Abstract:

People tend to give priority to negative information and allocate more cognitive resources such as perception, attention and memory to negative, compared to positive, information. This phenomenon is called "negativity bias", which is well established across toddlers, children, adolescents and adults. However, this emotional bias remains controversial in infants, especially in young infants that are less than six months old. Furthermore, it is still unclear whether the emotional bias changes from no bias or positivity bias to negativity bias during infants’ development in the first year of life.

In this study, we used near-infrared spectroscopy to examine the neural responses to angry and happy prosodies in 45 neonates (22 girls and 23 boys, M_age = 5.2 ± 1.7 days old) and 45 infants (24 girls and 21 boys, M_age = 12.7 ± 1.4 months old). The experiment was conducted in the neonatal ward of Peking University First Hospital. NIRS data were recorded when the infants were at active sleeping or staying quietly (Figure 1; Table 1). Using a passive listening task, we investigated the brain functional connectivity during automatic processing of emotional prosodies of anger and happiness. The experiment was divided into three emotional blocks (using angry, happy and neutral prosodies, respectively). The order of the three blocks was counterbalanced among the participants. Each block contained 10 sentences, which were repeated six times, that is, 60 sentences were presented during the experiment in a random order.

The results showed that emotional category had a significant main effect on 60 pairs of functional connectivity (F (2, 176) ≥ 8.518, p _(FWE) ≤ 0.044), which revealed that angry and happy prosodies evoked stronger functional connectivity than neutral prosody, whereas there was no significant difference between the angry and happy conditions. The observed significant functional connectivity was mainly distributed within the right hemisphere or across bilateral hemispheres (Figure 2). More importantly, the interaction between emotion and group was significant in 21 pairs of brain connectivities (F(2, 176) ≥ 8.355, p_(FWE) ≤ 0.040; Table 2). In the neonatal group, it was found that: 1) the main effects of the above emotions were significant in the brain connectivities of 3/21 pairs (anger > neutral; t (44) ≥ 2.932, p_(FWE) ≤ 0.039) and 21/21 pairs (happiness > neutral; t (44) ≥ 3.543, p_(FWE)≤ 0.040); 2) Significant positive bias was found in 14/21 pairs of brain connectivities (happiness > anger; t (44) ≥ 3.932, p_(FWE) ≤ 0.046). Of these 14 pairs of brain functional connectivities, 12 pairs were distributed in the right hemisphere (connectivities between the right hemisphere's inferior frontal gyrus, superior temporal gyrus, supramarginal gyrus and angular gyrus), and 2 pairs were transhemispheric brain functional connectivities. In the infant group, it was found that: 1) the main effect of emotion was significant in the brain connectivities of 20/21 pairs (anger > neutral; t (44) ≥ 3.926, p _(FWE) ≤ 0.047) and 3/21 pairs (happiness > neutral; t(44) ≥ 4.123, p_(FWE) ≤ 0.023); 2) Significant negative bias was found in 10/21 pairs of brain connectivities (anger > happiness; t (44) ≥ 4.055, p_(FWE) ≤ 0.030). Of these 10 pairs of brain functional connectivities, 8 pairs were distributed in the right hemisphere (connectivities between the right hemisphere's inferior frontal gyrus, superior temporal gyrus, supramarginal gyrus and angular gyrus), and 2 pairs were transhemispheric brain functional connectivities.

In addition, the group comparison was carried out in three emotional conditions (Figure 3, Table 2). Under the condition of happy prosody, it was found that 5/21 pairs of brain connectivities were stronger in the neonatal group than in the infant group (neonate > infant; t (44) ≥ 3.820, p_(FWE) ≤ 0.022). Four of these five pairs of brain functional connectivities were distributed in the right hemisphere (the connectivities between the right hemisphere's inferior frontal gyrus, superior temporal gyrus, supramarginal gyrus and angular gyrus), and one pair is the cross-hemisphere brain functional connectivity. Under the condition of angry prosody, 11/21 pairs of brain connectivities were found to be stronger in the infant group than in the neonatal group (infant > neonate; t(44) ≥ 3.588, p_(FWE) ≤ 0.049). Of these 11 pairs of brain functional connectivities, 7 pairs were distributed in the right hemisphere (connectivities between the right hemisphere's inferior frontal gyrus, superior temporal gyrus, supramarginal gyrus and angular gyrus), and 4 pairs were transhemispheric brain functional connectivities. No significant inter-group differences were found under neutral prosodies.

By examining the neural response to emotional prosodies at two time points (0 and 1 year old), this study revealed for the first time the changes of emotional bias in a developmental perspective. We found that emotional processing has a positive bias at the beginning of postnatal period, revealed by the stronger functional connectivity for happy than for angry prosodies at the right hemisphere of the superior temporal gyrus, the inferior frontal gyrus, the supramarginal gyrus, and the angular gyrus. However, the emotional processing bias reverses in 1-year-old infants, that is, the brain functional connectivity within the above mentioned brain regions is stronger for angry than that for happy prosodies. Therefore, the reliable phenomenon of “negativity bias” is not innate, although it is always observed in adults and children. Instead, we propose that there is a developmental change from positivity bias to negativity bias in the first year of human life.

Key words: negativity bias, neonate, infant, affective prosody, superior temporal gyrus, inferior frontal gyrus