A meta-analytic investigation of the relationship between attentional bias and subjective craving in substance use: A preregistered update of evidence

Background: Attentional bias towards alcohol and substance-related stimuli is proposed to be associated with craving, and an influential meta-analysis previously supported this relationship (r = 0.19, [95% CI: 0.15 to 0.23]). The aim of the current meta-analysis was to update this evidence with contemporary literature. Design: A preregistered systematic review of literature, replicating the search terms and analysis strategy of Field et al.’s (2009) meta-analysis on attentional bias and craving in individuals who use alcohol / substances. Method: Following a literature search and data extraction informed by PRISMA, a random-effects meta-analysis was conducted to update the previous meta-analysis. A total of 140 effects (from 110 studies) were included in the analyses. Results: The updated pooled association between attentional bias and subjective craving was r = 0.15 ([95% CI: 0.12 to 0.19], I2 = 39%, Tau2 = 0.016). Moderate heterogeneity led to wide prediction intervals (95% PI: -0.10 to 0.38). Corrections for small study / publication bias including Trim and Fill, and PET/PEESE reduced the pooled estimate (ranging from r = 0.08 to 0.14), however it remained statistically significant each time. Subgroup analysis confirmed previous observations that associations were greater for direct measures of attentional bias (vs indirect), and in samples experiencing higher craving. The overall magnitude of the association has reduced over time, in line with a decline effect. Conclusions: Updating an influential meta-analysis to include 15 years of new literature demonstrated that the association between attentional bias and subjective craving in individuals who use alcohol/substances is small but robust, supporting psychological models of addiction and the continued focus on attentional bias as a key mechanistic process.