Gait Assessment of Pain and Analgesics: Comparison of the DigiGait™ and CatWalk™ Gait Imaging Systems

Abstract

Investigation of pain requires measurements of nociceptive sensitivity and other pain-related behaviors. Recent studies have indicated the superiority of gait analysis over traditional evaluations (e.g., skin sensitivity and sciatic function index [SFI]) in detecting subtle improvements and deteriorations in animal models. Here, pain-related gait parameters, whose criteria include (1) alteration in pain models, (2) correlation with nociceptive threshold, and (3) normalization by analgesics, were identified in representative models of neuropathic pain (spared nerve injury: coordination data) and inflammatory pain (intraplantar complete Freund's adjuvant: both coordination and intensity data) in the DigiGait™ and CatWalk™ systems. DigiGait™ had advantages in fixed speed (controlled by treadmill) and dynamic SFI, while CatWalk™ excelled in intrinsic velocity, intensity data, and high-quality 3D images. Insights into the applicability of each system may provide guidance for selecting the appropriate gait imaging system for different animal models and optimization for future pain research.

Keywords: Analgesic; CatWalk™; DigiGait™; Gait analysis; Inflammatory pain; Neuropathic pain.

Fig. 1

Schematic diagrams of DigiGait^™ (A) and CatWalk^™ (B) gait systems. The gait apparatus is outlined above and a representative image is shown below.

Fig. 2

Mechanical allodynia and representative gait images in the SNI experiment. A Schematic diagrams of the procedures. B Time-course of punctate allodynia of the ipsilateral hindpaw after saline or PGB injection in SNI rats (group effect: non-significant, two-way ANOVA; post-test, ***P < 0.001, Sidak’s multiple comparisons test). C Time-course of dynamic allodynia of the ipsilateral hindpaw after saline or PGB injection in SNI rats (group effect: non-significant, two-way ANOVA; post-test: ***P < 0.001, Sidak’s multiple comparisons test). D Consecutive graphs within a duty cycle of the left hindpaw (LH) and right hindpaw (RH) after saline or PGB injection. E Three-dimensional graphs of LH and RH prints boxed by the dots in D [x-axis: length, 0.926 mm/unit; y-axis: width, 0.926 mm/unit; z-axis: intensity arbitrary units (A.U.)].

Fig. 3

Coordination data are pain-related gait parameters in the SNI model. A Correlation analysis between mechanical threshold (x-axis) and DigiGait-swing (left y-axis) and DigiGait-duty cycle (right y-axis). Linear regression, *P < 0.05. B Correlation analysis between mechanical threshold (x-axis) and CatWalk-swing (left y-axis) and CatWalk-duty cycle (right y-axis). Linear regression, *P < 0.05. C, D DigiGait-swing (C) and DigiGait-duty cycle (D) in SNI rats after saline or PGB injection. Two-way ANOVA, treatment effect: **P < 0.01, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, *P < 0.05, shown below the data points. E Correlation between DigiGait-swing and DigiGait-duty cycle. Linear regression, ***P < 0.001; post-test to compare the lines: non-significant. F, G CatWalk-swing (F) and CatWalk-duty cycle (G) in SNI rats after saline or PGB injection. Two-way ANOVA, treatment effect: ***P < 0.001, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, *P < 0.05, ***P < 0.001, shown below the data points. H Correlation between CatWalk-swing and CatWalk-duty cycle. Linear regression, ***P < 0.001; post-test to compare the lines: non-significant.

Fig. 4

Area and intensity data failed to satisfy the criteria for pain-related gait parameters in the SNI model. A Correlation analysis between mechanical threshold (x-axis) and DigiGait-projected area (y-axis). Linear regression. B Correlation analysis between mechanical threshold (x-axis) and CatWalk-max contact area (left y-axis) and CatWalk-print area (right y-axis). Linear regression. C–E DigiGait-projected area (C), CatWalk-max contact area (D) and CatWalk-print area (E) in SNI rats after saline or PGB injection. Two-way ANOVA, treatment effect: non-significant, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, non-significant, shown below the data points. F Correlation between CatWalk-max contact area and CatWalk-print area. Linear regression, ***P < 0.001; post-test to compare the lines: **P < 0.01. G Correlation analysis between mechanical threshold (x-axis) and CatWalk-mean intensity (y-axis). Linear regression. H CatWalk-mean intensity in SNI rats after saline or PGB injection. Two-way ANOVA, treatment effect: non-significant, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, non-significant, shown below the data points.

Fig. 5

Mechanical allodynia and representative gait images in the CFA model. A Schematic diagrams of the procedures. B Time course of punctate allodynia of the ipsilateral hindpaw after saline or tramadol injection in CFA rats. Two-way ANOVA, group effect: non-significant; post-test: Sidak’s multiple comparisons test, ***P < 0.001. C Time course of the dynamic allodynia of the ipsilateral hindpaw after saline or tramadol injection. Two-way ANOVA, group effect: non-significant; post-test: Sidak’s multiple comparisons test, ***P < 0.001. D Consecutive graphs within a duty cycle of the left hindpaw (LH) and right hindpaw (RH) after saline or tramadol injection. E Three-dimensional graphs of LH and RH print boxed by the dotted boxes in D. x-axis: length (0.926 mm per unit); y-axis: width (0.926 mm per unit); z-axis: intensity, arbitrary units (A.U.).

Fig. 6

Coordination data are pain-related gait parameters in the CFA model. A Correlation analysis between mechanical threshold (x-axis) and DigiGait-swing (left y-axis) and DigiGait-duty cycle (right y-axis). Linear regression, **P < 0.01. B Correlation analysis between mechanical threshold (x-axis) and CatWalk-swing (left y-axis) and CatWalk-duty cycle (right y-axis). Linear regression, **P < 0.01. C, D DigiGait-swing (C) and DigiGait-duty cycle (D) in CFA rats after saline or tramadol injection. Two-way ANOVA, treatment effect: **P < 0.01, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, *P < 0.05, shown below the data points. E Correlation between DigiGait-swing and DigiGait-duty cycle. Linear regression, ***P < 0.001; post-test to compare the lines: non-significant. F, G CatWalk-swing (F) and CatWalk-duty cycle (G) in CFA rats after saline or tramadol injection. Two-way ANOVA, treatment effect: ***P < 0.001, *P < 0.05, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, *P < 0.05, shown below the data points. H Correlation between CatWalk-swing and CatWalk-duty cycle. Linear regression, ***P < 0.001, post-test to compare the lines: non-significant.

Fig. 7

Intensity data, but not area data satisfy the criteria for pain-related gait parameters in the CFA model. A Correlation analysis between mechanical threshold (x-axis) and DigiGait-projected area (y-axis). Linear regression. B Correlation analysis between mechanical threshold (x-axis) and CatWalk-max contact area (left y-axis) and CatWalk-print area (right y-axis). Linear regression. C–E DigiGait-projected area (C), CatWalk-max contact area (D) and CatWalk-print area (E) in CFA rats after saline or PGB injection. Two-way ANOVA, treatment effect: non-significant, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, non-significant, shown below the data points. F Correlation between CatWalk-max contact area and CatWalk-print area. Linear regression, ***P < 0.001; post-test to compare the lines: **P < 0.01. G Correlation analysis between mechanical threshold (x-axis) and CatWalk-mean intensity (y-axis). Linear regression, *P < 0.05. H CatWalk-mean intensity in CFA rats after saline or tramadol injection. Two-way ANOVA, treatment effect: ***P < 0.001, shown in the upper right corner; post-test: Sidak’s multiple comparisons test, **P < 0.01, ***P < 0.001, shown below the data points.