Training-Induced Functional Gains following SCI

Abstract

We previously demonstrated that daily, hour-long training sessions significantly improved both locomotor (limb kinematics, gait, and hindlimb flexor-extensor bursting patterns) and nonlocomotor (bladder function and at-level mechanical allodynia) functions following a moderate contusive spinal cord injury. The amount of training needed to achieve this recovery is unknown. Furthermore, whether this recovery is induced primarily by neuronal activity below the lesion or other aspects related to general exercise is unclear. Therefore, the current study objectives were to (1) test the efficacy of 30 minutes of step training for recovery following a clinically relevant contusion injury in male Wistar rats and (2) test the efficacy of training without hindlimb engagement. The results indicate that as little as 30 minutes of step training six days per week enhances overground locomotion in male rats with contusive spinal cord injury but does not alter allodynia or bladder function. Thirty minutes of forelimb-only exercise did not alter locomotion, allodynia, or bladder function, and neither training protocol altered the amount of in-cage activity. Taken together, locomotor improvements were facilitated by hindlimb step training for 30 minutes, but longer durations of training are required to affect nonlocomotor systems.

Figure 1

Weekly open field locomotor scoring of trained, nontrained, and forelimb SCI male rats. Statistically significant locomotor recovery occurred from week 1 to week 2 for all groups. At 2 weeks post-SCI training began. Only the trained group had significant increases from pretraining. Shams not shown, BBB = 21 (∗ versus pretraining; W2; repeated ANOVA with Bonferroni post hoc t-tests: nontrain n = 7; forelimb n = 9; train n = 13).

Figure 2

Percent animals with consistent weight support. Significant differences were found between sham and both the nontrained and forelimb groups (∗) at weeks 1–8, except week 7, as well as between the trained and nontrained groups (#) at weeks 3, 5, 6, and 8. No differences were found between sham and trained animals after week 3 (binomial proportion test: nontrain n = 7; forelimb n = 9; train n = 13).

Figure 3

(a) Kinematic illustration of the hindlimb (iliac crest, greater trochanter, knee, lateral malleolus, and metatarsophalangeal joint) during stepping. (b) The maximum (extension) and minimum (flexion) angles as well as excursion (range of motion) of the hip-ankle-toe and iliac crest-hip-ankle angles were calculated and compared between trained (n = 13), nontrained (n = 7), forelimb (n = 9), and sham (n = 4) (ANOVA with Bonferroni post hoc t-tests). Significant differences were found for ankle extension and hip and ankle excursion (∗ sham versus nontrained and forelimb). Sham was not different from trained.

Figure 4

The regularity index revealed a significant difference between the trained and nontrained groups (∗) (Mann-Whitney U test). The plantar stepping index revealed a significant difference between the sham and both the nontrained and forelimb groups (∗) (ANOVA with Bonferroni post hoc t-tests). Parameters of gait were not significant: stride length, stride time, base of support, and toe velocity. Nontrain n = 5; train n = 13; forelimb n = 6; sham n = 4.

Figure 5

No animals were sensitive to touch or gentle squeeze prior to injury. Immediately after SCI, all rats exhibited a moderate degree of evoked at-level allodynia. There were no differences between groups regarding the course of at-level allodynia.

Figure 6

Home cage activity: in-cage activity assessments with an infrared activity monitor show no differences in the amount of ambulatory movements up to week 8 (W8), suggesting that the increased locomotor recovery of the trained group was a result of the step training paradigm and not due to “self-training.” All animals significantly decreased activity with time from surgery (∗ preoperation versus W8 for each group; repeated ANOVA with Bonferroni post hoc t-tests; nontrain n = 7; train n = 13; forelimb n = 9; sham n = 4).

Figure 7

Representative spinal cord segments. Histological assessment did not reveal any differences between groups when analyzing total white or gray matter from the epicenter to 1.0 mm rostrally or when further subdividing the sections into areas of ventral and ventral lateral funiculi (ANOVA: nontrain n = 7; train n = 13; forelimb n = 9).