A Systematic Review and Meta-Analysis of the Relationship Between Social Dominance Status and Common Behavioral Phenotypes in Male Laboratory Mice

Abstract

Background: Social dominance status (e.g., dominant or subordinate) is often associated with individual differences in behavior and physiology but is largely neglected in experimental designs and statistical analysis plans in biomedical animal research. In fact, the extent to which social dominance status affects common experimental outcomes is virtually unknown. Given the pervasive use of laboratory mice and culminating evidence of issues with reproducibility, understanding the role of social dominance status on common behavioral measures used in research may be of paramount importance.Methods: To determine whether social dominance status—one facet of the social environment—contributes in a systematic way to standard measures of behavior in biomedical science, we conducted a systematic review of the existing literature searching the databases of PubMed, Embase, and Web of Science. Experiments were divided into several domains of behavior: exploration, anxiety, learned helplessness, cognition, social, and sensory behavior. Meta-analyses between experiments were conducted for the open field, elevated plus-maze, and Porsolt forced swim test.Results: Of the 696 publications identified, a total of 55 experiments from 20 published studies met our pre-specified criteria. Study characteristics and reported results were highly heterogeneous across studies. A systematic review and meta-analyses, where possible, with these studies revealed little evidence for systematic phenotypic differences between dominant and subordinate male mice.Conclusion: This finding contradicts the notion that social dominance status impacts behavior in significant ways, although the lack of an observed relationship may be attributable to study heterogeneity concerning strain, group-size, age, housing and husbandry conditions, and dominance assessment method. Therefore, further research considering these secondary sources of variation may be necessary to determine if social dominance generally impacts treatment effects in substantive ways.