BehaviorMID task performance

We found no effects of oxytocin on MID accuracy for win (t(49) = 0.32, p = 0.75; oxytocin: M = 0.67 SD ± 0.04; placebo: M = 0.67 SD ± 0.03), avoid loss (t(49) = 1.01, p = 0.32; oxytocin: M = 0.70 SD ± 0.01; placebo: M = 0.69 SD ± 0.02), or neutral (t(49) = 0.56, p = 0.57; oxytocin: M = 0.63 SD ± 0.19; placebo: M = 0.61 SD ± 0.18). We also found no effects of oxytocin on MID reaction time for win (t(49) = 0.76, p = 0.45; oxytocin: M = 0.19 SD ± 0.01; placebo: M = 0.20 SD ± 0.01), avoid loss (t(49) = 0.70, p = 0.48; oxytocin: M = 0.13 SD ± 0.02; placebo: M = 0.13 SD ± 0.02), or neutral (t(49) = 0.28, p = 0.78; oxytocin: M = 0.20 SD ± 0.01; placebo: M = 0.20 SD ± 0.01).

SID task performance

We found no effects of oxytocin on SID accuracy for win (t(49) = 0.35, p = 0.73; oxytocin: M = 0.65 SD ± 0.04; placebo: M = 0.65 SD ± 0.04), avoid loss (t(49) = 0.21, p = 0.84; oxytocin: M = 0.65 SD ± 0.03; placebo: M = 0.66 SD ± 0.06), or neutral (t(49) = 0.36, p = 0.72; oxytocin: M = 0.67 SD ± 0.19; placebo: M = 0.65 SD ± 0.18). We also found no effects of oxytocin on SID reaction time for win (t(49) = 0.34, p = 0.73; oxytocin: M = 0.20 SD ± 0.02; placebo: M = 0.20 SD ± 0.01), avoid loss (t(49) = 0.08, p = 0.94; oxytocin: M = 0.17 SD ± 0.01; placebo: M = 0.17 SD ± 0.01), or neutral (t(49) = 0.23, p = 0.82; oxytocin: M = 0.21 SD ± 0.02; placebo: M = 0.20 SD ± 0.01).

Mood states related to monetary incentives

We found no effects of oxytocin related to self-perceived happiness across feedback conditions (F(1, 52) = 0.07, p = 0.79), excited (F(1, 53) = 0.15, p = 0.70), unhappiness (F(1, 53) = 0.15, p = 0.70), fearfulness (F(1, 53) = 0.01, p = 0.95), nor motivation (F(1, 53) = 0.06, p = 0.81) in relation to monetary incentives.

Mood states related to social incentives

We found no effects of oxytocin on self-reported happiness (F(1, 52) = 0.35, p = 0.56), excited (F(1, 52) = 1.37, p = 0.25), unhappiness (F(1, 52) = 0.001, p = 0.97), fearfulness (F(1, 52) = 0.03, p = 0.87), or motivation (F(1, 52) = 0.06, p = 0.80) in relation to social incentives.

Oxytocin effects on affective state (PANAS)

We found no effects of oxytocin on positive or negative affect (F(1, 53) = 2.16, p = 0.15).

Neural responses during the tasks

See Table 1 for complete linear mixed model of the MID and Table 2 for the complete linear mixed model of the SID.

Main effects of time

For both the MID (χ2(3) = 276.74, p < 0.0001) and SID (χ2(3) = 146.56, p < 0.0001) we found an effect of time, with both tasks eliciting the expected peak BOLD responses at 6–8 s in both ROIs (see Figs. 4 and 5).

Fig. 4

Time-series of oxytocin effects across all conditions

Table 1 Linear mixed model of hemodynamic response in ROIs with oxytocin versus placebo: Monetary incentive delay taskTable 2 Linear mixed model of hemodynamic response in ROIs with oxytocin versus placebo: social incentive delay taskResponses vary across condition and brain region

For the MID we found a two-way interaction effect of time X condition, χ2(6) = 37.14, p < 0.0001 and a main effect of condition, χ2(2) = 13.79, p = 0.001. For the SID we found a main effect of condition, χ2(2) = 43.39, p < 0.0001, with both tasks in general eliciting greater responses during avoid loss and win trials, over neutral trials (see Fig. 5). We found a condition X ROI interaction for both the MID (X2(2) = 12.18, p = 0.002) and the SID (χ2(2) = 6.08, p = 0.048). Figure 5 illustrates this variation between ROIs and across condition for each task.

Fig. 5

Response versus condition, Collapsed across oxytocin and placebo

Effects of oxytocin on neural responsesOxytocin does not affect response amplitude or shape when collapsing across condition

We saw no main effects of oxytocin on hemodynamic response in either the MID, X2(1) = 1.84, p = 0.18, or the SID, X2(1) = 0.49, p = 0.48. We also saw no oxytocin X time interactions in either the MID, X2(3) = 5.81, p = 0.12, or the SID, X2(3) = 1.49, p = 0.68, indicating the amplitude and shape of BOLD response remains constant across oxytocin and tasks (see Fig. 4).

Oxytocin modulates condition-specific responses

For both the MID (χ2(2) = 7.53, p = 0.02) and the SID (χ2(2) = 6.64, p = 0.04), we saw an oxytocin X condition interaction, showing that oxytocin modulates condition-specific responses during both monetary and social incentive processing (see Fig. 6). For the MID we also found a three-way interaction effect of oxytocin X condition X ROI, χ2(2) = 7.24, p = 0.03, suggesting an ROI-specific effect of oxytocin only during the MID (see Fig. 6).

MID contrasts with the neutral condition subtracted (Fig. 6) indicated that oxytocin tended to reduce BOLD signal in VTA/SN and increase BOLD signal in NAc. The largest effect of oxytocin was an enhancement of the BOLD undershoot at 12 s in the VTA/SN during the monetary loss condition. With the SID task, contrasts with the neutral condition subtracted (Fig. 6) showed that oxytocin tended to flatten the BOLD response across conditions in both VTA/SN and NAc. The most prominent effect was an oxytocin-induced suppression of the BOLD signal peak at 8 s in the NAc during the social win condition.

Fig. 6

Time-series of oxytocin effects by condition after subtraction of neutral control

Neural-behavioral correlations

For the VTA/SN during the MID, increases in feelings of happiness for monetary reward were correlated with decreases in BOLD response from oxytocin (oxytocin – placebo) at 10–12 s post-onset, r(17) = -0.61, pcorr < 0.007. For the NAc during the SID, increases in feelings of motivation for social reward were correlated with increases in BOLD response from oxytocin (oxytocin – placebo) at 8–10 s post-onset, r(17) = 0.64, pcorr < 0.007.

All other correlations of BOLD response and task-related mood states or PANAS scales across our 6–14 post-onset time window of interest were not statistically significant (pcorr > 0.007). For significant uncorrected neural-behavioral correlates, see Supplemental Table 1.