Combined Attention Bias Modification and Working Memory Training for Reducing Craving in Methamphetamine Use Disorder: A Randomized Controlled Study

Background: Craving is central to relapse risk in methamphetamine use disorder (MUD), yet the affective–cognitive processes related to craving remain unclear. Objectives: This study examined whether negative affect was associated with craving and whether attentional bias and working memory capacity (WMC) were involved in this association, and subsequently evaluated the effectiveness of a combined emotion-cue–based attentional bias modification (ABM) and adaptive working memory training (WMT) intervention.Methods: Fifty individuals with MUD completed measures of negative affect (DASS-21), emotional attentional bias, WMC, and craving, which were analyzed using moderated mediation. Subsequently, 40 participants were randomized to a 4-week combined ABM and WMT intervention or treatment as usual, with effects assessed using linear mixed models.Results: Overall negative affect was not significantly associated with craving. Among its components, stress was positively related to both attentional bias and craving. Moderated mediation analyses revealed that attentional bias statistically mediated the association between stress and craving, and WMC significantly moderated the association between stress and attentional bias. The conditional indirect association between stress and craving via attentional bias was significant at low and mean levels of WMC but non-significant at high levels.Linear mixed model analyses showed significant Group × Time interactions for stress, total negative affect, craving, attentional bias, and WMC, and these effects remained significant after FDR correction. Compared with the control group, the experimental group showed greater reductions in stress, total negative affect, craving, and attentional bias, as well as a greater increase in WMC. The Group × Time interactions for depression and anxiety were not significant.Conclusions: These findings support an affective–cognitive framework of craving and provide evidence for a mechanism-informed digital intervention in MUD.