A three-way RM ANOVA assessing RTs indicated that the main effects were not significant [emotion: F(1, 19) = 1.20, p = 0.287, \(_^\) = 0.06; behavior: F(1, 19) = 1.88, p = 0.186, \(_^\) = 0.09; TMS condition: F(1.349, 25.640) = 0.79, p = 0.418, \(_^\) = 0.04]. By contrast, there was a significant interaction between emotion and behavior [F(1, 19) = 12.07, p = 0.003, \(_^\) = 0.39]. Post-hoc tests indicated shorter RTs for approaching positive image (1062.3 ± 225.7 ms) than for avoiding [1212.6 ± 360.7 ms, p = 0.003, d = 0.48, 95% CI = (58.7, 241.9)], and shorter RTs for avoiding negative image (1058.4 ± 250.7 ms) than for approaching [1248.8 ± 406.4 ms, p = 0.004, d = 0.56, 95% CI = (69.8, 311.1)]. The RTs of each behavior are shown in Fig. 2a.

Reaction times for each TMS condition sorted by behavior (a) and block (b). Error bars indicate standard errors of means (SEMs). **p < 0.01

The RTs sorted by block are shown in Fig. 2b. The results of a two-way RM ANOVA showed a significant main effect of block [F(1, 19) = 12.07, p = 0.003, \(_^\) = 0.39], with the post-hoc analysis indicating that the RT of the automatic block (1060.4 ± 250.3 ms) was significantly shorter than that of the regulated block [1230.7 ± 380.5 ms; d = 0.53, 95% CI = (67.7, 272.0)]. No other significant main effects or interactions were detected.

The mean TMS intensity needed to generate an approximately 1.0-mV MEP was 49.8% of the maximum stimulator output (range, 38–74%). The MEP amplitude recorded at rest which served as baseline was 1.02 ± 0.34 mV. The MEP amplitudes for three TMS conditions are shown in Table 1 and the MEP ratios (% of baseline) are shown in Fig. 3a. The results of a three-way ANOVA of MEP ratios showed a significant main effect of TMS condition [F(1.424, 27.047) = 34.94, p < 0.001, \(_^\) = 0.65]. Post-hoc analyses indicated that the MEP ratio under the ICF condition (213.6 ± 113.8%) was higher than those for the TS alone condition [150.5 ± 80.8%; p < 0.001, d = 0.64, 95% CI = (33.2, 93.0)] and the SICI condition [111.3 ± 77.7%; p < 0.001, d = 1.05, 95% CI = (61.2, 143.5)]; and that the MEP ratio under the SICI condition was lower than that for the TS alone condition [p = 0.001, d = 0.50, 95% CI = (15.4, 63.0)]. The main effects of emotion and behavior were not significant. The two-way interaction between emotion and behavior was significant [F(1, 19) = 16.4, p = 0.001, \(_^\) = 0.46], and the three-way interaction was also significant [F(1.342, 25.499) = 5.67, p = 0.017, \(_^\) = 0.23]. Post-hoc analyses indicated that under the TS alone condition and the ICF condition, the MEP ratio for the positive approaching behavior was significantly higher than that for the positive avoiding behavior [TS alone condition: p = 0.040, d = 0.24, 95% CI = (1.1, 41.1); ICF condition: p = 0.005, d = 0.39, 95% CI = (16.1, 77.6)]; and the MEP ratio for negative avoiding behavior was significantly higher than that for negative approaching behavior [TS alone condition: p = 0.011, d = 0.25, 95% CI = (5.2, 36.0); ICF condition: p = 0.002, d = 0.51, 95% CI = (24.9, 97.8)]. No significant differences were detected for the SICI condition.

MEP ratio (% of baseline) for each TMS condition sorted by behavior (a) and block (b). Error bars indicate SEM. The dashed line indicates the MEP ratio at baseline (100%). Ratios <100% indicate inhibition, and ratios >100% indicate facilitation. *p < 0.05; **p < 0.01

The MEP ratios (% of baseline) sorted block are shown in Fig. 3b. The results of a two-way RM ANOVA showed significant main effects of TMS condition [F(1.424, 27.048) = 34.94, p < 0.001, \(_^\) = 0.65] and block [F(1, 19) = 16.44, p = 0.001, \(_^\) = 0.46]. The interaction of TMS condition and block was also significant [F(1.342, 25.499) = 5.67, p = 0.017, \(_^\) = 0.23]. Post-hoc analyses indicated that under the TS alone condition and the ICF condition, the MEP ratio of the automatic block was higher than that of the regulated block [TS alone condition: p = 0.006, d = 0.25, 95% CI = (6.6, 35.1); ICF condition: p = 0.002, d = 0.46, 95% CI = (6.6, 39.3)]. No significant difference was found for the SICI condition.

For M1 intracortical circuits, the MEP ratios (% of TS alone) under pair-pulse conditions are shown in Fig. 4a. The results of a three-way ANOVA of MEP ratios revealed a significant main effect of TMS condition [F(1, 19) = 47.93, p < 0.001, \(_^\) = 0.72]. Paired-sample t-tests comparing the different TMS conditions indicated that the MEP ratios (% of TS alone) were significantly higher for the ICF (145.2 ± 32.8%) than for the SICI (72.7 ± 33.5%), confirming the effects of the paired-pulse protocol. No other significant main effects were observed [emotion: F(1, 19) = 1.60, p = 0.222, \(_^\) = 0.08; behavior: F(1, 19) = 0.04, p = 0.835, \(_^\) = 0.002]. There was a significant two-way interaction between emotion and behavior [F(1, 19) = 4.90, p = 0.039, \(_^\) = 0.21], and a significant three-way interaction [F(1, 19) = 7.85, p = 0.011, \(_^\) = 0.29]. Paired-sample t-tests indicated that the ICF effect of positive approaching behavior was significantly stronger than positive avoiding [p = 0.037, d = 0.49, 95% CI = (1.4, 39.2)], and the ICF effect of negative avoiding behavior was significantly stronger than negative approaching [p = 0.027, d = 0.58, 95% CI = (3.2, 48.1)]. No significant difference among the different behaviors was observed for the SICI effect.

MEP ratio (% of TS alone) for each TMS condition sorted by behavior (a) and block (b). Error bars indicate SEM. Dashed lines indicate the MEP ratio for the TS alone condition (100%). Ratios <100% indicate inhibition, and ratios >100% indicate facilitation. *p < 0.05; **p < 0.01

The MEP ratios (% of TS alone) sorted by block are shown in Fig. 4b. The results of a two-way RM ANOVA based on block showed significant main effects of TMS condition [F(1, 19) = 47.93, p < 0.001, \(_^\) = 0.72] and block [F(1, 19) = 4.90, p = 0.039, \(_^\) = 0.21] and a significant interaction between TMS condition and block [F(1, 19) =  7.85, p = 0.011, \(_^\) = 0.29]. Post-hoc analyses indicated that under the ICF condition, the MEP ratio of the automatic block was higher than that for the regulated block [p = 0.008, d = 0.62, 95% CI = (−5.2, 20.9)]. No significant difference was observed under the SICI condition.

Pearson correlation analysis showed a significant negative correlation between RT and ICF for the automatic block (r = −0.45, p = 0.048, Fig. 5a). There was no significant correlation for the regulated block (r = −0.16, p = 0.49, Fig. 5b).

Correlation between RTs and MEP ratios (% of TS alone) during the intracortical facilitation condition for the automatic block (a) and regulated block (b)