The hand movement reaction time and movement time (Fig. 1.B) and surface EMG activities of 10 muscles (Fig. 2) were recorded concomitantly during voluntary pointing movements. These movements were initiated from a standing posture, towards a close or a distant visual target (Fig. 1.A) while a picture was presented along the horizontal component of the point movement. Six conditions were tested: 2 Target Distance (close/distant) ⋅ 3 Affective Picture Content (pleasant/neutral/unpleasant). For each participant, the scores of the computerized 9-point version of the Self-Assessment Manikin (SAM) scale were used to control and attest the affective dimension by gender. All numerical results are expressed as the mean ± standard deviation (SD); all graphical results are presented as the mean ± standard error (SE).
Hand movement time is longer in an unpleasant affective context than in a pleasant one.
Reaction time values (Fig. 2.A) did not follow a normal distribution even after log-transformation (Shapiro-Wilk W-tests; P < 0.01). Friedman chi-squared tests were conducted and did not reveal any significant Target Distance effect (χ2 = 3.5, p = 0.36) or Affective Picture Content effect (χ2 = 2.3, p = 0.10).
Hand movement time values (Fig. 2.B) followed a normal distribution (Shapiro-Wilk tests, p > 0.46) and showed equality of sphericity (Mauchly’s tests, p > 0.37). Two-way ANOVA with repeated measurements were conducted and revealed a significant Target Distance effect (F(1,15) = 167.70, p < 0.001; η2 = 0.91), a significant Affective Picture Content effect (F(2,30) = 4.12, p < 0.05; η2 = 0.30), but no interaction effect (F(2,30) = 0.14, p = 0.87). Tukey-HSD post-hoc analysis revealed that hand movement time was significantly longer in an unpleasant context than in a pleasant one (0.35 ± 0.10 s vs. 0.33 ± 0.10 s, p < 0.01). Hand movement time in the neutral context (0.34 ± 0.11 s) was not significantly different from that in a pleasant or an unpleasant context (p = 0.62 and p = 0.63, respectively).
IMC in postural muscle pairs is lower in an unpleasant context than in neutral and pleasant ones.
IMC values for those 24 relevant muscle pairs, which include at least tibialis anterior (TA), the anterior section of the deltoïdus (DA), and the biceps brachii (BB), were computed using the method described by Bigot et al. [43] and Charissou et al. [44, 45]. These IMC values were computed in pleasant, neutral, and unpleasant affective contexts from unrectified muscular activations responsible for the initiation of the voluntary pointing movements (Fig. 3.A).
These 24 muscle pairs were classified into 3 groups. Firstly, the “Posturo-Postural” group (PP) comprising muscle pairs formed only with muscles considered as mainly postural: TA-rectus femoris (RF), TA-soleus (SO), TA-biceps femoris (BF), TA-erector spinae (ES), TA- anterior portion of the rectus abdominis (RA). Secondly, a “Foco-Focal” (FF) group comprising muscle pairs formed only with muscles considered as mainly focal in the present motor task: DA-the posterior section of the deltoïdus (DP), DA-BB, DA-triceps brachii (TP), BB-DP, BB-TB. Thirdly, a “Posturo-Focal” (PF) group with the muscle pairs formed by a mainly postural muscle and a focal muscle TA-DA, TA-DP, TA-BB, TA-TB, DA-SO, DA-RF, DA-BF, DA-RA, DA-ES, BB-SO, BB-RF, BB-BF, BB-RA, BB-ES.
IMC values between these 24 possible muscle pairs were computed during the initiation of the hand pointing movement, more precisely during the [t0-80: t0] period and the [t0:t0 + 80] period (see Fig. 2 and Fig. 3). The first period lying in the anticipatory postural adjustment, and the second corresponded to the very beginning of the movement initiation without, or with very few, online control of the hand. For each participant and in each experimental condition, intermuscular coherence (IMC) across 15 performed movements was calculated as previously described in detail [44, 45]. For each muscle pair, time-frequency IMC was quantified as the volume under the magnitude-coherence curve in both time windows of interest, where the correlation between the EMG time series was detected as significant on the wavelet cross-spectrum [43, 44, 45]. For each time period of interest, the IMC value of each group of muscle pairs (i.e., PP, FF and PF) was the mean of the volumes obtained across the appropriate muscle pairs. IMC values were computed in both the “beta” (β, [15–35] Hz) and the “gamma” (γ, [35–60] Hz) frequency bands (βIMC and γIMC, respectively).
βIMC and γIMC values followed a normal distribution (Shapiro-Wilk tests, p > 0.33) and showed equality of sphericity (Mauchly’s tests, p > 0.21). Three-way ANOVA with repeated measurements were conducted to test the effects of i) Target Distance (close/distant), ii) Affective Picture Content (pleasant/neutral/unpleasant) and iii) Time-Frequency Area (βIMC on [t0-80:t0]/βIMC on [t0:t0 + 80]/γIMC on [t0-80:t0]/γIMC on [t0:t0 + 80]), separately for each group of muscle pairs. Tukey-HSD post-hoc were applied when necessary.
In the PP group of muscle pairs (Fig. 4), the results showed a significant effect of Affective Picture Content (F(2,30) = 4.21, p < 0.05, η2 = 0.40), but no significant Target distance effect (F(1,15) = 0.12, p = 0.54, no Time-Frequency Area effect (F(3,45) = 2.49, p = 0.18) or interaction (F(6,90) = 0.70, p = 0.90). Tukey-HSD post-hoc analysis revealed that the mean IMC in muscle pairs of the PP group significantly decreased in the unpleasant context (0.12 ± 0.06 a.u.), compared to both neutral (0.15 ± 0.06 a.u.; t(127) = 3.29, p < 0.01, η2 = 0.23) and pleasant (0.148 ± 0.062 a.u.; t(127) = 4.06, p < 0.001, η2 = 0.33) contexts. No significant difference was found on IMC between the neutral and pleasant conditions (t(127) = 1.16, p = 0.12 ). No significant effect was highlighted on IMC in both PF (in average: F = 0.74, p = 0.84) and FF (in average: F = 0.11, p = 0.54) muscle pair groups.
Movement time and IMC in posturo-focal muscle pairs are negatively associated in an unpleasant context.
Finally, the relationship between hand movement parameters and IMC values averaged over the two target distances in the groups of muscle pairs was studied in each of the three emotional contexts involved (see Fig. 5 for βIMC). These relationships highlighted significant negative correlations between the hand movement time and IMC in the PF group for both target distances in the unpleasant context only: a significant linear association was found between hand movement time and βIMC during the [t0-80:t0] period (DoF = 30; R2 = 0.52, p < 0.01), and between hand movement time and both βIMC and γIMC during the [t0:t0 + 80] period (DoF = 30; R2 = 0.53, p < 0.01 and R2 = 0.36, p < 0.05, respectively in the γ band). No other significant effect was highlighted in the neutral (DoF = 30; [max R2 = 0.1: min R2 = 0.001], p = 0.98 in average) or pleasant (DoF = 30; [max R2 = 0.15: min R2 = 0.0001], p = 0.86 in average) context.