29-30 January, 2020 - Szeged, Hungary


Abstract details

Trait or state anxiety: refined sampling methods of emotion-related behavior in rodent models


Zoltán K Varga1, Diána Pejtsik1, Zoltán Balogh1, Manó Aliczki1, Máté Tóth1, Éva Mikics1

1 Laboratory of Translational Behavioral Neuroscience, Institute of Experimental Medicine, Budapest

Mental disorders, particularly those associated with anxiety represent a serious burden on both the individual and the society. Preclinical animal models are the predominant approaches to devise therapy, however, available tests more successfully describe populational features than traits of specific subjects, as indicated by their low inter-test correlations and intra-test predictions at the individual level. Thus, current indirect approaches to measure trait anxiety may be strongly biased by the permanently changing internal states of individuals. Here, we present a novel approach, based on serial sampling of behavior, to reduce state-dependent variance and to refine the characterization of individual traits. We conducted a four-time sampling series, applying the three most widely used anxiety tests for rats, the elevated plus-maze, light-dark box and open-field, in a semi-random design, under neutral and highly aversive conditions and constructed summary measures (SuM) from the spatio-temporal and ethological readouts of the tests. Principal component analysis revealed that SuMs, such as averages and slopes of avoidance behavior, explain greater proportion of the total variance of the data than single measures (SiM). In addition, SuMs, contrary to SiMs, show consistently higher, significant inter-test correlation and stronger predictive potential to the behavioral output of same type, as well as novel anxiety tests under aversive conditions. We suggest that the appearance of such associations between SuMs of different test types and conditions support the idea of an emotional trait as a common drive of behavior in rodent anxiety tests that is measurable by the refined sampling method presented here.