PREreview of The Association between Autism and Psychosis and the Tools Used to Measure it: An Updated Systematic Review and Meta-Analysis

This systematic review evaluates whether methodological and measurement issues are part of the link between autism and psychosis. Disentangling the two diagnostic constructs is complex, and authors clearly investigate different explanations for the common co-occurrence. The review shows the heterogeneity of the literature and explanations of the association between autism and psychosis, and emphasises that transparent reporting helps evidence synthesis. There is little justification for carrying out this meta-analysis aside from a gap of some years since the last on this topic. More detailed explanations of why this investigation was done would be beneficial.

Major strengths

• This is measurement-focused investigation into transdiagnostic clinical traits which is an understudied area of real clinical significance.

• Risk of bias criteria, study selection criteria, and review process are clearly presented.

Major issues

• Justification for an updated meta-analysis in the context of any changes to the literature.

Minor issues

• No clear delineation between planned and unplanned (sub-/sensitivity) analyses.

• Reporting using the ‘SWiM’ – Synthesis Without Meta-Analysis – guidelines may bolster the claims made about the studies included in this systematic review.

• The authors state in the discussion that the study "provides a thorough attempt of examining the moderating effect of measures used," but were unable to examine measurement as a moderator in analyses 2 and 3 due to limited data. While this limitation is appropriately acknowledged in the results and limitations sections, consider slightly softening the language in the abstract and discussion to reflect that moderation analyses were only possible for the correlational data (analysis 1), to more precisely represent the scope of the analyses conducted, given this was a main aim. 

• While the authors acknowledge considerable heterogeneity across all three analyses, the discussion of potential explanations remains somewhat limited. The authors note that age, sex, and measurement type do not fully explain this heterogeneity. The authors could expand their discussion of other potential moderators – authors may have more insight into potential explanations that would benefit the reader and future research.

• Given that analyses 2 and 3 combined various psychotic conditions (schizophrenia, first episode psychosis, clinical high risk, schizotypal personality disorder), authors also could have explored whether diagnostic category moderates the association, particularly given that clinical high risk represents a different stage than established psychotic disorders.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.