Final Report Summary - GAIT-2-OA (Can biomechanical measurements of joints predict severity of osteoarthritis?)
Gait changes may non-invasively monitor osteoarthritis (OA) progression. We have recently found that OA-prone Str/ort mice show increasing non-compliance in a treadmill task with age and OA development. The aim herein was to evaluate use of asymmetry and imbalance indices, as predictors of treadmill task non-compliance in Str/ort mouse OA development. Twenty-two male Str/ort and 20 CBA mice (non-OA prone) were monitored longitudinally and gait measurements (DigigaitTM system) taken at 4wk intervals from age 8-36wks or until treadmill task non-compliance; a period encompassing pre-OA and overt, active OA phases in Str/ort mice. Spatiotemporal gait measurements including swing, brake, propel and stance times, stride length and frequency, and paw area were each used to calculate: symmetry index (SI), and symmetry ratio (SR) bewteen right and left contralateral hind- or fore-limbs. We also defined gait moderation index (GMI) as an imbalanced measures between contralateral hind-fore limbs (diagonal limbs). Linear mixed effects models were employed to assess difference in gait symmetry between CBA and Str/ort mice. Models included the effects of strain, age and strain/age interaction, as well as random mouse effect. Time-dependent Cox regression models were used to assess effects of SI, SR and GMI on early treadmill task non-compliance in Str/ort aged 16-28wks. All 20 CBA, but only 3/22 Str/ort mice completed treadmill tasks until 36wks of age; 9 Str/ort mice dropped-out between 16-28wks and a further 10 at 32 wks. Results were consistent for both SI and SR for all gait characteristics. Longitudinal pattern of SI differed in hind- and fore-limbs for swing:stance ratio, in hind-limb paw area and stance time in CBA and Str/ort mice. Str/ort had lower fore-paw area SI than CBA mice. SI did not differ in CBA and Str/ort for stride length, frequency, swing, propel or brake time; the majority remained stable in hind- but not in fore-limbs. Longitudinal GMI patterns differed between strains (swing:stance, paw area and propel time) except stride length and frequency. SI for hind-limb swing:stance and GMI for paw area were predictors of treadmill non-compliance in Str/ort mice. Longitudinal differences in symmetry and imbalance patterns between CBA and Str/ort mice indicate different gait modification during normal growth periods and OA development. Assuming treadmill task non-compliance is linked with OA, our data suggest that hind-limb asymmetry is more closely related to OA development, but that fore-limb asymmetry tempers these hind-limb OA effects. Vitally, capacity for swing:stance asymmetry and paw area gait moderation index to predict Str/ort mouse non-compliance in the treadmill task supports their use as early OA markers.