Challenges in cross-test comparison of anxiety-related outcomes in rodents: A network meta-analysis

Abstract

Anxiety-like complex behavioral and psychological constructs are difficult to evaluate in rodents. Many studies have investigated which techniques are appropriate for measuring anxiety and related physiological parameters. Here, we used network meta-analysis to compare the current methods of assessing anxiety. We performed a comprehensive review and network meta-analysis by searching PubMed, EMBASE, the Cochrane Library, SCOPUS, and Web of Science for studies involving rodents with anxiety-related behaviors undergoing behavioral tests with certain keywords: The common parameters that emerged were total distance traveled, fecal boli count, and rearing behavior. In the 46 studies reviewed, the open-field test (OFT) and elevated plus maze (EPM) test appeared most often (in 45 and 43 studies, respectively), while the light-dark box (LDB) and elevated zero maze tests appeared less frequently (in two studies and one study, respectively Subsequently, the tests were ranked based on their likelihood of being the most effective measure for each outcome. For total distance traveled, the OFT showed a significant disadvantage over the EPM and LDB. For fecal boli, there was a significant difference between the LDB and OFT. There were no variations between tests in terms of rearing. Our findings reinforce the importance of considering each behavioral test's unique characteristics when selecting appropriate measures for anxiety-like behaviors. Researchers should exercise caution when interpreting single-measure outcomes and adopt a holistic approach that integrates multiple test results to achieve reliable and relevant conclusions. Network meta-analysis is a powerful tool for identifying the highlights, complexities, and inconsistencies of anxiety-related behaviors in rodents in preclinical anxiety models.