A step beyond: age and sex-dependent changes in multidimensional locomotor and gait patterns of voluntary motion of C57BL/6 mice

Age-related decreases in physical activity drive frailty and reduced intrinsic capacity, promoting adverse health outcomes such as falls. Experimental tools for quantifying these metrics in preclinical models have lagged behind those used clinically; single-endpoint activity assessments (e.g., rotarod performance, treadmill activity) of rodents often require invasive handling and/or forced performance cues that may increase stress and variability, especially in aged animals. Here, we evaluated whether automated analysis of spontaneous movement using machine-learning-based software (Palmreader) could identify age-related changes in locomotor activity, gait, paw loading, or posture in male and female C57BL/6 mice across young, adult, and aged timepoints (4.5-, 7-, and 22 months of age). Parameters were quantified using the BlackBox One system during a 7-minute open-field recording. Principal component analysis demonstrated age-related separation of movement phenotypes, with individual parameters demonstrating increased hindlimb duty factor and longer stride lengths with age. Age-related changes in forelimb-to-hindlimb duty-factor imbalance, paw contact, and postural organization were also observed, and several gait alterations with age showed different patterns between males and females. Conventional measures of physical activity (e.g., distance traveled, walking time), in contrast, showed few age-related differences, particularly in females. These findings suggest that aging more consistently alters movement patterns rather than movement magnitude, and that gait and postural features may represent important age-related functional changes not captured by activity measurements alone. Automated assessment of spontaneous movement may therefore represent a useful complement to traditional physical activity testing, with the added benefit of minimal animal handling and short recording times.