Commonly-used rodent tests of anxiety-like behavior lack predictive validity for human sex differences

Abstract

Women are more likely to develop an anxiety disorder than men. Yet, preclinical models of anxiety were largely developed in male rodents, with poorly understood predictive validity for sex differences. Here, we investigate whether commonly-used anxiety-like behavior tests, elevated plus maze (EPM) and open field (OF), represent the human sex difference in adult Sprague-Dawley rats. When interpreted by EPM or OF, female rats displayed less anxiety-like behavior compared to males, as they spent twice as much time in the open arms of the EPM or the center of the OF compared to males. However, they also displayed vastly different levels of locomotor activity, possibly confounding interpretation of these locomotion-dependent tests. To exclude locomotion from the assessment, the acoustic startle response (ASR) test was used. When interpreted by the ASR test, females displayed more anxiety-like behavior compared to males, as indicated by a nearly two-fold higher startle amplitude. The observed sex differences were not driven by gonadal steroids. Overall, all but one of the tests fail to mirror the sex difference in anxiety reported in humans. Our findings suggest that the ASR might be a better fit in modelling female anxiety-like behavior.