Table 2.
Fully lagged regression analyses evaluating whether the abstinence-based social network mediates the association between having a sponsor and substance use
| Conditions for mediation | β | B (SE) | p |
|---|---|---|---|
| Condition 1 | |||
| Having a sponsor (3 mo.) predicting substance use (9 mo.)1 | |||
| PDAA | .20 | .13 (.06) | .03 |
| DPDD | −.02 | −.32 (1.60) | .84 |
| Abstinence from alcohol | .76 (.42) | .07 | |
| % days use of most frequently used drug | .04 | .02 (.05) | .64 |
| Condition 2 | |||
| Having a sponsor (3 mo.) predicting abstinence-based social network (6 mo.) controlling for baseline abstinence-based social network | .13 | .10 (.07) | .13 |
| Having a sponsor (3 mo.) predicting abstinence-based social network (6 mo.) not controlling for baseline abstinence-based social network | .19 | .15 (.07) | .04 |
| Condition 3 | |||
| Abstinence-based social network (6 mo.) predicting substance use (9 mo.)2 | |||
| PDAA | −.09 | −.07 (.08) | .37 |
| DPDD | .16 | 3.81 (2.20) | .09 |
| Abstinence from alcohol | −1.06 (.59) | .07 | |
| % days use of most frequently used drug | .02 | .02 (.07) | .82 |
| Condition 4 | |||
| Sponsorship (3 mo.) predicting substance use (9 mo.) controlling for abstinence-based social network (6 mo.)1 | |||
| PDAA | .21 | .14 (.06) | .02 |
| DPDD | −.05 | −.86 (1.62) | .60 |
| Abstinence from alcohol | .94 (.44) | .03 | |
| % days use of most frequently used drug | .04 | .02 (.05) | .68 |
Note. PDAA = Proportion of days abstinent from alcohol. DPDD = Drinks per drinking day.
1
Controlled for baseline substance use measure; PDA, DPDD, abstinence from alcohol, and % days use of most frequently used drug respectively.
2
Controlled for having a sponsor at 3 mo. and baseline substance use measure; PDA, DPDD, abstinence from alcohol, % days use of most frequently used drug respectively.