Experimenter-free pain assessment in mice using a thermal gradient ring and functional linear models.

Introduction: Traditional preclinical laboratory animal tests of pain can be suboptimal for many reasons, such as experimenter bias, interference of motor effects, and a lack of emotional or motivational pain process engagement. These shortcomings may also impede the search for novel pain therapies. Recently, thermal preference or gradient tests have gained popularity, and several apparatuses are either commercially available or can be self-made. The thermal gradient ring (TGR) offers a high-resolution thermal preference assessment and removes spatial confounds that may arise in two-chamber or linear designs. Objectives: The aim of this study was to determine the effectiveness of the TGR as a tool to assess pharmacotherapies and develop an appropriate statistical framework for data analysis. Methods: We examined the effects of 17 treatment conditions, including pronociceptive and antinociceptive compounds, in male and female mice. Behavioral data were analyzed using functional linear models (FLM), enabling hypothesis testing of temperature preference and motor activity at each timepoint of the TGR session. Results: Functional linear models captured the temporal dynamics and magnitudes of treatment effects and identified locomotor side effects that were obscured by traditional session-averaged analyses. Minute-by-minute comparisons revealed distinct temporal progressions of treatment effects and locomotor side effects. Conclusion: Our findings demonstrate that the TGR-FLM combination is a robust framework and a valuable tool for screening novel pain therapies.