All means (± standard error) are available for each kinematic variable in Supplementary Table 1, but for reasons of brevity and focussing on our main question of interest (group differences) only group means are presented in Figs. 2 and 3.
Transport phase; MT, PV, DT, FAP, NOAA, NOAF
Participants with a cSCI (m = 1757 ms) produced movements of a longer duration than AMC (m = 961 ms) [F(1,29) = 9.06, p < 0.01, η2 = 0.48] and reached a lower PV [F(1,19) = 5.83, p < 0.05, η2 = 0.31] (cSCI 513 mm/s, AMC 672 mm/s) (Fig. 3a, b). Participants with a cSCI (DT 75.80%, FAP 33.22%) also spent a longer proportion of the movement decelerating [F(1,19) = 7.73, p < 0.05, η2 = 0.29] and in the final adjustment phase [F(1,19) = 5.97, p < 0.05, η2 = 0.31] (Figs. 1, 2, 3c, d) compared to AMC (DT 69.12%, FAP 18.89%) and made more adjustments than AMC (NOAA 1.44 vs 4.16, NOAF 2.27 vs 8.44) in both the approach [F(1,19) = 10.34, p < 0.01, η2 = 0.54] and final adjustment phase [F(1,19) = 4.34, p = 0.05, η2 = 0.19] (Fig. 3e, f).
For PV [F(3,17) = 89.42, p < 0.001, η2 = 0.94], FAP [F(3,17) = 12.28, p < 0.001, η2 = 0.68] and NOAF [F(3,17) = 4.26, p < 0.05, η2 = 0.43], a significant condition by limb interaction emerged. These interactions emerged as in asymmetrical conditions (condition three NF and four FN) the limb reaching to the far object (NP/MI limb in condition three = 686.58 mm/s, P/LI limb in condition four = 670.62 mm/s), reached a higher peak velocity than the limb reaching to the near object (P/LI limb in condition three = 599.54 mm/s, NP/MI limb in condition four = 539.05 mm/s). Additionally, in asymmetrical conditions, the limb reaching to the near object spent a longer proportion of the movement in the final adjustment phase [condition three (NF) P/LI = 31.02% NP/MI = 22.08%, condition four (FN) P/LI = 24.58% NP/MI = 30.26%] and made more adjustments [condition three (NF) P/LI = 5.33 NP/MI = 3.96, condition four (FN) P/LI = 4.47 NP/MI = 6.01] in this phase compared to the limb moving to the far object.
Grasp phase: MGA, tMGA, TrG
There was no significant main effect of group, condition or limb for MGA or tMGA. However, participants with a cSCI (m = 334 ms) produced less coupled transport and grasp phases compared to AMC (m = 114 ms) [F(1,19) = 8.49, p < 0.01, η2 = 0.45].
Interlimb synchrony
At movement onset, there was no significant difference between groups [F(1,19) = 3.14, p > 0.05, η2 = 0.14] despite variance seen in participants with a cSCI (Fig. 4a). At ToPV [F(1,19) = 8.66, p < 0.01, η2 = 0.31] (Fig. 4b) and FAPstart [F(1,18) = 7.50, p < 0.05, η2 = 0.30] (Fig. 4c), participants with a cSCI (PV 76 ms, FAP 198 ms) were less synchronous than AMC (PV 32 ms, FAP 82 ms). However, at the end of the movement, the main effect of group did not reach significance [F(1,20) = 2.14, p > 0.05, η2 = 0.10] (Fig. 4d). The main effect of condition did not reach significance at any phase of the movement and no significant interactions emerged.
EMG
Peak muscle activity in relation to kinematic events occurred closest to the start of the final adjustment phase (FAPstart) for all muscles tested (see Fig. 2), therefore, this was the focus of the subsequent statistical analyses (Table 2). We also calculated the timing of peak muscle activity in relation to all other kinematic events but for reasons of brevity include these data as Supplementary Table 2.
Table 2 Group means for the timing of peak muscle activity (as a percentage of movement time) minus the timing of the start of the final adjustment phase (FAPstart) (as a percentage of movement time) for each bimanual condition collapsed across limb [difference in timing as a percentage of movement time is presented (%)]
There was no significant main effect of group for any of the muscles tested except the triceps brachii [F(1,15) = 10.75, p < 0.05, η2 = 0.42] as peak triceps brachii activity was significantly closer to FAPstart for participants with a cSCI (m = 0.19%) compared to AMC (m = − 14.40%).
For the anterior deltoid, there was a significant main effect of condition [F(3,45) = 3.29, p < 0.05, η2 = 0.18] as peak anterior deltoid activity was closer to FAPstart in condition four (FN = 2.46%) compared to condition one (NN = − 10.52%).
A condition by group interaction emerged for the anterior deltoid [F(3,45) = 3,28, p < 0.05, η2 = 0.18], extensor digitorum superficialis [F(3,45) = 3.12, p < 0.05, η2 = 0.17] and triceps brachii [F(3,45) = 3.13, p < 0.05, η2 = 0.17]. Repeated-measures ANOVAs for each group and each muscle revealed no significant main effect of condition. Independent T tests revealed that for the symmetrical trials (NN and FF) there was no significant difference between groups for any of the muscles. However, for the asymmetrical trials (NF and FN), there was a significant difference between groups. In condition three (NF) peak anterior deltoid, extensor digitorum superficialis and triceps brachii activity were after FAPstart in participants with a cSCI (AD = 5.21%, E = 1.65%, T = 4.93%) but before FAPstart for AMC (AD = − 21.60%, E = − 24.22%, T = − 32.88%). In condition four (FN) peak triceps brachii and anterior deltoid activity occurred before FAPstart for participants with a cSCI (T = 4.74%, AD = 9.47%) but after FAPstart in AMC (T = − 12.27%, AD = − 6.02%).
In terms of agonist–antagonist muscle activity patterns, there was a significant main effect of group [F(1,12) = 14.16, p < 0.05, η2 = 0.54] as the time of peak triceps brachii activity and lowest biceps brachii activity was less coupled in participants with a cSCI (m = 24.59%) than AMC (m = 3.14%).