Does metacognitive training for psychosis (MCT) improve neurocognitive performance? A systematic review and meta-analysis

Background: Metacognitive training for psychosis (MCT) offers benefits for addressing hallmark deficits/symptoms in schizophrenia spectrum disorders including reductions in cognitive biases and positive/negative symptoms as well as improvements in social cognition and functioning. However, differing results exist regarding the relationship between MCT and neurocognition. A comprehensive understanding of the nature of this relationship would significantly contribute to the existing literature and our understanding of the potential added value of MCT as a cognitive intervention for psychosis.

Methods: Across eleven electronic databases, 1312 sources were identified, and 14 studies examining MCT and neurocognition in psychosis were included in this review. Measures of estimated effect sizes were calculated with Hedge's g, moderator analyses used Cochrane's Q statistic and significance tests to measure group differences according to control conditions.

Results: Twelve studies, 11 randomized controlled trials (RCTs) and 1 non-RCT, were included in the main meta-analyses, consisting of 673 participants (n_MCT = 345, n_control = 328). When comparing MCT against control interventions, non-significant differences in estimated effect sizes were observed across all neurocognitive domains when evaluating pre-post changes (g ≤ 0.1, p > .05). Two additional studies corroborated these results in a narrative review.

Conclusion: These findings suggest that when compared against control conditions, MCT does not pose a statistically meaningful benefit to neurocognitive performance. General practice/learning effects are likely the main contributor that explains improvement in neurocognitive performance, and not a difference of intervention allocation when considering MCT against the included control comparators. These findings help establish the specificity of the effects of MCT.