High-speed imaging of evoked rodent mechanical behaviors yields variable results that are not predictive of inflammatory injury.

Few analgesics identified using preclinical models have successfully translated to clinical use. These translational limitations may be due to the unidimensional nature of behavioral response measures used to assess rodent nociception. Advances in high-speed videography for pain behavior allow for objective quantification of nuanced aspects of evoked paw withdrawal responses. However, whether videography-based assessments of mechanical hypersensitivity outperform traditional measurement reproducibility is unknown. First, we determined whether high-speed videography of paw withdrawal was reproducible across experimenters. Second, we examined whether this method distinguishes behavioral responses exhibited by naive mice and mice with complete Freund's adjuvant (CFA)-induced inflammation. Twelve experimenters stimulated naive C57BL/6 mice with varying mechanical stimuli. Paw withdrawal responses were recorded with high-speed videography and scored offline by one individual. Our group was unable to replicate the original findings produced by high-speed videography analysis. Surprisingly, ∼80% of variation was not accounted for by variables previously reported to distinguish between responses to innocuous and noxious stimuli (paw height, paw velocity, and pain score), or by additional variables (experimenter, time-of-day, and animal), but rather by unidentified factors. Similar high-speed videography assessments were performed in CFA- and vehicle-treated animals, and the cumulative data failed to reveal an effect of CFA injection on withdrawal as measured by high-speed videography. This study does not support using paw height, velocity, or pain score measurements from high-speed recordings to delineate behavioral responses to innocuous and noxious stimuli. Our group encourages the continued use of traditional mechanical withdrawal assessments until additional high-speed withdrawal measures are validated in established pain models.