Enough money for “extras”) were included as predictors of interpersonal dynamic trajectories (Najdzionek et al., 2023).

Recovery Capital.

Recovery capital was assessed using the Brief Assessment of Recovery Capital (BARC-10)(Vilsaint et al., 2017), which is a validated 10-item scale abridged from the Assessment of Recovery Capital (ARC) Scale (Groshkova et al., 2013). Items were rated on a 6-point scale from 1 (Strongly Disagree) to 6 (Strongly Agree) and assessed the level of broader personal, social, physical, and professional resources in an individual’s environment that may be used to support recovery (i.e., “I get lots of support from friends.”; α=.84; ω=.84).

Recovery Commitment.

Baseline recovery commitment was measured with the commitment to sobriety scale (CSS), a validated, brief 5-item self-report measure assesses the degree of commitment for abstinence (i.e., “Staying sober is the most important thing in my life.”). Items were rated on a 6-point scale from 1 (Strongly Disagree) to 6 (Strongly Agree; α=.94; ω=.94). The CSS has been shown to have strong convergent and discriminant validity in patients undergoing residential treatment for substance use disorder (Kelly & Greene, 2014).

Recovery Self-Efficacy.

Recovery Self-Efficacy was measured with the 20-item alcohol abstinence self-efficacy scale, which measures belief in one’s ability to remain abstinent in various scenarios. Items were rated on a 6-point scale from 1 (Not at all confident) to 5 (Extremely confident; DiClemente et al., 1994; α=.95; ω=.95).

AUD Symptoms.

AUD Severity and diagnostic status was measured with the Alcohol Use Disorder from the Diagnostic Assessment Research Tool, a validated semi-structured interview (AUD DART)(Garber et al., 2024). The first question was “In the past 3 months, have you consumed any alcohol?” If the participant answered ‘yes’, the DART would pertain to the past 3 months. At the final assessment, if the participant answered ‘no’, they would be asked if they consumed alcohol during the past 12 months, and, if yes, the DART would pertain to the past 12 months. We used a sum score of all AUD symptoms as the independent time-invariant predictor variable.

Social network recovery engagement.

Baseline network recovery engagement was measured using an index derived from an egocentric social network analysis. The participant listed the names of 20 individuals important in their life (e.g., family, friends, coworkers, treatment providers, etc.) based on two criteria: all individuals (1) must be greater than 12 years of age, and (2) must have been in contact with the participant in the past three months. Participants answered the following questions about each of the 20 individuals: gender, length of time known by participant, relationship to participant, frequency of interaction with participant, closeness to participant, alcohol use in past month, cannabis use in past month, illicit drug use in the past month, and whether member of a recovery group. We used the proportion of people in the person’s network who is a member of a recovery group as an index of baseline network recovery engagement. All alter data was provided by the ego and thus is based on the ego's perception.

Data Analysis

Analyses were conducted in Mplus v8 (Muthén & Muthén, 2017). Missing data were handled in Mplus with Full Information Maximum Liklihood (FIML). Overall, retention was strong (89.6% at 1-month, 88.4% at 3-month, 81.2% at 6-month, 76.6% at 9-month, and 79.8% at 12-month), and 84% completed at least 4 assessments (Supplemental Table 1). We created latent negative and positive interpersonal dynamic factors and tested longitudinal structural invariance across time to ensure that change is not accounted for by variance in measurement across assessments (Supplemental Tables S2 and S5). We used Dynamic Fit Indices to evaluate model fit for our measurement models (McNeish & Wolf, 2023). To reduce model complexity, factor scores were extracted and used as indicators of interpersonal dynamics across subsequent models.

First, in a deviation from the pre-registered analyses, we explored differences in raw negative and positive interpersonal dynamic scores at baseline and the 12-month follow up split by remission group using paired samples t-tests to characterize raw change in interpersonal dynamics across the study period in each group. Remission status was calculated to create 3 distinct groups, reflective of the NIAAA definition of remission (Kelly et al., 2025): (i) those in abstinent remission; (ii) those in non-abstinent remission; and iii) those who do not meet the criteria for remission.

Next, unconditional latent growth curve models (LGCM) were estimated separately for positive and negative interpersonal dynamics. We evaluated multiple forms of growth (e.g., intercept only, linear, latent basis, quadratic). The Satorra-Bentler chi-square difference test (for non-normally distributed outcome variables modeled using Maximum Likelihood Robust estimation)(Satorra & Bentler, 2001) was used to evaluate the best fitting model. In a second deviation from the pre-registered analyses, we then examined latent growth curves split by AUD remission groups to further characterize changes in interpersonal dynamics as a function of AUD remission status at 12-month.

Next, we used time-invariant baseline variables as predictors of the previously established growth models for positive and negative interpersonal dynamics. Each predictor was first examined in a separate model, and significant predictors were included in a final omnibus model. In a third deviation from the protocol, we also explored whether baseline time-invariant variables were more or less predictive based on AUD remission status.

Results

Demographics and Raw Change in Interpersonal Dynamics Split by AUD Remission

Correlations amongst all study variables can be found in supplemental table S6. Of the full sample, 66.1% were not in remission at the 12-month follow up, whereas 17.6% met criteria for non-abstinent remission and 16.6% met criteria for abstinent remission. Changes in negative and positive interpersonal dynamics from baseline to 12-month follow-up can be found in Figure 1. Among those who did not meet criteria for remission, there was a significant, but small decline in negative interpersonal dynamics from baseline to 12-month follow up (t[208]=1.98, p=.049, Cohen’s d=.14), and small-to-medium significant declines in negative interpersonal dynamic declines for those meeting criteria for non-abstinent remission (t[53]=2.91, p=.003, Cohen’s d=.40) or abstinent remission (t[61]=3.20, p=.001, Cohen’s d=.41). There were no differences in positive interpersonal dynamics from baseline to the 12-month follow up for the no remission group (t[197]=−0.72, p=.47, Cohen’s d=−.05), the non-abstinent remission group (t[51]=.34, p=.74, Cohen’s d=.05), or the abstinent remission group (t[64]=−1.22, p=.23, Cohen’s d=−.15).

Figure 1. Differences in Negative Interpersonal Dynamics at Baseline and 12-month Follow Up by 12-month Remission Status.

Figure 1

Open in a new tab

Note. Differences in negative (left) and positive (right) interpersonal dynamics from baseline to 12-month follow up, split by remission group at the 12-month follow up. Paired samples t-test indicates a significant medium reduction in negative interpersonal dynamics among those in abstinent remission, a significant but small reduction in negative interpersonal dynamics among those in non-abstinent remission, and a significant but negligible reduction in negative interpersonal dynamics among those who did not achieve remission. There were no significant differences in positive interpersonal dynamics across all three remission groups from baseline to 12-month follow up.

Negative Interpersonal Dynamics

Fit statistics for the negative interpersonal dynamics LGCM are reported in supplemental table S7. A nonlinear growth model (i.e. latent basis) with the slope loading freely estimated at the 1-month, 3-month, 6-month, and 9-month follow-ups provided the best fit to the data (χ2 [12]=22.77, p=.03, CFI=0.99, TLI=0.99, RMSEA=.04, SRMR=.03). Visual representation of the raw means and estimated model fit for negative interpersonal dynamics is illustrated in Figure 2. Negative interpersonal dynamics decreased significantly over the course of the year-long follow-up (Mean [M] = −.37, standard error [SE] = .06, p <.001). There was significant variability in negative interpersonal dynamics at the baseline assessment (Variance [σ2]=.80, [SE]=.09, p<.001) and in change over time (σ2=.17, SE=.06, p<.001).

Figure 2. Trajectories of Positive (Green) and Negative (Red) Interpersonal Dynamics among Adults in Early Recovery from Alcohol use Disorder.

Figure 2.

Open in a new tab

Note. Mean-level trajectories of negative interpersonal dynamics (red) and positive interpersonal dynamics (green) over the first year of recovery. Negative and positive interpersonal dynamics are standardized.

Estimated and raw means for negative interpersonal dynamics split by remission group can be found in Figure 3. Slope variance across all three groups was fixed to zero to achieve convergence. There was significant variability in negative interpersonal dynamics at the baseline assessment for the no remission (σ2=.73, SE=.06, p<.001), non-abstinent remission (σ2=.64, SE=.11, p<.001), and abstinent remission groups (σ2=.66, SE=.12, p<.001). In the no remission group, there was a negligible decline from baseline to the 1-month follow up, and no additional decline across the remainder of the study period (Slope M = −.14, SE = .03, p <.001). Negative interpersonal dynamics were the lowest at baseline for the non-abstinent remission group, declined from baseline to the 1-month follow up, and remained stable across the remainder of the study period (Slope M=−.20, SE=.05, p<.001). Negative interpersonal dynamics at baseline were the highest across all three remission groups in the abstinent remission group, declined steeply from baseline to 1-month, and continued a shallow decline across the remainder of the study period, such that negative interpersonal dynamics were below that of the no remission group (Slope M=−.35, SE=.09, p<.001).

Figure 3. Negative and Positive Interpersonal Dynamic Trajectories Split by Remission Status.

Figure 3.

Open in a new tab

Note. Trajectories of negative (left) and positive (right) interpersonal dynamics, split by remission group at the 12-month follow up. Among those in non-abstinent remission, there was a sharp decrease in negative interpersonal dynamics from baseline to 1-month, which decreased slightly across the remaining study period. Among those in non-abstinent remission, negative interpersonal dynamic started at a lower point than dynamics for the other two groups, and decreased significantly, primarily from baseline to 1-month. Among those who did not achieve remission at 12-month, negative interpersonal dynamics started below those in abstinent remission group and above those in the non-abstinent remission group. However, the trajectory remained relatively stable with a negligible reduction, and ended with the highest rates of negative interpersonal dynamics at 12-months. There was no significant change in positive interpersonal dynamics across the study period; However, those in the abstinent remission group had the highest rates of positive interpersonal dynamics, followed by those in the non-abstinent remission group and the no remission group, respectively.

Positive Interpersonal Dynamics

Fit statistics for the positive interpersonal dynamics LGCM are reported in supplemental table S8. A nonlinear growth model with the slope loading freely estimated at the 6-month and 9-month follow-ups provided the best fit to the data (χ2 [14]=30.25, p=.007, CFI=0.99, TLI=0.99, RMSEA=.05, SRMR=.02). Visual representation of the raw means and estimated model fit for positive interpersonal dynamics is illustrated in Figure 1. There was a positive but nonsignificant increase in positive interpersonal dynamics over the course of the study (M=.07, SE=.08, p=.37). There was significant variability in positive interpersonal dynamics at the baseline assessment (σ2=.74, SE=.05, p<.001) and in change over time (σ2=.11, SE=.05, p<.001).

Estimated and raw means for positive interpersonal dynamics split by remission group can be found in Figure 3. There was significant variability in positive interpersonal dynamics at the baseline assessment for the no remission (σ2=.78, SE=.06, p<.001), non-abstinent remission (σ2=.80, SE=.14, p<.001), and abstinent remission groups (σ2=.60, SE=.12, p<.001). The no remission group had the lowest positive interpersonal dynamics across the study period, followed by the non-abstinent remission group and the abstinent remission group. As demonstrated by non-significant mean slopes, positive interpersonal dynamics remained relatively flat for the no remission group (Slope M=.02, SE=.03, p=.66), the non-abstinent remission group (Slope M=.03, SE=.05, p=.55), and the abstinent remission group (Slope M=.08, SE=.08, p=.33). There was significant variability in positive interpersonal dynamic slopes for the no remission group (σ2=.11, SE=.04, p=.01), but no variability for the non-abstinent remission group (σ2=.12, SE=.07, p=.10) or the abstinent remission group (σ2=.26, SE=.16, p=.11).

Baseline Predictors of Negative Interpersonal Dynamic Trajectories

Results of univariate and omnibus models evaluating associations between baseline time-invariant predictors and growth factors can be found in Table 2. In univariate models, baseline negative interpersonal dynamics were significantly negatively associated with baseline perceived financial status, recovery capital, and abstinence self-efficacy, and significantly positively associated with AUD symptom count, recovery motivation, and the number of people in the social network attending mutual help groups. Higher baseline AUD symptoms and recovery motivation were significantly associated with greater reductions in negative interpersonal dynamics over the assessment period. In an omnibus model that included all significant predictors of a negative interpersonal dynamic growth factor, baseline associations between financial status, AUD symptoms, recovery capital, and recovery motivation remained; only recovery capital was a significant predictor of change in negative interpersonal dynamics over time. Lower recovery capital at baseline was associated with greater reductions in negative interpersonal dynamics over the course of the study (Figure 2).

Table 2.

Time-Invariant Baseline Predictors of Negative Interpersonal Dynamic Trajectories.

Intercept Slope
Variable Est. SE p-value Est. SE p-value
Individual Models – Full Sample
Sex 0.08 0.05 0.12 −0.07 0.07 0.32
Perceived Income −0.26 0.05 <.001 0.08 0.06 0.21
AUD Symptoms 0.25 0.04 <.001 −.13 0.06 0.02
Recovery Capital −0.43 0.05 <.001 0.11 0.06 0.07
Recovery Commitment 0.14 0.05 0.004 −0.13 0.06 0.046
Recovery Self-Efficacy −0.11 0.06 0.047 −0.04 0.06 0.47
Social Network Recovery Engagement 0.13 0.05 0.01 −0.08 0.07 0.21
Omnibus Model – Full Sample
Perceived Income −0.17 0.05 <.001 0.06 0.07 0.36
AUD Symptoms 0.19 0.07 <.001 −0.11 0.07 0.09
Recovery Capital −0.49 0.05 <.001 0.18 0.07 0.008
Recovery Commitment 0.13 0.06 0.03 −0.06 0.08 0.47
Recovery Self-Efficacy 0.08 0.06 0.17 −0.12 0.07 0.08
Social Network Recovery Engagement 0.002 0.05 0.69 −0.04 0.08 0.64
Omnibus Model – No Remission
Perceived Income −0.14 0.05 .007 0.01 0.07 .84
AUD Symptoms 0.10 0.06 .10 −0.07 0.07 .29
Recovery Capital −.40 0.06 <.001 0.12 0.07 .06
Recovery Commitment 0.12 0.07 .09 −0.05 0.09 .57
Recovery Self-Efficacy −0.01 0.06 .94 −0.01 0.06 .94
Social Network Recovery Engagement −0.04 0.09 .65 −0.04 0.09 .65
Omnibus Model – Non-abstinent Remission
Perceived Income −0.28 0.09 .003 0.39 0.12 .001
AUD Symptoms 0.28 0.09 .001 −0.17 0.11 .12
Recovery Capital −0.57 0.08 <.001 0.06 0.11 .57
Recovery Commitment 0.04 0.10 .70 0.28 0.11 .01
Recovery Self-Efficacy 0.14 0.09 .11 0.04 0.10 .68
Social Network Recovery Engagement 0.04 0.08 .65 −0.03 0.12 .81
Omnibus Model – Abstinent Remission
Perceived Income −0.08 0.10 .44 −0.02 0.13 .85
AUD Symptoms 0.36 0.08 <.001 −0.17 0.08 .045
Recovery Capital −0.54 0.09 <.001 0.35 0.10 <.001
Recovery Commitment 0.20 0.09 .04 −0.22 0.11 0.05
Recovery Self-Efficacy 0.16 0.09 .07 −0.34 0.13 .008
Social Network Recovery Engagement −0.12 0.10 .23 0.02 0.11 .85
Open in a new tab

Note. Est. = Estimate; S.E. = Standard Error.

Results of the baseline (time-invariant) predictor analyses split by remission status can be found in Table 2. Among those who were not in remission after 12 months, higher baseline income and recovery capital was associated with less negative interpersonal dynamics at baseline; however, no other association was significant. Among those in non-abstinent remission after 12 months, higher baseline income and recovery capital, and lower baseline AUD severity was associated with less negative interpersonal dynamics at baseline. Further, greater income and recovery commitment was associated with a shallower decline in negative interpersonal dynamics over time. Among those in abstinent remission after 12 months, higher baseline recovery capital, and lower baseline AUD severity and recovery commitment, were associated with less negative interpersonal dynamics at baseline. Greater AUD symptoms and recovery commitment at baseline was associated with a steeper decline in negative interpersonal dynamics, whereas higher recovery capital was associated with shallower declines in negative interpersonal dynamics.

Baseline Predictors of Positive Interpersonal Dynamic Trajectories

Results of univariate and omnibus models evaluating associations between baseline (time-invariant) predictors and growth factors can be found in Table 3. In univariate models, higher baseline positive interpersonal dynamics were significantly positively associated with baseline financial status, recovery capital, and abstinence self-efficacy. Higher baseline recovery capital and abstinence self-efficacy significantly predicted greater, but shallower, increases in positive interpersonal dynamics over the assessment period. In an omnibus model including all significant predictors of a positive interpersonal dynamic growth factor, baseline associations between perceived financial status and recovery capital remained. In the omnibus model, only recovery capital was a significant predictor of change in positive interpersonal dynamics over time. Higher recovery capital at baseline was associated with shallower increases in positive interpersonal dynamics over the course of the study (Figure 4).

Table 3.

Time-Invariant Baseline Predictors of Positive Interpersonal Dynamic Trajectories.

Intercept Slope
Variable Est. SE p-value Est. SE p-value
Sex 0.05 0.05 0.29 0.05 0.8 0.53
Perceived Income 0.17 0.05 <.001 0 0.08 0.99
AUD Symptoms −0.04 0.05 0.38 0.07 0.08 0.44
Recovery Capital 0.56 0.03 <.001 −0.27 0.07 <.001
Recovery Commitment 0.02 0.05 0.68 −0.01 0.13 0.95
Recovery Self-Efficacy 0.21 0.05 <.001 −0.17 0.07 0.02
Social Network Recovery Engagement −0.002 0.29 0.99 0.06 0.15 0.68
Omnibus Model – Full Sample
Perceived Income 0.11 0.04 0.01 0.05 0.08 0.51
Recovery Capital 0.57 0.04 <.001 −0.25 0.08 0.003
Recovery Self-Efficacy −0.04 0.05 0.4 −0.06 0.08 0.47
Omnibus Model – No Remission
Perceived Income 0.11 0.05 .04 −0.03 0.10 .80
Recovery Capital 0.56 0.05 <.001 −0.37 0.10 <.001
Recovery Self-Efficacy −0.07 0.06 .24 −0.03 0.10 .24
Omnibus Model – Non-abstinent Remission
Perceived Income 0.22 0.10 .02 −0.19 0.16 .25
Recovery Capital 0.56 0.10 <.001 0.14 0.21 .50
Recovery Self-Efficacy −0.04 0.15 .79 −0.58 0.23 .01
Omnibus Model – Abstinent Remission
Perceived Income −0.002 0.11 .99 0.41 0.12 .001
Recovery Capital 0.55 0.12 <.001 −0.18 0.13 .17
Recovery Self-Efficacy 0.10 0.12 .40 −0.001 0.13 .99
Open in a new tab

Note. Est. = Estimate; S.E. = Standard Error.

Figure 4. Trajectories of Negative and Positive Interpersonal Dynamics over One Year Split by Recovery Capital at Baseline.

Figure 4.

Open in a new tab

Note. Mean-level trajectories of negative interpersonal dynamics (left) and positive interpersonal dynamics (right) over the first year of recovery split at the median into low recovery capital (red) and high recovery capital (green). Negative and positive interpersonal dynamics are standardized.

Among those who were not in remission after 12 months, higher baseline income and recovery capital was associated with greater positive interpersonal dynamics at baseline, and greater recovery capital was associated with shallower improvements in positive interpersonal dynamics over time. Among those in non-abstinent remission after 12 months, higher baseline income and recovery capital was associated with greater positive interpersonal dynamics at baseline, and higher baseline self-efficacy was associated shallower improvements in positive interpersonal dynamics over time. Among those in abstinent remission after 12 months, only higher baseline recovery capital was associated with greater positive interpersonal dynamics at baseline. Further, higher baseline perceived income was associated with a steeper improvement in positive interpersonal dynamics.

Discussion

The current study characterized changes in self-reported interpersonal dynamics in the first year of a recovery attempt from AUD. Consistent with our hypotheses, we found evidence for decreases in negative interpersonal dynamics (perceived rejection and hostility) in the first month of recovery that were sustained through the 12-month follow-up. These increases were negligible for those who were not in remission at the 12-month follow up and were small-to-medium in magnitude for those who were in AUD remission (either non-abstinent or abstinent remission). However, we found that there were no changes in positive interpersonal dynamics across all three remission groups. Omnibus time-invariant models examining baseline predictors found that perceived income, AUD symptoms, recovery capital, and recovery commitment at baseline were all associated with negative interpersonal dynamics at baseline or with subsequent growth trajectories, although there were differences in their predictive utility across remission groups. Similarly, perceived income and recovery capital were associated with positive interpersonal dynamics, with differences detected in their predictive validity across remission groups.

Previous research has identified negative interpersonal experiences (e.g., family or friend conflict, marital trouble, embarrassing family, and disapproval from others) as among the most common reasons people provide for wanting to reduce their alcohol consumption (Prestigiacomo et al., 2024) and a common driver of seeking treatment. Therefore, as might be anticipated, our findings suggest that reducing or eliminating problematic alcohol use and achieving AUD remission is likely to result in reductions in negative interpersonal dynamics, which may encourage more ambivalent individuals to consider making a change. It is important to note that mean level reductions were relatively small in magnitude. That said, it is unclear to what degree the kind of magnitude interpersonal improvements on the dimensional measures observed in this study may translate into real-world subjective and functional social improvements. Further research using multimodal assessments may help uncover this.

In contrast to the observed salutary changes in negative interpersonal dynamics, our study did not find increases in positive interpersonal dynamics (friendship and emotional support), which was contrary to our hypothesis. Further, changes did not occur in any remission group. This suggests that in the first year following an AUD recovery attempt, it may be easier to decrease negative interpersonal dynamics – especially if one is able to reduce or eliminate problematic alcohol use – but more difficult to improve positive interpersonal dynamics, which may take more time. These findings mirror others (Kelly et al., 2011) which similarly have found that substantial reductions in relapse risk for AUD patients were more strongly associated with reductions in negative heavy drinking social network members than in the accrual of positive abstinent/recovering social network members. Unraveling some of the nuanced reasons for this is likely to require further incisive qualitative investigations.

The lack of observed increases in positive interpersonal dynamics is concerning due to the already complex and difficult biobehavioral challenges associated with early recovery. Those reducing or abstaining from alcohol often must contend with craving, withdrawal, post-acute withdrawal symptoms, and, at times, anhedonic response to pleasurable experiences unrelated to alcohol or other drug use. Recovery may be enhanced by attempting to increase immediate reinforcement from recovery-oriented, and more broadly, non-drug, activities and experiences (Acuff et al., 2023; Acuff, Oddo, et al., 2024; McKay, 2017), and one identifiable target may be to improve positive interpersonal experiences as has been identified in behavioral couples therapies for alcohol/drug use disorders (O’Farrell & Schein, 2011).

In the full sample, only baseline recovery capital was associated with changes in negative interpersonal dynamics over time in the omnibus model. Data descriptives suggest a strong association between recovery capital and the negative interpersonal dynamic intercept, such that those with high recovery capital at baseline have approximately a half standard deviation lower negative interpersonal dynamic at baseline. As a result, those with low recovery capital have much greater subsequent declines in negative interpersonal dynamics, particularly during initial recovery, but still report significantly greater negative interpersonal dynamics across the study period. However, time-invariant predictor results differed across remission groups, particularly when predicting the negative interpersonal dynamic growth trajectory. None of the predictors were associated with negative interpersonal dynamics in the no remission group, perhaps suggesting that current alcohol use may better account for variance in negative interpersonal dynamics than these baseline variables.

Among those in non-abstinent remission, baseline higher perceived income was associated with a lower intercept and a shallower change trajectory. This suggests that, among those who ultimately achieve non-abstinent remission, a higher perceived income may help protect against negative interpersonal dynamics. Further, higher commitment to sobriety among those who achieved non-abstinent remission was associated with less change in negative interpersonal dynamics. This may be because the measure focuses on commitment to sobriety/abstinence, and those who were highly committed to abstinence at the start of their recovery but who continue to drink, even if achieving AUD remission, may still experience some frustration or conflict within their immediate social network (e.g., spouse). Among those in abstinent remission, higher baseline recovery self-efficacy and greater baseline AUD symptomology was associated with larger reductions in negative interpersonal dynamics. This suggests that someone with more severe AUD who achieves substantive change in drinking will likely experience greater reductions in negative interpersonal dynamics, negatively reinforcing recovery-oriented behaviors and help people with severe AUD maintain sobriety. Recovery capital also predicted negative interpersonal dynamics in the abstinent remission group, the only group that mirrored the effect of recovery capital in the overall sample. This suggests that low baseline recovery capital is associated with greater subsequent negative interpersonal dynamic declines but overall greater absolute levels of negative interpersonal dynamics across the study period.

Within the full sample, only income and recovery capital were significantly associated with the positive interpersonal dynamic intercept, and only recovery capital was associated with changes in positive interpersonal dynamic trajectories over time in models controlling for relevant covariates. Similar to our findings with negative interpersonal dynamics, the results were in the opposite direction than hypothesized: there were greater improvements in positive interpersonal dynamics for those with low recovery capital, yet intercept differences between these two groups demonstrate that positive interpersonal dynamics remain much lower for those with low recovery capital, compared to those with high recovery capital, across the study period. However, predictor analyses differed by remission group. Among those who did not achieve any form of remission, the results were the same as the full sample. However, results differed within the non-abstinent and abstinent remission groups. In the non-abstinent remission group, while perceived income and recovery capital were positively associated with positive interpersonal dynamics at baseline, only recovery self-efficacy was associated with positive interpersonal dynamic trajectories, such that higher recovery self-efficacy at baseline was associated with flatter increases in positive interpersonal dynamics over time. Among those in the abstinent recovery group, only recovery capital was positively associated with positive interpersonal dynamics at baseline, whereas greater perceived income at baseline was associated with greater increases in positive interpersonal dynamics over time. These results suggest that those with greater perceived income may yield additional benefits from abstinent recovery potentially due to increased resources to engage in positive substance-free reinforcement (Acuff, Ellis, et al., 2024) and positive social interactions with their social network.

Findings underscore a recognized need to target interpersonal interactions among people attempting to recover from AUD (McCrady, 2004; Wright et al., 2023). Negative interpersonal dynamics only reduced, on average, by less than half a standard deviation among those in the abstinent remission group over the course of the study. Further, positive interpersonal dynamics improved, on average, only one-tenth of a standard deviation among those in the abstinent remission group over the study period, and this change was not significant. From a behavioral perspective, negative interactions prior to a remission attempt, if linked clearly to alcohol consumption, may increase the chances the individual will be motivated to reduce drinking. Reductions in these negative interpersonal dynamics after starting the remission attempt could motivate the individual to continue their recovery journey. However, remission rates might be improved if family and friends were trained to positively reinforce recovery-oriented behavior through positive social interaction (McCrady, 2004, 2012). Treatments such as Community Reinforcement Approach and Family Training, or CRAFT, are useful approaches that provide such training, typically prior to an individual initiating a recovery attempt (Meyers et al., 1998; Smith et al., 2008). However, few loved ones receive CRAFT. Future research may identify novel approaches to improving interpersonal dynamics that can be more effectively disseminated, or that target family members after or during treatment to help them maximize likelihood of remission.

Limitations & Future Directions

The current study used complex analyses to investigate reciprocal associations between interpersonal dynamics and recovery outcomes in a large sample over the course of one year. Our study is, however, not without limitations. First, our data did not ascertain the recovery goal (i.e., reduction or abstinence) and we are unable to determine recovery success or differences in interpersonal dynamics by recovery goal (rather than outcome). Second, interpersonal dynamics were not split by recovery support status. It is likely that positive and negative interpersonal dynamics are likely to change differentially for those who are supportive of recovery and those who encourage continued use. Future research may consider parsing apart interpersonal dynamics across the social network. Third, interpersonal dynamics likely play out on a more fine-grained temporal resolution than when assessed over mostly 90-day periods, as was done in this study. Future research may consider daily or even momentary assessment to determine whether associations are stronger on a more temporally constrained time window. Fourth, while our study characterized interpersonal dynamic change trajectories and tested baseline predictors of these trajectories, we are unable to answer how interpersonal dynamics, and their association with remission outcomes, are targeted or affected by recovery support services or treatment. It is possible that some behavior change techniques may be particularly helpful in facilitating change in interpersonal dynamics, and future research should explore novel approaches to changing interpersonal dynamics.

Conclusions

The current study is novel in that it focused on the interpersonal quality of relationships rather than merely the alcohol-using status of social network members finding that negative interpersonal dynamics decreased during the first year following a recovery attempt, and that this effect was much greater for those who achieved abstinent or non-abstinent remission compared to those who did not achieve remission from AUD. Positive interpersonal dynamics did not change across the study period. Predictors of interpersonal dynamics differed across remission groups, though perceived income, recovery capital, and recovery self-efficacy appeared to be the most influential predictors. Future research may consider exploring varying interpersonal dynamics across different contingents within the social network, including those that do and do not support AUD recovery. Also, the use of more granular assessment approaches (e.g., ecological momentary assessment) may prove fruitful in uncovering the exact nature of the interplay among positive and negative interpersonal dynamics and changes in alcohol use within the recovery process.

Supplementary Material

Supplemental

Table 1.

Socio-demographics and measures at baseline.

Variable % M (SD)
Age 41.42 (11.22)
Sex (% female) 57.5%
Income
 Not enough to pay some bills no matter how hard you try 14.9%
 Enough to pay bills, but have had to cut back 30.5%
 Enough to pay bills without cutting back but no “extras” 23.1%
Enough money for “extras” 31.5%
AUD Symptoms 8.15 (2.52)
Recovery Capital (10-60) 42.72 (8.14)
Recovery Commitment (5-30) 20.69 (6.75)
Recovery Self-Efficacy (0-80) 57.24 (16.38)
Social Network Recovery Engagement 8.63% (14.07%)
Negative Interpersonal Dynamics
 Perceived Hostility (8-40) 18.54 (6.25)
 Perceived Rejection (8-40) 18.04 (6.91)
Positive Interpersonal Dynamics
 Friendship (8-40) 25.73 (8.14)
 Emotional Support (8-40) 30.39 (7.83)
Open in a new tab

Note. Est. = Estimate; SD = Standard Deviation.

References

  1. Acuff SF, Ellis JD, Rabinowitz JA, Hochheimer M, Hobelmann JG, Huhn AS, & Strickland JC (2024). A brief measure of non-drug reinforcement: Association with treatment outcomes during initial substance use recovery. Drug and Alcohol Dependence, 256, 111092. 10.1016/j.drugalcdep.2024.111092 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Acuff SF, MacKillop J, & Murphy JG (2023). A contextualized reinforcer pathology approach to addiction. Nature Reviews Psychology, 2, 309–323. 10.1038/s44159-023-00167-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Acuff SF, Oddo LE, Johansen AN, & Strickland JC (2024). Contextual and psychosocial factors influencing drug reward in humans: The importance of non-drug reinforcement. Pharmacology Biochemistry and Behavior, 241, 173802. 10.1016/j.pbb.2024.173802 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Acuff SF, Stoops WW, & Strickland JC (2021). Behavioral economics and the aggregate versus proximal impact of sociality on heavy drinking. Drug and Alcohol Dependence, 220(March 2021), 108523. 10.1016/j.drugalcdep.2021.108523 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Akyunus M, Gençöz T, & Aka BT (2021). Age and sex differences in basic personality traits and interpersonal problems across young adulthood. Current Psychology, 40(5), 2518–2527. 10.1007/s12144-019-0165-z [DOI] [Google Scholar]
  6. American Psychiatric Association. (2013). DSM-5. In Diagnostic and statistical manual of mental disorders. American Psychiatric Publishing. [Google Scholar]
  7. Anderson M, Devlin AM, Pickering L, McCann M, & Wight D (2021). “It’s not 9 to 5 recovery”: The role of a recovery community in producing social bonds that support recovery. Drugs: Education, Prevention & Policy, 28(5), 475–485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Andersson HW, Wenaas M, & Nordfjærn T (2019). Relapse after inpatient substance use treatment: A prospective cohort study among users of illicit substances. Addictive Behaviors, 90, 222–228. 10.1016/j.addbeh.2018.11.008 [DOI] [PubMed] [Google Scholar]
  9. Best D, Beckwith M, Haslam C, Alexander Haslam S, Jetten J, Mawson E, & Lubman DI (2016). Overcoming alcohol and other drug addiction as a process of social identity transition: The social identity model of recovery (SIMOR). Addiction Research & Theory, 24(2), 111–123. 10.3109/16066359.2015.1075980 [DOI] [Google Scholar]
  10. Bianchi EC, & Vohs KD (2016). Social Class and Social Worlds: Income Predicts the Frequency and Nature of Social Contact. Social Psychological and Personality Science, 7(5), 479–486. 10.1177/1948550616641472 [DOI] [Google Scholar]
  11. Brecht M-L, & Herbeck D (2014). Time to relapse following treatment for methamphetamine use: A long-term perspective on patterns and predictors. Drug and Alcohol Dependence, 139, 18–25. 10.1016/j.drugalcdep.2014.02.702 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bullers S, Cooper ML, & Russell M (2001). Social network drinking and adult alcohol involvemen: A longitudinal exploration of the direction of influence. Addictive Behaviors, 26(2), 181–199. 10.1016/S0306-4603(00)00099-X [DOI] [PubMed] [Google Scholar]
  13. Cloud W, & Granfield R (2008). Conceptualizing recovery capital: Expansion of a theoretical construct. Substance Use & Misuse, 43(12–13), 1971–1986. 10.1080/10826080802289762 [DOI] [PubMed] [Google Scholar]
  14. Cyranowski JM, Zill N, Bode R, Butt Z, Kelly MAR, Pilkonis PA, Salsman JM, & Cella D (2013). Assessing Social Support, Companionship, and Distress: NIH Toolbox Adult Social Relationship Scales. Health Psychology : Official Journal of the Division of Health Psychology, American Psychological Association, 32(3), 293–301. 10.1037/a0028586 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. DiClemente CC, Carbonari JP, Montgomery RP, & Hughes SO (1994). The Alcohol Abstinence Self-Efficacy scale. Journal of Studies on Alcohol, 55(2), 141–148. 10.15288/jsa.1994.55.141 [DOI] [PubMed] [Google Scholar]
  16. Eddie D, & Kelly JF (2017). How many or how much? Testing the relative influence of the number of social network risks versus the amount of time exposed to social network risks on post-treatment substance use. Drug and Alcohol Dependence, 175, 246–253. 10.1016/j.drugalcdep.2017.02.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fairbairn CE, & Sayette MA (2014). A Social-Attributional Analysis of Alcohol Response. Psychological Bulletin, 140(5), 1361–1382. 10.1037/a0037563 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Garber ML, Belisario K, Levitt EE, McCabe RE, Kelly J, & MacKillop J (2024). Psychometric validation of the Diagnostic Assessment Research Tool: Alcohol use disorder module. Alcohol and Alcoholism (Oxford, Oxfordshire), 60(1), agae088. 10.1093/alcalc/agae088 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Groshkova T, Best D, & White W (2013). The Assessment of Recovery Capital: Properties and psychometrics of a measure of addiction recovery strengths. Drug and Alcohol Review, 32(2), 187–194. 10.1111/j.1465-3362.2012.00489.x [DOI] [PubMed] [Google Scholar]
  20. Hagman BT, Falk D, Litten R, & Koob GF (2022). Defining Recovery From Alcohol Use Disorder: Development of an NIAAA Research Definition. American Journal of Psychiatry, 179(11), 807–813. 10.1176/appi.ajp.21090963 [DOI] [PubMed] [Google Scholar]
  21. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, McLeod L, Delacqua G, Delacqua F, Kirby J, & Duda SN (2019). The REDCap Consortium: Building an International Community of Software Platform Partners. Journal of Biomedical Informatics, 95, 103208. 10.1016/j.jbi.2019.103208 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Harrison AJ, Timko C, & Blonigen DM (2017). Interpersonal Styles, Peer Relationships, and Outcomes in Residential Substance Use Treatment. Journal of Substance Abuse Treatment, 81, 17–24. 10.1016/j.jsat.2017.07.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hennessy EA (2017). Recovery capital: A systematic review of the literature. Addiction Research & Theory, 25(5), 349–360. 10.1080/16066359.2017.1297990 [DOI] [Google Scholar]
  24. Holt-Lunstad J, Smith TB, & Layton JB (2010). Social relationships and mortality risk: A meta-analytic review. PLoS Medicine, 7(7), e1000316. 10.1371/journal.pmed.1000316 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hopwood CJ, Harrison AL, Amole M, Girard JM, Wright AGC, Thomas KM, Sadler P, Ansell EB, Chaplin TM, Morey LC, Crowley MJ, Durbin CE, & Kashy DA (2020). Properties of the Continuous Assessment of Interpersonal Dynamics across sex, level of familiarity, and interpersonal conflict. Assessment, 27(1), 40–56. 10.1177/1073191118798916 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kelly JF, Belisario K, & MacKillop J (2025). Prevalence, predictors, correlates, and dynamic changes in the NIAAA-defined ‘recovery’ definition. Alcohol: Clinical & Experimental Research. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kelly JF, Brown SA, Abrantes A, Kahler C, & Myers M (2008). Social Recovery Model: An 8-Year Investigation of Adolescent 12-step Group Involvement following Inpatient Treatment. Alcoholism, Clinical and Experimental Research, 32(8), 1468–1478. 10.1111/j.1530-0277.2008.00712.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kelly JF, & Greene MC (2014). Beyond Motivation: Initial validation of the Commitment to Sobriety Scale. Journal of Substance Abuse Treatment, 46(2), 10.1016/j.jsat.2013.06.010. 10.1016/j.jsat.2013.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kelly JF, Greene MC, Bergman BG, White WL, & Hoeppner BB (2019). How Many Recovery Attempts Does it Take to Successfully Resolve an Alcohol or Drug Problem? Estimates and Correlates From a National Study of Recovering U.S. Adults. Alcoholism: Clinical and Experimental Research, 43(7), 1533–1544. 10.1111/acer.14067 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. 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(2), 289–299. 10.1111/j.1360-0443.2011.03593.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. 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(6), e100121. 10.1371/journal.pone.0100121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kelly JF, Stout RL, Magill M, & Tonigan JS (2011). The role of Alcoholics Anonymous in mobilizing adaptive social network changes: A prospective lagged mediational analysis. Drug and Alcohol Dependence, 114(2–3), 119–126. 10.1016/j.drugalcdep.2010.09.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Leach D, & Kranzler HR (2013). An Interpersonal Model of Addiction Relapse. Addictive Disorders & Their Treatment, 12(4), 183–192. 10.1097/ADT.0b013e31826ac408 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lin Y, Dowell N, Godfrey A, Choi H, & Brooks C (2019). Modeling gender dynamics in intra and interpersonal interactions during online collaborative learning. Proceedings of the 9th International Conference on Learning Analytics & Knowledge, 431–435. 10.1145/3303772.3303837 [DOI] [Google Scholar]
  35. Lopez-Quintero C, Hasin DS, de los Cobos JP, Pines A, Wang S, Grant BF, & Blanco C (2011). Probability and predictors of remission from lifetime nicotine, alcohol, cannabis, or cocaine dependence: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Addiction (Abingdon, England), 106(3), 657–669. 10.1111/j.1360-0443.2010.03194.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Luong G, Charles ST, & Fingerman KL (2011). Better With Age: Social Relationships Across Adulthood. Journal of Social and Personal Relationships, 28(1), 9–23. 10.1177/0265407510391362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Maisto SA, Hallgren KA, Roos CR, & Witkiewitz K (2018). Course of remission from and relapse to heavy drinking following outpatient treatment of alcohol use disorder. Drug and Alcohol Dependence, 187, 319–326. 10.1016/j.drugalcdep.2018.03.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Martinelli TF, van de Mheen D, Best D, Vanderplasschen W, & Nagelhout GE (2021). Are members of mutual aid groups better equipped for addiction recovery? European cross-sectional study into recovery capital, social networks, and commitment to sobriety. Drugs: Education, Prevention & Policy, 28(5), 389–398. [Google Scholar]
  39. McCrady BS (2004). To Have But One True Friend: Implications for Practice of Research on Alcohol Use Disorders and Social Network. Psychology of Addictive Behaviors : Journal of the Society of Psychologists in Addictive Behaviors, 18, 113–121. 10.1037/0893-164X.18.2.113 [DOI] [PubMed] [Google Scholar]
  40. McCrady BS (2012). Treating Alcohol Problems With Couple Therapy. Journal of Clinical Psychology, 68(5), 514–525. 10.1002/jclp.21854 [DOI] [PubMed] [Google Scholar]
  41. McKay JR (2017). Making the hard work of recovery more attractive for those with substance use disorders. Addiction, 112(5), 751–757. 10.1111/add.13502 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. McNeish D, & Wolf MG (2023). Dynamic fit index cutoffs for confirmatory factor analysis models. Psychological Methods, 28(1), 61–88. 10.1037/met0000425 [DOI] [PubMed] [Google Scholar]
  43. Meisel SN, Hennessy EA, Jurinsky J, & Kelly JF (2023). Improving social recovery capital research to enhance clinical utility: A proposed agenda. Addiction Research & Theory, 0(0), 1–7. 10.1080/16066359.2023.2224964 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Meyers RJ, Miller WR, Hill DE, & Tonigan JS (1998). Community reinforcement and family training (CRAFT): Engaging unmotivated drug users in treatment. Journal of Substance Abuse, 10(3), 291–308. 10.1016/S0899-3289(99)00003-6 [DOI] [PubMed] [Google Scholar]
  45. Muthén LK, & Muthén BO (2017). Mplus User’s Guide. Eighth Edition. Los Angeles, CA: Muthén & Muthén. 10.1111/j.1600-0447.2011.01711.x [DOI] [Google Scholar]
  46. Najdzionek P, McIntyre-Wood C, Amlung M, & MacKillop J (2023). Incorporating socioeconomic status into studies on delay discounting and health via subjective financial status: An initial validation in tobacco use. Experimental and Clinical Psychopharmacology, 31(2), 475–481. 10.1037/pha0000628 [DOI] [PubMed] [Google Scholar]
  47. Nguyen L-C, Durazzo TC, Dwyer CL, Rauch AA, Humphreys K, Williams LM, & Padula CB (2020). Predicting Relapse After Alcohol Use Disorder Treatment in a High-Risk Cohort: The Roles of Anhedonia and Smoking. Journal of Psychiatric Research, 126, 1–7. 10.1016/j.jpsychires.2020.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. O’Farrell TJ, & Schein AZ (2011). Behavioral Couples Therapy for Alcoholism and Drug Abuse. Journal of Family Psychotherapy, 22(3), 193–215. 10.1080/08975353.2011.602615 [DOI] [Google Scholar]
  49. Prestigiacomo C, Fisher-Fox L, & Cyders MA (2024). A systematic review of the reasons for quitting and/or reducing alcohol among those who have received alcohol use disorder treatment. Drug and Alcohol Dependence Reports, 13, 100300. 10.1016/j.dadr.2024.100300 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Roessler KK, Mau M, & Ekstrøm CT (2018). Interpersonal problems of alcohol use disorder patients undergoing a physical exercise intervention – a randomised controlled trial. Nordic Psychology, 70(3), 245–255. 10.1080/19012276.2017.1418414 [DOI] [Google Scholar]
  51. SAMHSA. (2022). Key substance use and mental health indicators in the United States: Results from the 2021 National Survey on Drug Use and Health (HHS Publication No. PEP22-07-01-005). https://www.samhsa.gov/data/report/2021-nsduh-annual-national-report
  52. Satorra A, & Bentler PM (2001). A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66(4), 507–514. 10.1007/BF02296192 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Shin H, & Park C (2023). Gender differences in social networks and physical and mental health: Are social relationships more health protective in women than in men? Frontiers in Psychology, 14, 1216032. 10.3389/fpsyg.2023.1216032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Smith JE, Meyers RJ, & Austin JL (2008). Working with Family Members to Engage Treatment-Refusing Drinkers: The CRAFT Program. Alcoholism Treatment Quarterly, 26(1–2), 169–193. 10.1300/J020v26n01_09 [DOI] [Google Scholar]
  55. Stout RL, Kelly JF, Magill M, & Pagano ME (2012). Association between social influences and drinking outcomes across three years. Journal of Studies on Alcohol and Drugs, 73(3), 489–497. 10.15288/jsad.2012.73.489 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Strickland JC, & Acuff SF (2023). Role of social context in addiction etiology and recovery. Pharmacology Biochemistry and Behavior, 229, 173603. 10.1016/j.pbb.2023.173603 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Sturgeon JA, Zautra AJ, & Okun MA (2014). Associations between financial stress and interpersonal events: A daily diary study of middle-aged adults and their life circumstances. Psychology and Aging, 29(4), 803–813. 10.1037/a0037961 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Vila J. (2021). Social Support and Longevity: Meta-Analysis-Based Evidence and Psychobiological Mechanisms. Frontiers in Psychology, 12, 717164. 10.3389/fpsyg.2021.717164 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Vilsaint CL, Kelly JF, Bergman BG, Groshkova T, Best D, & White W (2017). Development and validation of a Brief Assessment of Recovery Capital (BARC-10) for alcohol and drug use disorder. Drug and Alcohol Dependence, 177(February), 71–76. 10.1016/j.drugalcdep.2017.03.022 [DOI] [PubMed] [Google Scholar]
  60. Wright AGC, Pincus AL, & Hopwood CJ (2023). Contemporary integrative interpersonal theory: Integrating structure, dynamics, temporal scale, and levels of analysis. Journal of Psychopathology and Clinical Science, 132(3), 263–276. 10.1037/abn0000741.supp [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental
NIHMS2192136-supplement-Supplemental.docx (53.5KB, docx)

RESOURCES