Effects of oxytocin on behavioral and ERP measures of recognition memory for own-race and other-race faces in women and men

Abstract

Oxytocin has been shown to affect human social information processing including recognition memory for faces. Here we investigated the neural processes underlying the effect of oxytocin on memorizing own-race and other-race faces in men and women. In a placebo-controlled, double-blind, between-subject study, participants received either oxytocin or placebo before studying own-race and other-race faces. We recorded event-related potentials (ERPs) during both the study and recognition phase to investigate neural correlates of oxytocin's effect on memory encoding, memory retrieval, and perception. Oxytocin increased the accuracy of familiarity judgments in the recognition test. Neural correlates for this effect were found in ERPs related to memory encoding and retrieval but not perception. In contrast to its facilitating effects on familiarity, oxytocin impaired recollection judgments, but in men only. Oxytocin did not differentially affect own-race and other-race faces. This study shows that oxytocin influences memory, but not perceptual processes, in a face recognition task and is the first to reveal sex differences in the effect of oxytocin on face memory. Contrary to recent findings in oxytocin and moral decision making, oxytocin did not preferentially improve memory for own-race faces.

Keywords: ERP; Other-race faces; Own-race faces; Oxytocin; Recognition memory; Remember/know; Sex differences.

Figure 1

Geodesic sensor net layout. Electrode sites are numbered. Highlighted clusters are regions of interest included in analyses. L = left, R = right, F = frontal, A = anterior, C = central, P = parietal, M = medial, S = superior, I = inferior.

Figure 2

Most important behavioral effects of oxytocin. Top: Main effects of oxytocin show increased IRK familiarity hits minus false alarms and reduced high-confidence false alarms for oxytocin as compared to placebo. Data is averaged across the factors subject sex, stimulus race, and stimulus facial expression. Bottom: Sex × drug interactions show that oxytocin significantly reduced recollection hits minus false alarms and recollection hits for men but not for women. Data is averaged across the factors stimulus race and stimulus facial expression.

Figure 3

Effects of oxytocin on ERPs from the study phase. A) Memory encoding is shown as mean amplitudes at frontal, central, and parietal regions for subsequently “recollected,” subsequently “familiar,” and subsequently forgotten own-race and other-race faces in the oxytocin and placebo group. For these three regions ERPs are averaged across medial and left and right superior regions of interest. Vertical lines indicate the time window between 750-1000 ms where effects of oxytocin were observed. Here, only oxytocin caused a significant memory encoding effect between subsequently “familiar” and subsequently forgotten faces. Over parietal areas, this effect was only found for own-race faces (see also panel B for topographical maps). Perceptual ERPs are shown at LPI and RPI. No effects of oxytocin were observed for perceptual ERPs. B) Topographical maps of the ERP difference of subsequently “familiar” minus subsequently forgotten items between 750-1000 ms showing significant Dms for own-race and other-race faces over frontal and central areas, but a significant Dm over parietal areas only for own-race faces. C) Scatter plot (regression equation: Y = −0.29 * X + 0.15, R2 = .084, p = .037) depicting the negative correlation between high-confidence false alarms, which are controlled for differences in general memory, and the familiarity-related encoding effect (ERP difference wave of subsequently “familiar” minus subsequently forgotten items) averaged across own-race and other-race faces and all regions of interest. D) Scatter plot (regression equation: Y = 0.35 * X + 0.24, R2 = .126, p = .010) depicting the positive correlation between familiarity judgments measured as hits minus false alarms, which are controlled for differences in general memory, and the familiarity-related encoding effect (ERP difference wave of subsequently “familiar” minus subsequently forgotten items) for own-race faces averaged across all parietal regions of interest.

Figure 4

Effects of oxytocin on ERPs from the recognition test phase. Shown are mean amplitudes for “recollected” old, “familiar” old, and correctly rejected new own-race and other-race faces in the oxytocin and placebo group. Vertical lines indicate the time windows for statistical analyses. Only oxytocin led to significant recollection- and familiarity-related old/new effects between 300 and 500 ms. It also caused generally higher mean amplitudes starting at 500 ms (best seen at CM). Perceptual ERPs are shown at LPI and RPI. No effects of oxytocin were observed for perceptual ERPs.