It is believed that face processing involves the activation of the face fusiform area [1] (FFA) mediating the analysis of face structure, the superior temporal sulcus mediating the analysis of eye gaze and facial expressions [2] and the inferior occipital gyrus, being particularly responsive to face parts [3]. Some researchers suggest that face processing predominantly activates the right hemisphere [1], whereas others hold that face processing involves both left and right FFA [4] although each may contribute in a different way [5]. It was hypothesized that left FFA is involved in unfamiliar face coding, whereas right FFA would recognize familiar faces [6]. In another study [7] it was found that FFA lateralization depended on handedness: FFA activation was right lateralized in right-handers but not in left-handers (24/32 participants were females). Rhodes et al., [8] found that both left and right FFAs were activated more strongly to faces than to objects while the total volume of activation was significantly larger in the right than the left hemisphere. However, in a recent fMRI study performed only in women [9] they found a clearly bilateral activation of FG in emotional face processing. Overall, the literature is highly inconsistent about FFA hemispheric lateralization [10] [11], [12] since virtually no neuroimaging study has ever considered viewer's sex.

Electromagnetic recordings have identified a posterior negative response peaking at about 170 ms (N170) that is larger to faces than other visual objects over the right hemisphere, and thought to reflect processes involved in the structural encoding of faces. The combination of electromagnetic and neuroimaging data identified the N1 generator in the ventral occipito-temporal cortex [13]–[16], suggesting that N1 might be the manifestation of FFA activity [17]. A closer examination of the literature shows that face-specific N170 topographic distribution is often but not always right-sided in right-handed individuals. It is of great interest that N170 response was found to be bilateral or even left-sided in studies involving a sample in which women were the majority [18] [19] [20] [21] [22]. The study's aim was to investigate whether there are sex-related hemispheric asymmetries for face processing. Face-sensitive N1 responses were measured over the occipital/temporal cortices in 50 right-handed observers. In two previous studies [23], [24] we found a bilateral activation of occipito/temporal cortex in women and right–lateralized activation in men during infant face processing as indexed by sensory ERP responses. Since in those studies all stimuli were infant faces, it lacked a control condition with non-face objects. We devised a paradigm in which face processing of persons of various age was contrasted with that of technological objects.

Fig. 2 shows grand-average waveforms recorded at occipito/temporal electrode sites as a function of the viewer's gender and stimulus content. Strong gender differences are visible, especially in the degree of N170 lateralization and discriminative response.

The ANOVA performed on the peak amplitude values of N170 revealed a significant effect of stimulus content (F3,144 = 90.4; p<0.000001), showing a larger response to childish (toddler = −3.95 µV, SE = 0.37; child = −3.44 µV) than adult faces (−2.7 µV, SE = 0.33), as proved by post-hoc comparisons (p<0.01). Furthermore, N170 to faces was much larger (p<0.01) than to objects (−0.32 µV, SE 0.26). The further significance of hemisphere x sex (F1,48 = 7.57; p<0.008) showed the presence of a sex difference in N170 lateralization, with larger N170 over the right (−2.92 µV, SE = 0.55) than the left hemisphere in men (−1.95 µV, SE = 0.42), as indicated by post-hoc tests (p<0.01), and no significant asymmetry in women (RH = −2.35 µV, SE = 0.55, LH = −3.21 µV, SE = 0.42).

ANOVA yielded the significance of stimulus content x hemisphere x sex (F 3,144 = 3.51, p<0.018). Post-hoc tests revealed no sex difference in the amplitude of N170 to objects. Furthermore, they showed a much larger face-specific response over the right than left hemisphere (Tod-obj = 5.4 µV at P10 and 2.41 at P9) in men (p<0.01), and a bilateral face specific response in women (Tod-obj = 3.79 µV at P10 and 3.0 µV at P9). It was also found a lack of face coding effect over the left hemisphere in men, with no difference in N170 to faces as a function of person's age. Conversely, a significant age-coding effect of over both hemispheres was found in women, with a larger N170 to toddler than adult faces (p<0.01) at both P9 and P10 sites and no hemispheric difference in the amplitude of N170 to faces. On the other hand, results showed larger right than left hemispheric responses to faces in men (adults, p<0.04; toddlers and children, p<0.00001). The latter effect is visible in topographical maps of Fig. 3, computed for N170 surface voltages recorded in response to adult faces, separately for men and women.

In order to locate the possible neural circuits subtending face coding in the two sexes, two different swLORETA source reconstructions were performed, separately for men and women, on 170 amplitude measured in the time window 135–185 ms, which are displayed in Fig. 4. The inverse solution showed that the processing of adult faces in women was associated with a significant activity in the left fusiform gyrus (possibly corresponding to FFA), left MOG, right cuneus (BA18), left posterior cingulate cortex, and anterior brain regions (BA10/11), as listed in Table 1. In men processing of adult face was associated with activation in the right fusiform gyrus, the left MOG, right posterior cingulate cortex and anterior brain areas (BA10/11).

Overall, electrophysiological and source localization data support previous literature about the existence of specific neural populations in the fusiform area (FG)[1] [2] and the middle occipital area devoted to face processing, as reflected by the amplitude of occipito/temporal N170 component of ERPs. Moreover, they seem to suggest a less marked lateralization in the activity of face-devoted brain regions in women than men. This finding, supported by the presence of a stronger left FG generator in women and right FG generator in men, results in a asymmetrical N170 surface amplitude in men, and a bilateral distribution in women (tending toward the left asymmetry). Moreover, over the left hemisphere N170 amplitude did not vary as a function of face age in men but only in women. And indeed N170 was of greater amplitude to toddlers' than adults' faces in both hemispheres of the female brain. This age coding effect might be due to a series of factor including a bias toward baby schema, for which toddlers' faces are judged as cuter and more attractive than adult faces [29], and have greater attention capturing effects, as well as to more perceptual face-specific mechanisms related to the fact that toddler's faces are more alike (perceptually similar) than adult faces. The sex difference in hemispheric asymmetry finding agrees with Glocker and coworkers's [30] evidence of a clear left FFA activation in women during processing of infant faces. On the other hand, they do not directly agree with a recent investigation [7] suggesting a relation between handedness and FFA lateralization, since our female participants, showing a bilateral response of face responsive areas, were indeed right-handed.

Overall, our results are in line with many studies that show differences between men and women in the degree of lateralization of cognitive and affective processes. Substantial data support greater hemispheric lateralization in men than women for linguistic tasks [31] and for spatial tasks [32]. Gender differences have also been found in the lateralization of visual-spatial processes such as object construction and mental rotation tasks [33], in which males are typically right hemisphere (RH) dominant and females bilaterally distributed. More relevant to the present experiment are the data provided by Bourne [34], who examined the lateralization of processing positive facial emotion in a group of 276 right-handed individuals. Subjects were asked to observe a series of chimeric faces with contrasting expressions and to decide which face they thought looked happier. The results showed that males were more strongly lateralized than women, showing a stronger perceptual asymmetry in favour of the left visual field (RH). There are also a number of studies that have found different degrees of lateralization in the cerebral response of men and women to emotional stimuli: men tend to demonstrate an asymmetric functioning, and women a bilateral functioning [35]–[38]. Notwithstanding the existing supporting literature (including sex differences in lateralized amygdala activity during happy and fearful face perception, e.g. [39]), is certainly not a shared knowledge in cognitive neuroscience that face processing is bilateral in the female brain and right-sided in the male brain, and such an assumption is not made anywhere in cognitive neuroscience manuals or clinical essays on prosopagnosia. For this reason, we feel that the present findings may make a great contribution toward the overall understanding of how faces are processed in the female and male brain in humans.