Examining the longer-term efficacy of brief, alcohol-focused personalized feedback interventions for individuals with internalizing distress: Secondary analysis of a randomized controlled trial

Abstract

Background and aims: Efficacy of brief alcohol interventions for young adults with internalizing distress (i.e. symptoms of depression, anxiety and stress) is unclear. We tested the moderating effect of internalizing distress on the efficacy of alcohol single- and multicomponent personalized feedback interventions (PFIs).

Design: Secondary data were analyzed from a randomized controlled trial (RCT) testing the efficacy of single and multicomponent PFIs, compared with an attention-only control condition.

Setting: Participants were sampled from two West Coast universities in the United States. All study protocols were completed online.

Participants: Participants (n = 1137) were college students (63% female; M_age = 20.1 years; 62.6% non-Hispanic white) who reported on internalizing distress at baseline. Some individuals reported clinically significant symptoms (depression: mild/moderate = 24.0%, severe/extremely severe = 10.5%; anxiety: mild/moderate = 19.6%, severe/extremely severe = 11.4%; and stress: mild/moderate = 37.5%, severe/extremely severe = 6.0%).

Interventions: There were four different alcohol single-component PFIs administered and an attention-only PFI control. Alcohol PFIs varied in their complexity and single-component PFIs (i.e. personalized normative feedback) were compared with multicomponent PFI (i.e. containing additional alcohol-focused psychoeducation).

Measurements: Baseline levels of internalizing distress were measured using the summed total of the Depression, Anxiety and Stress Scales (DASS). Drinking outcomes (alcohol consumption, peak eBAC, alcohol-related consequences) were measured at baseline, 3, 6 and 12 months post-intervention.

Findings: Alcohol PFI (compared with attention-only control) reduced alcohol consumption and related consequences at 6-month [rate ratio (RR)_Consumption = 0.85, P = 0.004] or 12-month follow-ups (RR_Consumption = 0.76, P < 0.001; RR_Consequences = 0.85, P = 0.020), regardless of baseline DASS score. Participants with higher DASS scores (compared with those with lower DASS scores) reported lower 6-month alcohol consumption after receiving a single-component intervention (RR = 0.80, P < 0.001). However, individuals with higher DASS scores (compared to those with lower DASS scores) reported more 6-month alcohol-related consequences after receiving a multicomponent intervention (RR = 0.78, P = 0.004).

Conclusions: Personalized feedback interventions may demonstrate efficacy towards reducing drinking in young adults and appear similarly beneficial across levels of internalizing distress, although lower-complexity interventions may be more efficacious.

Keywords: alcohol; alcohol consumption; alcohol‐related consequences; internalizing distress; personalized feedback interventions; tailoring.

FIGURE 1

Examples of intervention content and personalized normative feedback. (a) Descriptive normative feedback. (b) Injunctive normative feedback. (c) Psychoeducation on alcohol‐focused skills. Images depict examples of intervention content included in the original randomized controlled trial (see Larimer et al. [14]).

FIGURE 2

Marginal effect plots from three‐way interaction shown in Table 1 estimating number of drinks per week. The percent estimates shown in the figures represent the relative change in model‐predicted counts from the baseline value at a given follow‐up. BL, baseline; DASS, Depression, Anxiety and Stress Scale; DDQ, Daily Drinking Questionnaire.

FIGURE 3

Marginal effect plots from three‐way interaction shown in Table 1 estimating number of alcohol‐related consequences. The percent estimates shown in the figures represent the relative change in model‐predicted counts from the baseline value at a given follow‐up. BL, baseline; DASS, Depression, Anxiety and Stress Scale; DDQ, Daily Drinking Questionnaire.

FIGURE 4

Marginal effect plots from three‐way interaction shown in Table 2 estimating number of drinks per week. The percent estimates shown in the figures represent the relative change in model‐predicted counts from the baseline value at a given follow‐up. BL, baseline; DASS, Depression, Anxiety and Stress Scale; DDQ, Daily Drinking Questionnaire.

FIGURE 5

Marginal effect plots from three‐way interaction shown in Table 2 estimating number of alcohol‐related consequences. The percent estimates shown in the figures represent the relative change in model‐predicted counts from the baseline value at a given follow‐up. BL, baseline; DASS, Depression, Anxiety and Stress Scale, DDQ, Daily Drinking Questionnaire.