Using ecological momentary assessment to examine the relationship between craving and affect with opioid use in a clinical trial of clonidine as an adjunct medication to buprenorphine treatment

Abstract

Background: In a recent clinical trial (NCT00295308), we demonstrated that clonidine decreased the association between opioid craving and moderate levels of stress and affect in patients receiving buprenorphine-based opioid agonist therapy.

Objectives: To examine the relationship between illicit opioid use and craving and affect during the evaluation of clonidine as an adjunct medication in buprenorphine treatment for opioid use disorder. Secondarily, to examine whether those relationships are driven by within- or between-participant factors.

Methods: This was a secondary data analysis from our original trial. Participants (N = 108, female: n = 23, male n = 85) receiving buprenorphine were randomized to receive adjunct clonidine or placebo. Participants used portable electronic devices to rate stress, mood, and craving via ecological momentary assessment (EMA) four times randomly each day. To associate the EMA data with illicit opioid use, each EMA report was linked to participants' next urine drug screen (thrice weekly). We used generalized linear mixed models to examine the interaction between treatment group and illicit opioid use, as well as to decompose the analysis into within- and between-participant effects.

Results: Craving for opioids and cocaine was increased when participants were using illicit opioids; this effect was greater in the clonidine group. For affect, mood was poorer during periods preceding opioid-positive urines than opioid-negative urines for clonidine-treated participants, whereas there was no difference for placebo participants.

Conclusion: This secondary analysis provides evidence that for participants maintained on opioid agonist therapy, clonidine minimized the behavioral impact of moderate levels of negative affect and craving.

Keywords: Ecological momentary assessment; abstinence; clonidine; craving; opioid-agonist therapy; stress.

Figures

Figure 1.

The effect of the treatment condition and illicit opioid use on the intensity of reported craving. The y-axis represents the model-predicted mean of craving (range 0–3). Error bars represent standard error of the mean. The * represents a difference between positive and negative use periods with a p ≤ .05 adjusted for multiple comparisons using Tukey’s method.

Figure 2.

The effect of the treatment condition and illicit opioid use on the intensity of reported mood. The y-axis represents the model-predicted mean of mood (range 0–3). Error bars represent standard error of the mean. The * represents a difference between positive and negative use periods with a p ≤ .05 adjusted for multiple comparisons using Tukey’s method.