In the present study, for the first time, we explored the impact of prenatal and perinatal factors on behavioral traits within the delicate developmental phase, by posing special attention to the potential mediator role of brain anatomical features.
Our findings suggest a key role of obstetric complications in affecting both brain structure and behavior. Specifically, we found a link between complications during pregnancy, including infections, GM alterations in the right frontal gyrus, and withdrawn problems. These reciprocal associations were further explored via mediation models, providing valuable information that could be used for developing innovative strategies for psychiatric risk monitoring and prevention of transition from at-risk to disease states.
From a psychopathological perspective, we found that pregnancy complications were associated with higher withdrawn problems. From a neuroanatomical point of view, we observed associations between pregnancy complications and GMV reductions in the right superior medial frontal and anterior cingulate cortices. Finally, for the first time to the best of our knowledge, a mediation effect of the above neuroanatomical features on the influence of prenatal infections and the withdrawn dimension of CBCL was identified.
Overall, these results confirm the importance of the early management of infections and other morbidities during pregnancy, given their critical role in brain development and consequently in behavioral problems during childhood and adolescence periods (57).
Pregnancy complications and brain anatomy.
Our results suggest a critical role of maternal infections (like rubella, herpes, and toxoplasma) and other health problems (e.g., preeclampsia) in brain development, exerting a long-term impact on brain characteristics.
Specifically, critical areas contributing to superior cognitive functions, such as the rSFG and ACC resulted to be affected by pregnancy complications.
The SFG is a key brain region implicated in various superior cognitive processes, including mood, theory of mind, and self-awareness, and its alterations have been widely associated with psychiatric disorders (58, 59). Moreover, the ACC functions encompass motivation, decision making, learning, cost-benefit calculation, conflict, and error monitoring as well as social decision making (60, 61) ACC seems also to be implicated in empathy and prosocial behavior (62) and is altered in different psychiatric conditions such as depression (63–65) and bipolar disorder in adults (66, 67) and adolescents (65, 68, 69).
Therefore, it is not surprising that in our sample GMV deficits in the ACC corresponded to greater withdrawn behavior since this trait has been linked in previous studies to an altered volume and function in these structures (70–72).
Moreover, our results are in line with human and animal studies suggesting that infections during pregnancy might influence later brain development (73). Specifically, MRI studies in both the pre-and the post-natal periods detected alterations ranging from microcephaly, cystic evolution calcifications in the cortico-subcortical junction, frontal polymicrogyria, ventriculomegaly and extensive delayed myelination also in frontal lobes associated with infections during pregnancy (74, 75). Also, in a sample of preterm infants, infections during pregnancy were associated with alterations in both GM and WMV in the left frontal lobe (76). Similarly, animal models showed that maternal immune activation after infection during pregnancy was associated with impairments in tasks dependent on the prefrontal cortex and schizophrenia-like phenotype in the offspring (77). Moreover, a recent study (78) suggested that maternal immune activation might alter the expression of genes involved in neurodevelopment (e.g., genes involved in glutamatergic neurotransmission, mTOR signaling and potassium ion channel activity) and might have a negative impact on social behavior of the offspring, thus creating an autism-like phenotype.
More generally, PC seem to result in a wide spectrum of mental illnesses ranging from psychotic disorders to depression and anxiety in adulthood (5). In this regard, prenatal infections and inflammation seem to play an important role in schizophrenia (79) and a large body of evidence sustain the hypothesis that prenatal exposure to infections may cause changes in fetal brain development that further lead to a wide spectrum of cognitive deficits and neuropsychiatric disorders, including autism spectrum disorder, schizophrenia, depression, and bipolar disorder (57, 73, 80).
Pregnancy complications and behavior
Our findings suggest that complications during pregnancy might affect developmental behavior in the specific withdrawn dimension, ultimately leading to a higher score in the corresponding CBCL/6–18 subscale. This result is in line with existing evidence reporting an association between perinatal complications, including infections, and behavioral issues in children (30–33).
Indeed, internalizing behavior in offspring has been linked to prenatal maternal infection/immune activation (33). Furthermore, interleukin levels and maternal inflammation during pregnancy were found predictive of more severe internalizing symptoms, especially in female offspring (81). Similarly, maternal infections during the second trimester were significantly associated with depression during adolescence (82).
Regarding the CBCL, withdrawn behavior has been found to anticipate the development of numerous behavioral and psychiatric disorders, especially mood disorders (42, 83, 84). Since children with sub-clinical elevations in withdrawn scores have an increased risk of developing depression in the future (84), the withdrawn score elevation can be considered a clinical risk indicator for future depression.
However, the reasons why maternal immune activation can lead to alterations in offspring’s behavior is not clear yet. One of the proposed biological theories suggests that immune activation in a critical development period could alter the hypothalamus-pituitary axis, with consequent worse stress response, ultimately leading to depression and anxiety (5, 85).
Therefore, given the renowned association of withdrawn problems in childhood and adolescence with the future development of psychopathology (83, 86, 87), our and previous evidence remarks the need to prevent any infection during pregnancy. Indeed, in a cohort of more than 1 million children, the risk of psychiatric disorders in both childhood and adolescence was found to increase in a dose-response fashion with the number of maternal infections (88).
Neuroanatomy and behavior
In our study we found an association between the WD of the CBCL and reduced GMV in the frontal brain regions that resulted influenced by prenatal factors. These areas were the same affected by pregnancy complications, suggesting a relationship between these three components.
Previous neuroimaging studies performed in similar samples report widespread brain alterations associated with depressive and internalizing symptoms, encompassing both cortical and subcortical structures (70,89–92). In line with our findings, in a study conducted on children with a history of traumatic brain injury, right frontal gyrus cortical thickness was predictive of social problems (89). Similarly, the left frontal cortex was associated with worse internalizing problems in children with intractable epilepsy (90). Of note, in a recent mediation analysis (93), frontal gyrus volume was reported to significantly mediate the effect of adolescent-maternal relationship quality on both internalizing symptoms and adaptive functioning.
Similarly, functional studies confirmed the putative association between frontal areas and internalizing symptoms. In a study conducted on adolescents, resting state functional connectivity between amygdala and SFG was associated with higher internalizing symptoms (94). Altered amygdala connections to different cortical areas including the cingulate cortex have also been associated with social and emotional impairment in a sample of prematurely born (70), suggesting a role of this circuit in developing internalizing symptoms.
Role of frontal neuroanatomy in mediating the relationship between prenatal complications and withdrawn behavior.
The mediation analyses performed in our study suggested that neuroanatomy acts as a mediator in the association between obstetric complications and withdrawn scores measured via CBCL. To our knowledge, this is the first evidence of a possible mediator role of neuroanatomy in the developmental pathway from maternal infections and morbidities during pregnancy to adolescence internalizing symptoms. This finding could shed light on the underlying mechanism of the developmental origin of adults’ disease (also known as the fetal origin hypothesis) according to which many adult diseases find their basis in prenatal exposure (95, 96).
Although novel, these findings are consistent with the extant literature on the functions of the frontal cortex, highlighting its role in emotion regulation, which typically develop in adolescence (97). However, the mechanism underlying the mediation effect of the frontal cortex is still unknown. One of the possible putative causes behind the relation between maternal risk factors and withdrawn behavior in children is the dysregulation of the inflammatory response, including proinflammatory molecules (98). Immune activation in critical development periods could result in alteration in the hypothalamus-pituitary axis, with consequent worse stress response (5, 85). Similarly, perinatal stress might have an impact on the connection between the amygdala and frontal areas (99), thus leading to further alterations during adolescence (70). Given the cruciality of the adolescent period and the key role of the amygdala–prefrontal circuitry in emotional self-regulation, an alteration in this pathway could reflect an initial vulnerability to internalizing symptoms (70, 94) and further risk to develop anxiety or depression (100).
In this context, our results could help the study of intermediate phenotypes or endophenotypes that might clarify the etiology of otherwise complex syndromes. Endophenotypes have only minimally been utilized to explore the perinatal development of internalizing symptoms and psychiatric vulnerability (101). Regarding the risk of depression, frontal regions have been reported as putative areas that might be regarded as endophenotypes for the development of such psychiatric disorders (102).
However, more research is needed to clarify the effect of prenatal stressors on complex behaviors in adolescents and adults, particularly because our study included only one-time point. Nonetheless, in line with the gene-environment interaction model (103), our result points out the fundamental importance of studying the interplay between environmental risk –and protective factors and both brain structure and behavior. By unveiling endophenotypes that might further allow us to better define and stratify the risk for behavioral and psychiatric disorders, it would be possible to create specific interventions to be conducted in time windows that anticipate the transition from the at-risk state to the onset of disabling psychiatric disorders.
Limitations
Despite its potentially relevant contributions to the neuroscientific literature, this study has limitations that should be considered when interpreting the results. First, the sample size is limited by difficulties in conducting the MRI acquisitions, which required a high level of participants’ compliance that was not always achieved due to the young age and clinical characteristics of our sample. Consequently, our sample size is on the low end of what is recommended for moderated mediation models (104), therefore these results should be considered as a preliminary evidence suggestive of a mediation role of brain anatomy on the influence of prenatal issues, on internalizing behaviors. However, our sample characteristics are in line with those from other studies in the field (93, 105). The imbalance between females and males represents a confounder that might have affected the results even after the inclusion of sex as biological covariate in the analyses. Moreover, a limit to our analysis is given by the fact that, luckily, pre- and peri-natal complications are not common events, as they are reported in a minority of our sample. Therefore, larger samples are needed to obtain higher statistical power and our results should be tested on bigger replication samples to minimize the risk of false positives; multicentric approaches could reduce the possible sources of bias due to other environmental confounders, such as ethnicity, social environment, and other stressors acting during infancy and adolescence. In our statistical analyses, the consideration of age and sex (and MRI scanner and TIV in the GMV analyses) as the only covariates might have left residual confounders like postnatal environmental factors (maternal mental health, discrimination, school environment, social protective factors, etc.), which should be taken into account in future analyses.
Finally, future studies should include other behavioral measures, information on cognitive performances, and multimodal neuroimaging techniques to improve our understanding on the pathways leading from pre- and peri-natal disturbances to altered behaviors. As a future perspective, newer strategies, such as fetal MRI in combination with new genetic methods, could be useful in detecting even earlier structural alterations and to follow patients up with multiple scans in different periods of life.
In conclusion, our findings suggest a key role of obstetric complications in affecting both brain structure and behavior in children and adolescents Thanks to the use of a mediation design, we provided evidence for a mediator role of frontal cortex anatomy in the relationship between prenatal maternal issues and internalizing symptoms. Nonetheless, it would be desirable to promote prevention and early treatment programs for health during pregnancy, which we believe to be a primary goal of all care plans. Furthermore, early behavioral assessments in childhood should be encouraged to address in such a crucial time window the risk factors for the future development of disabling psychiatric disorders.