Identifying responders to gabapentin for the treatment of alcohol use disorder - an exploratory machine learning approach

Aug 4, 2026

Upload by Rowan

Background: Gabapentin has been proposed for treating alcohol use disorder (AUD), yet a multisite trial found no significant difference between gabapentin enacarbil extended-release (GE-XR; 600 mg twice daily) and placebo on the primary outcome of no heavy drinking days. Despite these null results, identifying who benefits from GE-XR is of strong interest. Methods: Using baseline moderators from the multisite trial (N=338) and qualitative interaction tree machine learning (QUINT), the study searched for predictors of responders and iatrogenic responders. Results: Key associated factors included baseline drinking levels, motivation for change, self-efficacy, cognitive impulsivity, and baseline anxiety. Conclusion: Baseline drinking and anxiety may relate to protracted withdrawal mechanisms, while motivation and self-efficacy emerge as clinically relevant predictors needing further study.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Abstract

Introduction

Background on gabapentin in AUD

Prior trials and withdrawal symptoms

Rationale for responder identification

Methods

QUINT approach and moderators

Results

Predictors of efficacy and iatrogenic response

Conclusion

Clinical implications for motivation and self-efficacy

UCLA UCLA Previously Published Works Title Identifying responders to gabapentin for the treatment of alcohol use disorder: an exploratory machine learning approach. Permalink [https://escholarship.org/uc/item/3jw3m6zk](https://escholarship.org/uc/item/3jw3m6zk) Journal Alcohol and Alcoholism, 60(3) Authors Ray, Lara Grodin, Erica Baskerville, Wave-Ananda et al. Publication Date 2025-03-25 DOI 10.1093/alcalc/agaf010 Copyright Information This work is made available under the terms of a Creative Commons Attribution License, available at [https://creativecommons.org/licenses/by/4.0/](https://creativecommons.org/licenses/by/4.0/) Peer reviewed [eScholarship.org](eScholarship.org) Powered by the California Digital Library University of California Identifying responders to gabapentin for the treatment of alcohol use disorder: an exploratory machine learning approach Lara A. Ray 1 ,2 ,3 , * , Erica N. Grodin 1 ,2 ,3 , Wave-Ananda Baskerville1 , Suzanna Donato1 , Alondra Cruz1 , Amanda K. Montoya1 1 Department of Psychology, University of California, Los Angeles, 1285 Franz Hall, Los Angeles, CA 90095, United States 2 Brain Research Institute, University of California, Los Angeles, 695 Charles E Young Drive S, Los Angeles, CA 90095, United States 3 Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, 760 Westwood Plaza, Los Angeles, CA 90095, United States *Corresponding author. Department of Psychology, University of California, Los Angeles, 1285 Franz Hall, Box 951563, Los Angeles, CA 90095-1563, United States. E-mail: [lararay@psych.ucla.edu](lararay@psych.ucla.edu) Abstract Background: Gabapentin, an anticonvulsant medication, has been proposed as a treatment for alcohol use disorder (AUD) . A multisite study tested gabapentin enacarbil extended-release (GE-XR; 600 mg/twice a day), a prodrug formulation, combined with a computerized behavioral intervention, for AUD. In this multisite trial, the gabapentin GE-XR group did not differ signiﬁcantly from placebo on the primary outcome of percent of subjects with no heavy drinking days. Despite the null ﬁndings, there is considerable interest in using machine learning methods to identify responders to GE-XR. The present study applies interaction tree machine learning methods to identify positive and iatrogenic (i.e. individuals who responded better to placebo than to GE-XR) treatment responders in the trial. Methods: Baseline characteristics taken from the multisite trial were examined as potential moderators of treatment response using qualitative interaction trees (QUINT; N = 338; 223 M/115F) . QUINT models are an exploratory decision tree approach that iteratively splits the data into leaves based on predictor variables to maximize a speciﬁc criterion. Results: Analyses identiﬁed key factors that are associated with the efﬁcacy (or iatrogenic effects) of GE-XR for AUD. Such factors are baseline drinking levels, motivation for change, conﬁdence in their ability to reach drinking goals (i.e. self-efﬁcacy), cognitive impulsivity, and baseline anxiety levels. Conclusion: Baseline drinking levels and anxiety levels may be associated with the protracted withdrawal syndrome, previously implicated in the clinical response to gabapentin. However, these analyses underscore motivation for change and self-efﬁcacy as predictors of clinical response to GE-XR, suggesting these established constructs should receive further attention in gabapentin research and clinical practice. Multiple studies using different machine learning methods are valuable as these novel analytic tools are applied to medication development for AUD. Introduction Gabapentin is an anticonvulsant medication approved by the United States Food and Drug Administration (FDA) for the treatment of partial epileptic seizures and for postherpetic neuralgia. Gabapentin’s mechanism of action is as a calcium channel–aminobutyric acid modu

FAQs

Why was responder identification pursued despite the trial’s null primary outcome?

The multisite trial showed no significant advantage of GE-XR over placebo on the primary endpoint. Interest remained in using machine learning to find subgroups who respond better or are harmed by GE-XR relative to placebo.

What machine learning method was used to identify treatment responders?

The study used qualitative interaction trees (QUINT), an exploratory decision-tree approach that splits data iteratively on predictor variables to maximize a chosen criterion.

Which baseline factors were associated with GE-XR efficacy or iatrogenic effects?

Key factors included baseline drinking levels, motivation for change, self-efficacy, cognitive impulsivity, and baseline anxiety levels, linking them to efficacy and iatrogenic outcomes.

Aug 4, 2026

Upload by Rowan