Variability in Infants' Functional Brain Network Connectivity Is Associated With Differences in Affect and Behavior

Variability in Infants' Functional Brain Network Connectivity Is Associated With Differences in Affect and Behavior. Caroline M. Kelsey, Katrina Farris and Tobias Grossmann. Front. Psychiatry, June 9 2021. https://doi.org/10.3389/fpsyt.2021.685754

Abstract: Variability in functional brain network connectivity has been linked to individual differences in cognitive, affective, and behavioral traits in adults. However, little is known about the developmental origins of such brain-behavior correlations. The current study examined functional brain network connectivity and its link to behavioral temperament in typically developing newborn and 1-month-old infants (M [age] = 25 days; N = 75) using functional near-infrared spectroscopy (fNIRS). Specifically, we measured long-range connectivity between cortical regions approximating fronto-parietal, default mode, and homologous-interhemispheric networks. Our results show that connectivity in these functional brain networks varies across infants and maps onto individual differences in behavioral temperament. Specifically, connectivity in the fronto-parietal network was positively associated with regulation and orienting behaviors, whereas connectivity in the default mode network showed the opposite effect on these behaviors. Our analysis also revealed a significant positive association between the homologous-interhemispheric network and infants' negative affect. The current results suggest that variability in long-range intra-hemispheric and cross-hemispheric functional connectivity between frontal, parietal, and temporal cortex is associated with individual differences in affect and behavior. These findings shed new light on the brain origins of individual differences in early-emerging behavioral traits and thus represent a viable novel approach for investigating developmental trajectories in typical and atypical neurodevelopment.

Discussion

The current study examined functional connectivity in brain networks using fNIRS and behavioral temperament using parental report in young infants. Our results show that functional connectivity in long-range cortical brain networks (FPN, DMN, and HIN) can be identified in very young infants and that functional connectivity in these networks varied considerably among infants. This supports the suitability of fNIRS in assessing functional connectivity and its variability in newborn infants. Importantly, our results also show that such variability in functional brain network connectivity systematically maps onto individual differences in infant behavioral temperament. Overall, the current findings provide novel insights into the brain origins of individual differences in affect and behavior, pointing to the early perinatal foundation of human temperament.

In line with our hypothesis, functional connectivity within the three brain networks (FPN, DMN, and HIN) was significantly greater than in the control network and significantly greater than a zero-value, indicating the existence of these long-range cortical brain networks in young infants. This provides further evidence that functional brain networks exist from early in ontogeny and are detectable in young infants (12, 17, 71). To our knowledge, this is the first study to demonstrate long-range functional connectivity in FPN and DMN in young infants, suggesting a remarkably early emergence of long-range connectivity in higher-order brain networks linked to cognitive control and self-referential processes, respectively. The current findings are noteworthy also in regard to the fact that both networks involve regions in prefrontal cortex, providing new evidence from newborns and 1-month-old infants supporting the view that prefrontal cortex plays a critical role in human brain function from very early in development (14, 72–75).

In addition to the general difference in connectivity between the functional and the (non-functional) control network, we also found that activity in the FPN was significantly greater than in the DMN and HIN (whereas there was no difference in connectivity levels found between the DMN and HIN). One possible interpretation of this finding is that functional connectivity in the FPN might have been enhanced when compared to the other functional networks because, like other resting-state studies with infants, the participants were presented with a video accompanied by music during the fNIRS measurement (37). In other words, the FPN might have been more engaged because infants were attending to external audio-visual stimuli [note that all infants were exposed to the same video (audio-visual) stimulus]. Here, it is important to mention that prior work with adults using fMRI shows that functional connectivity in higher-order cortical resting-state networks can be reliably acquired during the presentation of videos and corresponds to functional connectivity acquired in the absence of any stimulus (4, 76). Nonetheless, based on recent work showing that preterm infants display enhanced functional connectivity in higher-order cognitive networks in response to music (55), we speculate that enhanced functional connectivity in FPN might at least be partly explained by having newborn infants listen to music in the current study. Clearly, future research with infants which systematically compares stimulation protocols is needed to examine whether and how functional connectivity is influenced by the measurement context and the stimulation protocol used. Overall, our functional connectivity analysis supports the notion that intrinsic functional connectivity in cortical brain networks and its variability can be effectively mapped in newborn infants using fNIRS.

Having established functional connectivity in these brain networks as variable and distinct from a (non-functional) control network then allowed for the examination of specific associations between brain network connectivity and infant behavioral temperament. Our results confirmed our hypothesis and showed that infants' regulation/orienting behaviors were associated with functional connectivity in the FPN with greater connectivity in this network being associated with enhanced regulation and orienting. This result is in line with prior work linking functional connectivity in FPN to cognitive control of attention and behavior in adults (2, 48) and more recent work with infants (43). The current results further showed the opposite pattern of association for functional connectivity in DMN, with greater connectivity associated with reduced regulation and orienting, which is in agreement with our hypothesis based on the DMN previously being linked to self-referential, stimulus-independent thought and mind-wandering in adults (2, 48) and infants (43). To obtain such opposing effects of functional connectivity in FPN and DMN is reminiscent of seminal findings supporting the existence of anti-correlated brain networks in adults (77) and may suggest that similar organizational principles are at play in newborn infants. However, it should be emphasized that functional connectivity in the FPN and DMN in the current study was not anticorrelated as such, but rather had opposing effects on infants' behavioral and attentional regulation.

Our results concerning behavioral and attentional regulation and their functional connectivity correlates in infants are principally in line with prior research showing hyperconnectivity in the DMN and hypoconnectivity in the FPN in adults with negative emotionality and related mental health outcomes (33, 34, 48). Moreover, our data show that infants' functional connectivity in the HIN was associated with negative emotionality. Contrary to prior work with adults indicating that hypoconnectivity is associated with negative emotionality and depression (33, 35, 36) and work with infants indicating that corpus callosum length (thought to underly the HIN) is negatively associated with later emotion regulation abilities (45), the current infant data show that greater connectivity between homologous brain regions in both hemispheres was associated with greater negative affect. It is unclear why the direction of the association (positive vs. negative) would differ as a function of age (newborn infants in the current study and preschool aged and adults in previous work), but it is worth noting that the experience and display of negative affect only gradually emerges during infancy and may thus not be fully present in newborn infants (53).

Taken together, the current findings demonstrate specific associations between functional brain network connectivity and behavioral temperament in newborn infants. This suggests a remarkably early emergence of functional networks with behavioral relevance and highlights the importance of evaluating individual differences reflected in intrinsic brain connectivity. Although there are many advantages in the current approach of using fNIRS to examine functional brain connectivity, including its cost-effective and non-confining application, there are some limitations that need to be mentioned. First, because fNIRS is limited in monitoring activity from (superficial) cortical structures (78), our approach did not allow us to measure activity from deeper cortical and subcortical regions and include those in our network analyses. Second, from a developmental perspective, it should be noted that our analysis is limited to only one age group and comprised of very young infants. It is thus critically important to further assess the development of variability in these brain networks and their associations with behavioral temperament over developmental time to determine its long-term effects and the robustness of these associations (69). Third, it is important to note that these associations were assessed in a population of healthy infants, meaning there were no known birth or health complications at the time of the visit. Therefore, given the breadth of work examining how preterm birth and other medical complications (e.g., hypoxia) impact brain development, it will be important to test whether or not these associations generalize to other populations (16, 79, 80).

In conclusion, the current study provides novel insights into the use of fNIRS in identifying neural endophenotypes—variability in functional brain network connectivity—linked to behavioral temperament traits in early human development. The present findings support the notion that functionally distinct neural networks are implicated in regulatory and emotional behaviors already in newborn infants, adding a critical developmental component to efforts directed at mapping how the individual functional connectome links to affective, cognitive, and behavioral traits. The current findings shed light on the brain origins of individual differences in early-emerging behavioral traits and provide the basis for future research examining the genetic and environmental factors contributing to and the long-term developmental consequences of this brain-behavior correlation. More generally, the current study provides early ontogenetic evidence for the idea that studying functional brain network connectivity is an effective way in helping bridge the gap between brain and behavior.