Does working memory training improve emotion regulation and reduce internalizing symptoms? A pair of three-level meta-analyses

Emotional dysfunction is a core feature of many mental disorders. Working memory training (WM-T) is promising to improve emotion regulation and reduce internalizing symptoms (anxiety and depressive symptoms), but the results are mixed. Therefore, we conducted meta-analyses to clarify these mixed results. We searched Web of Science, PubMed, ScienceDirect, and EBSCO to identify relevant studies and screened the references. The effect size was calculated using Hedges' g. Three-level, random-effects models were run using metafor in R. The current study included 44 articles, of which 29 were involved with emotion regulation, and 30 were involved with internalizing symptoms. The results showed that WM-T could yield emotional benefits, but the benefits were confined to enhancing explicit emotional regulation capacity and reducing anxiety symptoms. For the meta-analysis regarding the effect of WM-T on emotion regulation, there is no significant moderator. For the meta-analysis regarding the effect of WM-T on internalizing symptoms, the emotional valence of the material and control group were statistically significant moderators. WM-T can yield certain emotional effects, but only to improve explicit emotion regulation capacity and reduce anxiety symptoms. In addition, some measures could enhance the effect, such as targeting specific populations, increasing the number of training sessions (>= 15) or duration (> 450 minutes), using negative material, and using n-back training tasks.