Intrauterine exposure to SARS-CoV-2 infection and early newborn brain development

Abstract

Epidemiologic studies suggest that prenatal exposures to certain viruses may influence early neurodevelopment, predisposing offspring to neuropsychiatric conditions later in life. The long-term effects of maternal COVID-19 infection in pregnancy on early brain development, however, remain largely unknown. We prospectively enrolled infants in an observational cohort study for a single-site study in the Washington, DC Metropolitan Area from June 2020 to November 2021 and compared these infants to pre-pandemic controls (studied March 2014–February 2020). The primary outcomes are measures of cortical morphometry (tissue-specific volumes), along with global and regional measures of local gyrification index, and sulcal depth. We studied 210 infants (55 infants of COVID-19 unexposed mothers, 47 infants of COVID-19-positive mothers, and 108 pre-pandemic healthy controls). We found increased cortical gray matter volume (182.45 ± 4.81 vs. 167.29 ± 2.92) and accelerated sulcal depth of the frontal lobe (5.01 ± 0.19 vs. 4.40 ± 0.13) in infants of COVID-19-positive mothers compared to controls. We found additional differences in infants of COVID-19 unexposed mothers, suggesting both maternal viral exposures, as well as non-viral stressors associated with the pandemic, may influence early development and warrant ongoing follow-up.

Introduction

Viral infections in pregnancy are known to have significant adverse effects on both the mother and fetus (Racicot and Mor 2017). Several such viral infections may result in overt structural injury and disruption of fetal neurodevelopment, while others are associated with neuropsychiatric disorders that may not manifest until later in life (Amgalan et al. 2021; Ganguli and Chavali 2021). Studies of respiratory viruses have shown that pregnant women are at higher risk for severe morbidity and mortality compared to nonpregnant adults, including increased rates of stillbirth, low birth weight, and preterm birth (Siston et al. 2010; Racicot and Mor 2017; Englund and Chu 2018). Mechanisms underlying these observations highlight both direct viral toxicity and viral-induced inflammation as potential mediators of adverse fetal neurodevelopment (Racicot and Mor 2017; Chudnovets et al. 2020; Amgalan et al. 2021).

The recent outbreak of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was declared a public health emergency by the World Health Organization in early 2020 and rapidly became a global pandemic (Cucinotta and Vanelli 2020). Despite reports of vertical transmission of COVID-19 from positive mothers to infants, intrauterine transmission rates generally remain low, and neonates are typically asymptomatic (Mullins et al. 2021; Sturrock et al. 2023). Large-scale studies note that maternal infection with COVID-19 was associated with preterm birth, with conflicting data about the risk of stillbirth and intrauterine growth restriction (Mullins et al. 2022; Smith et al. 2023). Certainly, intrauterine exposure to COVID-19 may have secondary consequences, particularly related to prematurity-related injury, as well as more subtle disruptions in early brain development that may not manifest until a later age (Stafstrom and Jantzie 2020; Shuffrey et al. 2022; Brum and Vain 2023). Emerging studies on early neurodevelopmental outcomes for infants born during the pandemic, again provide conflicting reports. At 3 months, one small study noted increased abnormalities in fine motor and personal-social domains for infants born to pregnant women with symptomatic COVID-19 compared to those born to pregnant women with asymptomatic COVID-19 (Mulkey et al. 2022), while a larger study of both exposed and unexposed infants at 6 months noted lower scores on gross motor, fine motor, and personal-social subdomains of early neurodevelopment compared to pre-pandemic controls (Shuffrey et al. 2022). A recent meta-analysis indicated that while overall neurodevelopmental delays were not significant at 1 year of age, communication delays were increased in all infants born during the pandemic, independent of maternal infection in pregnancy (Hessami et al. 2022). Consistent among all reports is that the intrauterine and early postnatal environment during the pandemic may disrupt early neurodevelopment, though the mechanisms remain unclear. To better understand the structural underpinnings of altered neurodevelopment, the objective of this longitudinal observational study was to compare quantitative measures of cortical development in neonates during the pandemic, both with and without prenatal exposure to COVID-19, to a healthy group of pre-pandemic newborns.

Materials and methods

Study participants

We recruited neonates with or without prenatal COVID-19 exposures between June 2020 and November 2021 as part of a prospective, observational study on neonatal brain development and reported our results according to STROBE cohort study guidelines (von Elm et al. 2007). This was complemented by a previous (pre-pandemic) study of normative early neurodevelopment in low-risk obstetric patients with normal pregnancy outcomes. Study procedures were identical across both enrollment periods. Exclusion criteria for both periods were infants with syndromic features and suspected or documented chromosomal abnormalities. In the pre-pandemic cohort of low-risk pregnancies, we excluded infants with suspected or documented congenital infections. Following approval by the Children’s National Hospital Institutional Review Board, written informed consent was obtained from the parents of all participants.

Clinical data

Neonatal and maternal demographics were extracted from medical records, including maternal age, gestational age (GA) at birth, infant anthropometric measures (weight, head circumference, length), and Apgar scores, with subsequent calculations of gestational and chronologic age at MRI. In addition, anthropometric measures were reported as absolute values (grams and centimeters) as well as z-scores using the Fenton growth calculator (Fenton and Kim 2013). Infant race and ethnicity were reported by parents at time of MRI. Families that selected more than one, entered a separate or declined to share race or ethnicity were categorized as other/unknown. Socioeconomic data included parental reports of education and employment status. Infants were classified as prenatal COVID-19 exposed with confirmed maternal COVID-19 infection by laboratory testing during prenatal care or at delivery, along with parental reports of known or potential exposures using the Coronavirus Perinatal Experiences—Impact Survey and Update (COPE-IS/COPE-IU) (Thomason and VanTieghem 2020; Thomason et al. 2020).

MRI data acquisition

Enrolled infants underwent an outpatient neonatal MRI during natural sleep using the feed and swaddle technique on a 3 T GE Discovery MR750 MR scanner (General Electric Medical Systems, Waukesha, WI) using an 8-channel head coil. Anatomical images were acquired using a T2-weighted 3D CUBE (GE: 3D Fast Spin Echo (FSE)) sequence, repetition time: 2500 ms, echo time: 64.7–89.9 ms, flip angle: 90, orientation: S/I, number of slices: 120, and resolution: 0.625 × 0.625 × 1 mm³. MRI acquisitions were scheduled within the first 6 weeks of term corrected age, optimal for feed and bundle, without the need for sedation or medications.

Image processing

Automatic segmentation of the 3D brain image brain tissues was implemented using the Developing Brain Region Annotation with Expectation–Maximization (Draw-EM) algorithm (Makropoulos et al. 2016). Two sets of tissue labels were generated from Draw-EM: the segmentation file with 9 labels (Makropoulos et al. 2014) and the parcellation file with 50 labels (Makropoulos et al. 2016) (Fig. S1). Manual correction of tissue labels of the segmentation and parcellation files was performed by a trained research scientist (K.K.) with more than 7 years of experience using ITK-SNAP in neonatal brain segmentation during the time of this work. Twenty percent (20%) scans were randomly chosen and segmented by a second trained examiner (N.R.A.) to evaluate the inter-rater reliability and team members were blinded to COVID-19 exposure status. The intraclass correlation coefficients for all measured regions between the two examiners were higher than 0.95.

Ten brain regions of both the right and left hemispheres were extracted from segmentation and parcellation files; the frontal, parietal, temporal, and occipital lobes; anterior and posterior cingulate gyrus, insula, and corpus callosum were extracted from the parcellation file, and the deep gray matter (DGM) and ventricles from the segmentation file. These brain regions were imported to BrainSuite version 18a to generate 3D surface mesh models (Shattuck and Leahy 2002). Each mesh model contained a set of 3D coordinates of the surface vertices and a set of triangular mesh. Every surface vertex was assigned to one of these 10 brain regions.

Neonatal brain volumes

Volumes for brain tissues and structures were determined based on the voxel sizes of the images. Cerebral tissues included cortical gray matter (CGM), white matter (WM), and DGM volumes (Brain Development Cooperative 2012) (Fig. S1). The DGM volume included the thalamus (Thal), caudate nucleus (CN), subthalamic nucleus (STN), lentiform nucleus (LN), ventrolateral nuclei (VLN), as well as the subcortical background containing the internal capsule (IC).

Cortical folding

The following cortical features were measured from the inner surface of the CGM/outer surface of WM to characterize 3D neonatal brain morphology: (a) local gyrification index (LGI) and (b) sulcal depth (SD) for overall cortical development (global) and each of the four lobes (frontal, parietal, temporal, and occipital lobes) (Fig. S1) (Hill et al. 2010; Clouchoux et al. 2012; Lefevre et al. 2016). To calculate LGI and SD, the convex hull surface of the 10 brain regions was first created (Van Essen 2005). The LGI was determined as the ratio between the area of a circular region of each vertex on the WM outer surface and the corresponding area on the convex hull for the vertex (Li et al. 2014). SD was computed as the distance from each vertex on the WM outer surface to the nearest point on the convex hull (Tosun et al. 2015).

Statistical analysis

Demographic data are presented as frequency and percent or mean and standard deviation. Initial COVID-19 exposure group comparisons for neonatal and maternal characteristics were performed using one-way ANOVA or Chi-square, where appropriate.

COVID-19 exposure group comparisons of morphometric data of brain volumes and cortical features were adjusted for gestational age at MRI, infant race, infant sex, and maternal SES and represented as adjusted means and standard error. Socioeconomic status (SES) data included maternal education (college or professional school graduate vs. other) and maternal employment (business professional vs. other). Overall, linear mixed models tested for significant differences between the three groups: pre-pandemic, the COVID-19 unexposed, and the COVID-19 exposed infants. Post hoc tests between pandemic exposure groups were performed. Lastly, we adjusted for multiple comparisons using the false discovery rate method based on the number of outcomes (regions or lobes) and report unadjusted P-values as well as adjusted q-values. All statistical analyses were conducted using SAS (ver. 9.4, Cary, NC). Statistical significance was considered P ≤0.05, two-tailed for demographic data, and for models of regional brain growth and morphometry, statistical significance was considered if q ≤0.05.

Results

Demographics

We recruited 210 infants into this study (108 pre-pandemic, 102 pandemic). For the pre-pandemic cohort, we included studies from our low-risk normative cohort sequentially recruited until before the pandemic onset (May 2015 to December 2019). Within the pandemic cohort, 55 infants were COVID-19 unexposed, and 47 were exposed to maternal COVID-19 in pregnancy. Of the 47 COVID-19 exposed infants, 19 (40%) were exposed in the perinatal period (within 72 h of delivery), versus 28 (60%) in the antenatal period, 2 (4%) in the first trimester, 13 (28%) in the second trimester, and 32 (68%) in the third trimester; 33 (70%) were symptomatic at testing and 7 (15%) required hospitalization for symptom management. None of the neonates with intrauterine COVID-19 exposure tested positive after delivery. Forty (39%) women delivered prior to when the first COVID vaccine became available in December 2020. Of the remaining 62 women, 27 were vaccinated during pregnancy, 22 were vaccinated after delivery, and 13 did not share vaccination status. Ten (<5%) of enrolled participants noted the occasional use of alcohol, and none disclosed the use of substances. Forty-two (20%) of enrolled participants noted the use of additional medications.

The COVID-19 exposed cohort was notable for a significantly younger maternal age compared to pre-pandemic and COVID-19 unexposed cohorts (30.1 years vs. 34.6 vs. 34.0, P < 0.0001), gestational age at delivery (37.8 weeks vs. 39.2 vs. 39.4, P = 0.0004), and birthweight (2994 g vs. 3288 vs. 3351, P = 0.005). Thirty-seven (81%) infants were born at term, 4 (8.5%) were moderate to late preterm (32–37 weeks gestation), 1 (2%) was very preterm (28–32 weeks gestation), and 3 (6%) were born extremely preterm (≤28 weeks gestation). Corrected gestational age (CGA) at MRI was 41.98 weeks for the pre-pandemic cohort, 45.15 for COVID-19 unexposed, and 43.85 for the COVID-19 exposed cohorts (P < 0.0001). Forty-three (91%) of the COVID-19 exposed infants had no evidence of structural injury or abnormality on MRI; 1 infant had evidence of subependymal heterotopias; 2 had evidence of prior intraventricular/choroid plexus hemorrhage with periventricular white matter injury; 1 had evidence of large acute left middle cerebral artery infarct of the temporal, parietal, occipital, and posterior frontal lobes (Supplementary Table 5). Overall, female sex comprised 49% of the cohort. Race and ethnicity were significantly different with the greatest percentage of Black and Hispanic neonates in the COVID-19 exposed group (P < 0.0001) (Table 1). All subsequent analyses were adjusted for CGA at MRI, maternal SES, infant race, and sex.

Table 1.

Demographic and clinical data of the cohort.

	Pre-pandemic	COVID-19 unexposed mothers	COVID-19 exposed mothers	P-value
Number of subjects	108	55	47
Female fetus	54 [50%]	23 [58%]	27 [57%]	0.52
Maternal age (years)	34.6 [±5.95]	34.0 [±4.10]	30.05 [±7.03]	<0.001
GA at birth (weeks)	39.3 [±1.4]	39.4 [±1.2]	37.8 [±4.0]	<0.001
Birth weight (g)	3288 [±480]	3351 [±472]	2994 [±849]	0.004
Birth weight (z-score)	−0.17 [±0.89]	−0.08 [±0.83]	−0.22 [±0.95]	0.94
Birth head circumference (cm)	34.2 [±1.5]	34.5 [±1.3]	32.8 [±3.2]	<0.001
Birth head circumference (z-score)	−0.27 [±1.02]	−0.09 [±0.85]	−0.27 [±0.92]	0.46
Birth length (cm)	50.7 [±2.4]	51.0 [±2.7]	49.03 [±5.6]	0.01
Birth length (z-score)	0.15 [±0.91]	0.32 [±1.02]	0.27 [±1.05]	0.50
Apgar score at 1 min	8 [±1.2]	8 [±1.4]	7 [±2.3]	0.09
Apgar score at 5 min	9 [±0.7]	9 [±0.3]	8 [±1.1]	0.03
CGA at MRI	41.98 [±2.04]	45.14 [±2.75]	43.85 [±3.41]	<0.001
DOL at MRI	18.58 [±11.97]	40.84 [±16.20]	45.55 [±27.78]	<0.001
Racea
Asian or Pacific Islander	3 (3%)	6 (12%)	0 (0%)	<0.001
Black	19 (18%)	6 (12%)	16 (36%)
White	64 (59%)	27 (55%)	7 (15%)
Other or Unknown	21 (20%)	10 (20%)	22 (49%)
Ethnicitya
Hispanic or Latino	13 (12%)	8 (16%)	21 (47%)	<0.001

Abbreviations: GA = gestational age; CGA = corrected gestational age; MRI = magnetic resonance imaging; DOL = day of life.

^aRace and ethnicity as reported by study participants where “other” refers to additional or multiple designations entered by study participants.

Brain tissue volumes

Brain volumes calculated from segmentation data were available for 192 (92%) of subjects; remaining studies failed due to ringing artifact from the 3D T2 cube acquisitions. CGM volume was greatest in the COVID-19 exposed infants, even when adjusting for multiple comparisons (Table 2); perinatal exposure as well as third-trimester exposures were associated with higher CGM volumes; however, these associations did not remain significant when adjusting for multiple comparisons (Tables S1 and S2). The DGM volume trended lower for both pandemic cohorts, with significant decreases in COVID-19 unexposed infants compared to pre-pandemic controls (Table 2) and infants exposed to COVID-19 in the antenatal period or trimesters 1 and 2 (Tables S1 and S2). Of these, decreased DGM volume for infants with antenatal exposure to COVID-19 remained significant after multiple corrections. We did not detect any differences in brain volumes between COVID-19 exposed infants with symptomatic mothers versus asymptomatic mothers.

Table 2.

Regional brain volumes and cortical features in neonates during the pandemic with and without maternal COVID-19 infections and pre-pandemic controls.

	Pre-pandemic	COVID-19 unexposed mothers		COVID-19 exposed mothers
	Adjusted mean	Adjusted mean	P-value	Adjusted mean	P-value
Volumes (cm3)
Cortical gray matter	167.29 ± 2.92	175.63 ± 4.00	0.04	182.45 ± 4.81	0.002*
Cortical white matter	170.18 ± 2.20	167.52 ± 3.01	0.39	174.085 ± 3.67	0.29
Deep gray matter	29.53 ± 0.36	28.45 ± 0.51	0.04	28.51 ± 0.64	0.11
Local gyrification index
Global brain	2.20 ± 0.05	2.40 ± 0.07	0.002*	2.29 ± 0.08	0.24
Frontal lobe	2.13 ± 0.05	2.35 ± 0.07	<0.001*	2.25 ± 0.08	0.09
Occipital lobe	2.15 ± 0.05	2.35 ± 0.07	0.002*	2.24 ± 0.08	0.22
Parietal lobe	2.39 ± 0.06	2.56 ± 0.08	0.019	2.43 ± 0.09	0.64
Temporal lobe	2.16 ± 0.05	2.33 ± 0.06	0.005*	2.23 ± 0.07	0.33
Sulcal depth (mm)
Global brain	4.70 ± 0.13	5.14 ± 0.16	0.005*	5.17 ± 0.18	0.01
Frontal lobe	4.40 ± 0.13	4.91 ± 0.17	0.002*	5.01 ± 0.19	0.001*
Occipital lobe	3.67 ± 0.10	3.89 ± 0.13	0.08	3.85 ± 0.15	0.22
Parietal lobe	5.88 ± 0.15	6.27 ± 0.18	0.03	6.35 ± 0.21	0.02
Temporal lobe	4.53 ± 0.15	5.02 ± 0.18	0.006	4.88 ± 0.21	0.09

Data presented as mean values ± standard error. All means adjusted for corrected gestational age at MRI, maternal socioeconomic status, infant race and sex. *Denotes q < 0.05 after adjusting for multiple comparisons.

Cortical maturation

Cortical feature analyses calculated from both segmentation and parcellation data were conducted for 169 (80%) subjects, excluding the COVID-19 exposed infants with evidence of hemorrhage and infarct. The sulcal depth of the frontal lobe was significantly higher for COVID-19 exposed infants (Table 2). Frontal, parietal, and global sulcal depths also were significantly deeper for those with perinatal exposures (Table S3), but did not observe significant differences in neonates with early prenatal exposure compared to late (Table S4). Compared to pre-pandemic controls, COVID-19 unexposed infants also were noted to have significantly higher gyrification indices between the frontal, occipital, temporal, and global regions, as well as deeper frontal and global sulcal depth (Table 2). As with brain volumes, we did not detect any differences in cortical maturation features between symptomatic and asymptomatic mothers in COVID-19 exposed infants.

Discussion

Summary of findings

This study utilized advanced 3D volumetric MRI to investigate the relationship between intrauterine COVID-19 exposure and early morphometric brain development. We observed a significant increase in CGM volume and accelerated sulcal depths of the frontal lobe in infants of COVID-19 exposed mothers compared to pre-pandemic controls. For infants of mothers with COVID-19 exposures, sub-group analyses did not detect significant differences in brain morphometry based on maternal symptoms nor based on early or late gestational exposures. Antenatal exposures, however, were associated with decreased DGM, while perinatal exposures were associated with increased sulcal depths of the frontal and parietal lobes, as well as the cortical surface overall. Interestingly, we also observed an increase in gyrification indices and sulcal depths in COVID-19 unexposed infants compared to pre-pandemic controls.

Given the complex temporal and spatial sequence of events in normal brain development that unfolds across gestation, the timing of intrauterine COVID-19 exposure may exert distinct effects on early brain development. This is supported by our finding that neonates with early antenatal (first–second trimester) exposures had significant impairment in volumetric brain growth, suggesting that earlier insults disrupt proliferative stages of brain development (Volpe and Volpe 2018). Conversely, late third-trimester exposures were associated with increased CGM growth and increased sulcal depths; whether these findings suggest accelerated maturation in the setting of perinatal stress or disrupted pruning and apoptosis remains unclear. The findings of accelerated cortical maturation were also seen in COVID-19 unexposed infants, which suggests that perhaps non-viral insults specific to the pandemic may be driving these changes. Similarly, we hypothesized greater differences in COVID-19 exposed infants with symptomatic mothers compared to those with asymptomatic mothers. While we did not observe any significant differences between these groups, only 11 of 47 infants with prenatal exposures reported asymptomatic mothers, limiting the interpretation of these findings.

Fetal origins of neuropsychiatric disease

Both animal and human literature have described a link between intrauterine exposures and subsequent neuropsychiatric disease (Barker and Thornburg 2013; Amgalan et al. 2021). Although the mechanisms behind these observations are not fully elucidated, infection, inflammation, hypoxia–ischemia, and toxic stress all may adversely disrupt early neurodevelopment (Amgalan et al. 2021). Inflammation, even in the absence of active viral infection in the fetus, is a potent mediator of adverse neurodevelopment (Estes and McAllister 2016; Meyer 2019; Stafstrom and Jantzie 2020), may persist across generations (Pollak and Weber-Stadlbauer 2020) and may be triggered by a variety of insults (Estes and McAllister 2016; Meyer 2019), including maternal stress and depression (Hantsoo et al. 2019; Amgalan et al. 2021). We and others have reported an association between pregnancy stress and depression and fetal brain development, which is notable in the setting of significant escalation of maternal psychosocial stress during the pandemic (Ostacoli et al. 2020; Wu et al. 2020; Lu et al. 2021). Given the socio-emotional and economic upheaval resulting from this pandemic (DiClemente et al. 2021; Kim 2021), non-infectious pandemic stressors may be related to the brain differences seen in the COVID-unexposed neonates. Since a robust and growing body of literature highlights the low intrauterine transmission rates from mother to fetus, our findings might reflect the sustained indirect effects of the pandemic.

Alterations in cortical tissue volumes

One of the most salient findings in this work is the regional increase in CGM volumes in infants with intrauterine COVID-19 exposures. In our previous work, we showed that CGM also was greater in fetuses of unexposed pregnant women during the pandemic compared to pre-pandemic controls, while cerebral WM and DGM volumes were lower (Lu et al. 2022); while only differences in fetal WM reached statistical significance in utero and CGM differences reached significance in this work, we observe the same directions of disturbed growth across all three tissue classes between the fetal and neonatal periods.

Increasingly, it is recognized that changes in gray matter volumes vary throughout the lifetime, with region- and sex-specific findings (Gogtay and Thompson 2010; Narvacan et al. 2017). Both DGM and CGM volumes are known to increase significantly in the second half of gestation and through early infancy (Gogtay and Thompson 2010; Andescavage et al. 2017; Narvacan et al. 2017) while deviations from these typical growth trajectories have been associated with neuropsychiatric and neurodegenerative disorders. Reductions in DGM have been described in preterm birth (Ball et al. 2012; Kidokoro et al. 2013; Omizzolo et al. 2014; Hintz et al. 2015), and have been shown to predict later neurodevelopmental impairments, including cognitive and memory delays (Woodward et al. 2006; Lind et al. 2011; Omizzolo et al. 2014). Furthermore, persistent reductions in DGM volumes through late childhood and adolescence are associated with increased psychosocial symptoms and emerging psychiatric diagnoses (Botellero et al. 2017; Lind et al. 2020). Several other studies similarly have demonstrated associations between reduced gray matter and behavioral problems including attention and conduct disorders, independent of prematurity (Norman et al. 2016; Rogers and De Brito 2016). Less is known on the clinical impact of increased CGM on neurodevelopmental outcomes. However, CGM volume is thought to reflect both cortical surface area and cortical thickness development (Winkler et al. 2010), so that increased volume is consistent with our findings of advanced cortical maturation. The potential consequences of increased CGM volume on long-term neurodevelopment in those with intrauterine COVID-19 exposure need to be elucidated, and these studies are currently underway.

Changes in cortical features

The human cortex undergoes significant and rapid changes in size and shape during early development, which allows for the detailed and quantitative analyses of the developing cortex (Limperopoulos 2009; Winkler et al. 2010; Clouchoux et al. 2012; Wu et al. 2015). Several studies have shown a positive association between CGM volumes and cortical maturation CGM, similar to the observations in this work. Increases in gyrification indices have been described in adults and children with neuropsychiatric disease (Sasabayashi et al. 2021). In previous studies, we also have shown increased gyrification indices and sulcal depth in fetuses exposed to elevated maternal stress and anxiety (Wu et al. 2020) as well as those with low parental SES (Lu et al. 2021). Much less is known about the relationship between cortical features and inflammation, though one study of healthy adult volunteers did identify a significant negative association between systemic inflammation and cortical surface area (Marsland et al. 2015). We postulate that maternal stress and anxiety, known to be increased during the pandemic, may contribute to the current findings of cortical maturation. While the functional significance of our observations on long-term offspring neurodevelopment remains unknown, this is currently under investigation.

Limitations

Despite the many strengths of our study, there are several limitations that deserve mention. First, as with many studies of early neurodevelopment, rapid changes in brain shape, size, and myelination pose unique challenges in evaluating the developing brain. While we used a robust and well-accepted pipeline for the automatic segmentation, technical advancements to improve the accuracy and reproducibility of early brain morphometry remain ongoing. Second, the COVID-19 exposures were determined by both maternal testing and self-reporting. However, it is possible that women may have had unknown exposures or subclinical infections outside the clinical testing windows. Third, the potential influence of viral exposures on the developing brain depends on the timing, severity, and duration of exposure; this variability is compounded by the known variants in SARS-CoV2 and differences in symptoms and acuity. Furthermore, our findings were limited to a single-site study that included 47 infants of COVID-19 exposed mothers and 55 infants of unexposed mothers; larger, multi-site studies are warranted to validate these findings. Similar to other reports, this cohort reflects an increase in preterm birth for mothers with confirmed SARS-CoV2 infection and the additional impact of prematurity on brain development warrants further study. Fourth, while we identified alterations in regional brain volumes and cortical maturation, these differences may not be solely related to the sequelae of maternal COVID-19 infections. Additional and alternate mechanisms of intrauterine stress resulting from the COVID-19 pandemic may contribute to these findings; indeed, the fact that there are significant morphometric differences in infants without maternal COVID-19 exposures compared to pre-pandemic controls is suggestive of additional pandemic factors that may influence early brain development. Though we included maternal education and employment data in the models, additional socioeconomic factors may also influence early brain development. Similarly, while we adjusted for gestational age at MRI, early extrauterine factors may have independently influenced these results; given the statistically significant differences between gestational age at MRI between cohorts, the potential for such exposures to influence our findings must be considered. These, along with differences in the time periods of recruitment for pre-pandemic and pandemic cohorts, suggest that a detailed examination of other pandemic and non-pandemic influences on early neurodevelopment is warranted. Lastly and most importantly, the long-term neurodevelopmental significance of our findings is unknown and warrants further study.

Conclusions

Adverse prenatal exposures are known risk factors for later neuropsychiatric disease and have been associated with altered brain morphology across the lifespan. We report differences in neonatal brain morphometry between infants born to COVID-19 unexposed and COVID-19 exposed mothers, suggesting that both viral and non-viral maternal exposures during the pandemic may influence early neurodevelopment. Despite known prenatal risk factors for developing neuropsychiatric conditions later in life, not all children born to mothers with viral infections develop a neuropsychiatric illness. There remains significant heterogeneity in the incidence, type, and severity of neuropsychiatric conditions that may manifest after such exposures (Meyer 2019). This work highlights the importance of continued research to understand whether these early life alterations in brain development converge to result in neuropsychiatric disease in order to identify early neurobehavioral impairments and facilitate access to care. Equally important will be the development of robust and targeted interventions for both the pregnant woman and child to attenuate these and other early risk factors that can disrupt early neurodevelopment and to optimize maternal–infant outcomes.

Author contributions

Drs. Andescavage and Limperopoulos had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Lu, Wu and Mr. Kapse contributed to investigation and formal analysis. Drs. Keller, Von Kohorn, Afifi, and Vezina, contributed to the investigation. Ms. Henderson, contributed resources and visualization. Drs. Wessel, du Plessis and Limperopoulos contributed to the conceptualization, funding acquisition, resources and supervision. Dr. Andescavage contribute the writing - original draft, and all authors contributed to the writing, review and editing.

Funding

This study was funded by National Institutes of Health (NHLBI R01 HL116585-01), Intellectual and Developmental Disabilities Research Center, and A. James & Alice B. Clark Foundation. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

 

Conflict of interest statement: None declared.

Data availability

Data can be made available from the corresponding author upon request.