Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jan 1.
Published in final edited form as: Brain Behav Immun. 2024 Nov 10;123:1071–1080. doi: 10.1016/j.bbi.2024.11.014

The association of maternal COVID-19-infection during pregnancy on the neonatal immune profile and associations with later diagnosis of neurodevelopmental disorders

Danielle HJ Kim 1,2, Lisa A Croen 3, Ana-Maria Iosif 4, Jennifer L Ames 3, Stacey Alexeeff 3, Yinge Qian 3, Robert H Yolken 5, Paul Ashwood 6,7, Judy Van de Water 1,7,*
PMCID: PMC11684471  NIHMSID: NIHMS2042041  PMID: 39532198

Abstract

Despite the prevalence and significant concern of COVID-19 in maternal and offspring health, little is known about the impact of COVID-19 during pregnancy on newborn immunity and neurodevelopment. This study aimed to examine 1) the relationship between maternal COVID-19 during pregnancy and newborn immune profiles and investigate the 2) associations between specific newborn immune profiles and the risk of subsequent diagnosis of a neurodevelopmental disorder (NDD) among children with prenatal exposure to COVID-19. Newborn dried bloodspots (NBS) from 545 children born at Kaiser Permanente Northern California between January 2020 and September 2021 (460 [223 males, 237 females] to COVID-19-infected [COVID+] mothers; 85 [45 males, 40 females] to COVID-19-uninfected [COVID-] mothers) were used to profile newborn immune molecules via a 42-plex cytokine/chemokine assay. Among the 460 children born to COVID+ mothers, 73 (47 males, 27 females) were later diagnosed with an NDD. In the first set of analyses examining the association between maternal COVID-19 infection during pregnancy and newborn immune profile, the results adjusted for covariates but uncorrected for multiple comparisons showed that newborns of COVID+ mothers had significantly higher levels of IL-22 (est.= 0.16, 95% Cl 0.01, 0.3, p= 0.04) and GM-CSF (estimate [est.] = 0.27, 95% Cl 0.09, 0.46, p= 0.004) compared to newborns of COVID− mothers. These differences were no longer statistically significant after multiple comparison adjustments. In the second analysis exploring the association between newborn profile and later diagnosis of NDD among newborns born to COVID+ mothers, the results adjusted for covariates revealed an association between higher neonatal levels of IL-22 (HR= 0.49, 95% Cl 0.33, 0.75, p= 0.001) and lower risk of a later diagnosis of an NDD, which remained significant after multiple comparison adjustments (p= 0.04). Other neonatal cytokines/chemokines/growth factors such as sCD40L (HR= 0.7, 95% Cl 0.57, 0.9, p= 0.009), IP-10 (HR= 0.46, 95% Cl 0.25, 0.83, p= 0.009), MIG (HR= 0.52, 95% Cl 0.3, 0.9, p= 0.02), FLT-3L (HR= 0.45, 95% Cl 0.24, 0.83, p= 0.01), PDGF AB/BB (HR= 0.56, 95% Cl 0.36, 0.99, p= 0.046), VEGF (HR= 0.57, 95% Cl 0.34, 0.98, p= 0.04), and IL-4 (HR= 0.48, 95% Cl 0.26, 0.93, p= 0.03) were no longer statistically significant after multiple comparison adjustments. Despite the imbalance between the number of COVID-19 exposed and unexposed newborns in this study cohort, our novel findings enhance our understanding of the potential impact of maternal COVID-19 infection during pregnancy on the developing neonatal immune system. Our findings highlight the role of immune molecules, beyond those considered to be pro-inflammatory, that may be crucial in maternal and newborn immunity against COVID-19 infection during pregnancy. Furthermore, our results suggest that reduced levels of neonatal immune molecules in newborns of COVID+ mothers may be linked to an increased risk of a subsequent diagnosis of an NDD.

Keywords: newborn immunity, cytokine, SARS-CoV-2, COVID-19 during pregnancy, neurodevelopmental disorder (NDD), maternal immune activation (MIA)

Introduction

Numerous clinical and preclinical studies have suggested adverse effects of maternal immune activation during pregnancy, including disruptions in cytokine/chemokine levels impacting fetal development, potentially leading to abnormal neonatal immune profiles and neurodevelopmental disorder (NDD) (Benesova, 1994; Gumusoglu and Stevens, 2019; Han et al., 2021; Shimizu et al., 2023). While confirmed COVID-19 cases in pregnant women are still occurring, the effects on child health, especially immune and nervous system development, remain largely unknown. Beyond investigating the link between maternal gestational COVID-19 infection and altered neurodevelopmental outcomes (Ayed et al., 2022; Ayesa-Arriola et al., 2023; Edlow et al., 2022; Hessami et al., 2022), there is a crucial need to understand underlying mechanisms, particularly immune-related factors, that contribute to offspring NDD following gestational COVID-19 exposure.

One of the detrimental clinical outcomes of COVID-19 infection is the excessive, uncontrolled levels of circulating pro-inflammatory cytokines in the infected individual. This, along with immune cell hyperactivation, can have a negative impact on the health of the pregnant woman and developing fetus (Cron et al., 2023). Certain maternal cytokines/chemokines can act on or pass through the placenta with the potential of having a deleterious effect on the fetus (Zaretsky et al., 2004) or the placental environment itself (Shook et al., 2022). It is of interest to examine how the maternal immune response to COVID-19 during pregnancy impacts fetal immune development, including neonatal cytokine/chemokine/growth factor profiles, and subsequent child neurodevelopment.

Using newborn screening bloodspot samples from newborns with or without prenatal exposure to COVID-19, the current study investigated the association between gestational exposure to COVID-19 and cytokine/chemokine/growth factor concentrations at birth. Further, we assessed the relationship between neonatal immune profiles and NDD among children with prenatal COVID-19 exposure. This study is one of the first to 1) assess 42 immune molecules of neonates at birth using neonatal bloodspots, 2) explore how maternal COVID-19 infection during pregnancy is associated with altered neonatal immunity, and 3) investigate how the immunity of newborns exposed to COVID-19 in utero is associated with a later diagnosis of an NDD.

Materials and methods

Participants

The study was set within Kaiser Permanente Northern California (KPNC), an integrated healthcare system serving 4.5 million health plan members in the San Francisco/Bay Area, Sacramento metropolitan area, and surrounding counties. To be eligible for the study, women needed to: 1) be members of KPNC, 2) be pregnant in 2020, 3) deliver a liveborn baby, and 4) participate in the online KPNC COVID-19 Pregnancy Survey (Ames et al., 2021). Approximately 15,000 women met all the eligible criteria (Supplementary Table 1), and we further identified all those with a positive COVID-19 PCR test anytime during pregnancy (COVID+) and all women who did not have COVID-19 during pregnancy (COVID-). A subset of COVID− women were randomly selected to match COVID+ women on maternal age, maternal race/ethnicity (participant self-report), and Neighborhood Deprivation Index. The initial study was designed to investigate newborn immune marker profiles in newborns with prenatal exposure to COVID-19.; and associations between these newborn immune profiles and NDD outcome only among children born to COVID+ mothers. While the original plan was to test 550 newborn bloodspots (NBS) of those with COVID+ mothers, we were able to identify only 465 NBS from newborns of exposed mothers. Thus, we were able to add 85 NBS samples obtained from newborns born to COVID− mothers to our analysis. While this resulted in an imbalance of exposed vs non-exposed samples, it afforded us the opportunity to examine the immune parameters among children born to COVID+ as well as COVID− mothers. The newborn blood spot samples were obtained from the California Newborn Screening Program specimen archive maintained by the Genetics Diseases Branch (GDB), California Department of Public Health. This study was approved by the KPNC and the State of California Institutional Review Boards.

COVID-19 exposure definition

Maternal COVID-19 infection during pregnancy was ascertained from the COVID-19 laboratory test results (PCR) recorded in the maternal KPNC electronic health record (EHR) during pregnancy. Pregnant patients with a positive SARS-CoV-2 PCR test anytime between the last menstrual period and the date of delivery were considered COVID+. Mothers were considered COVID− if they had a negative COVID-19 PCR test or were not tested due to not having symptoms during pregnancy and no self-report of COVID-19 infection during pregnancy on the KPNC COVID-19 Pregnancy Survey. Rapid test results were not included in the COVID-19 exposure definition.

Child neurodevelopmental outcome definition

KPNC implements a universal child developmental screening program using the Developmental Milestones Questionnaire from the validated Survey of Well-Being of Young Children (SWYC) at the 18-month well-child visit and The Parent’s Observation of Social Interactions form of the SWYC at the 18- and 24-month well-child visit. Most children who screened positive for developmental or autism spectrum disorder (ASD) concerns underwent a secondary screening using the Ages and Stages Questionnaire and the Modified Checklist for Autism in Toddlers - Revised with Follow-up. Children with developmental concerns following screening were referred to Developmental Behavioral Pediatrics or an ASD Evaluation Center for further assessment and diagnosis. All children with any of the following diagnoses recorded in their EHR (January 2020 – May 2024; age range until end of follow-up= 1–51 months) were considered to have an NDD: ASD (F84.x), developmental disorders (F80, F81, F82, F88, F89) including motor delay (F82), speech delay (F80), and learning disorder (F81), and cognitive disorders (F70-F73, F78-F79).

Dried neonatal bloodspot collection and elution

Dried NBS specimens were obtained by the heel-stick method and spotted onto a standardized filter paper within 48–72 hours of birth. Three 3 mm punches per sample were put into single wells in a 96-well plate and stored at −80°C until elution with 200 μl of elution buffer, consisting of phosphate-buffered saline (PBS), 0.5% bovine serum albumin, and protease inhibitors (Complete Protease Inhibitor Cocktail, Roche Diagnostics Corporation, Indianapolis, Indiana). Plates were shaken overnight at 4°C. Eluates were analyzed immediately following elution and a short incubation with protease inhibitors, dipeptidyl peptidase IV (DPPIV), and Perfabloc® SC (PEFBSC-RO Roche, Millipore Sigma) in PBS. 4 μl aliquot per eluted sample was used for bicinchoninic acid assay (BCA, Thermo Scientific, Rockford, IL) for total protein concentration analysis.

Cytokine/chemokine/growth factor measurement

Cytokine/chemokine/growth factor concentrations in NBS were measured using a 42-plex cytokine/chemokine/growth factor panel (MILLIPLEX MAP Human Cytokine/Chemokine Magnetic Bead Panel, Millipore Sigma, Burlington, MA) following the manufacturer’s instructions [16]. The panel included sCD40L, epidermal growth factor (EGF), eotaxin, fibroblast growth factor 2 (FGF-2), Fms-related tyrosine kinase 3 ligand (FLT-3L), fracktalkine, granulocyte colony-stimulation factor (G-CSF), granulocyte macrophage colony-stimulating factor (GM-CSF), GRO1 oncogene (GROα; CXCL1), interferon alpha 2 (IFNα2), IFNγ, interleukin (IL)-1α, IL-1β, interleukin-1 receptor antagonist (IL-1RA), IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-12p40, IL-12p70, IL-13, IL-15, IL-17A, IL-17E/IL-25, IL-17F, IL-18, IL-22, IL-27, interferon gamma-induced protein 10 (IP-10), monocyte chemoattractant protein-1 (MCP-1), MCP-3, macrophage colony-stimulating factor (M-CSF), macrophage-derived chemokine (MDC; CCL22), monokine induced by gamma (MIG; CXCL9), macrophage inflammatory protein-1 alpha (MIP-1α), MIP-1β, platelet-Derived Growth Factor AA (PDGF-AA), PDGF-AB/BB, regulated upon activation, normal T cell expressed and secreted (RANTES; CCL5), transforming growth factor alpha (TGFα), TGFβ, tumor Necrosis Factor alpha (TNFα), and vascular endothelial growth factor (VEGF). Each analyte concentration was normalized to the total protein concentration based on the BCA assay to standardize sampling variation in blood collection. Analytes with more than 55% below the lowest detection value (fracktalkine, COVID-19 exposed= 59%, COVID-19 unexposed= 55%; IL-10, exposed= 92%, unexposed= 98%; IL-17A, exposed= 64%, unexposed= 98%; IL-17F, exposed= 64%, unexposed= 67%; TNFβ, exposed= 62%, unexposed= 64%) were excluded from further analyses. For analytes with less than 40% values below the lowest detection, missing data were imputed by batch with lowest detected value (LDV)/√2 (Boss et al., 2019). The distribution of missing data across neonatal cytokines/chemokines/growth factors was similar between COVID-19 exposed and COVID-19 unexposed groups. Supplementary Table 2 summarizes the missingness across all analytes, by COVID-19 exposure. All analytes were natural log transformed before statistical analyses. Five samples were excluded from the analyses due to the bloodspot collection time being > 72 h after birth or missing from the health record.

Covariates

Covariate data were extracted from KPNC EHR and birth certificate files. Variables with known associations with COVID-19 infection, neonatal cytokine/chemokine/growth factor levels, or associated with exposure or outcome variables in our dataset such as child sex, child birth season and year, child’s age at bloodspot collection, child health insurance length, maternal race, pre-pregnancy obesity status, maternal gestational diabetes, maternal infection defined as not COVID-19, maternal autoimmunity, breastfeeding status, delivery method, and Luminex plate number were considered for inclusion in multivariable models. The selection of covariates included in the models varied depending on the analysis objective and the type of analysis, as described below.

Statistical analysis

Simple and multiple linear regression models were used to assess unadjusted and adjusted differences in neonatal cytokine/chemokine/growth factor concentrations between infants born to COVID+ mothers and those born to COVID− mothers (McCulloch et al., 2008). For each analyte (outcome variable), we started with an unadjusted model that included only maternal COVID-19 infection status as a predictor. Subsequently, we fitted a model with a term for maternal COVID-19 infection status (COVID+ or COVID-) and adjusted for child sex, birth season and year (a combination of birth season (Winter, January to March; Spring, April to June; Summer, July to September; Fall, October to December) and year, age at bloodspot collection, maternal race, pre-pregnancy obesity status, and Luminex plate number (Model 1). We chose these factors as covariates as they are strongly associated with neonatal immunity noted in the literature. Finally, we further adjusted Model 1 for maternal gestational diabetes; maternal prepregnancy obesity; maternal asthma, non-COVID-19 infection, or autoimmunity during pregnancy; breastfeeding status, and delivery method (Model 2) all factors previously associated with neonatal immune status and/or neurodevelopmental outcomes. The maternal COVID-19 infection status term in these linear regression models estimated the unadjusted and adjusted mean concentration difference between children born to COVID+ mothers and those born to COVID− mothers on a log scale. Exponentiating the estimate for maternal COVID-19 infection status gave the percent difference between the concentration levels of newborns born to COVID+ mothers and those born to COVID− mothers. Since the cytokines/chemokines had different ranges, to aid interpretation, we used the multiple linear regression models (Model 1) to calculate standardized effect sizes (Cohen’s d) that accounted for the imbalance in groups and covariates (Nakagawa and Cuthill, 2007). We used the following formula:

d=t(n1+n2)n1n2df

where t is the t value obtained for the COVID-19 infection status term (evaluating the difference between COVID+ and COVID-) from the multiple linear regression model, df is the degrees of freedom used for the t value, and n1 and n2 are the sample sizes for COVID+ and COVID− groups in the respective model.

Cox proportional hazards regression models were then used to examine associations between the neonatal cytokine/chemokine/growth factor levels and risk of NDD in the subsample of children born to COVID+ mothers. We employed a similar strategy by fitting both unadjusted and adjusted models, accounting for covariates known to be associated with NDD outcomes and the neonatal immune system. Children were followed from birth to an NDD diagnosis or a censoring event (leaving the KPNC health system or May 31, 2024, whichever occurred first). Each model included a natural log-transformed neonatal cytokine/chemokine/growth factor concentration as the independent variable. Due to the modest number of NDD cases in our sample, we minimally adjusted for the combination of child birth year and season, maternal age, gestational age, and Luminex plate number (Croen et al., 2024). Unadjusted and adjusted profiled likelihood hazard ratios (HR) and 95% confidence intervals (CI) were calculated to quantify the association between cytokine/chemokine/growth factor concentrations and the risk of NDD. To aid interpretation, we also calculated and graphed HR associated with a 1 SD increase in each of the analytes.

All analyses were implemented using SAS OnDemand 9.4 (SAS Institute Inc., Cary, NC). All tests were two-sided, and P-values< 0.05 were considered statistically significant. For each analysis, we applied false discovery rate (FDR) correction for multiple comparisons with α= 0.05 to mitigate the risk of Type I error. We also present the uncorrected p-values to offer a more comprehensive view of the potential associations or differences.

Results

The analytic sample included 545 children (268 males, 277 females) born between January 2020 and September 2021 and their mothers (Table 1). Among them, 460 children (223 males, 237 females) were born to COVID-19-infected (COVID+) mothers, and 85 (45 males, 40 females) were born to COVID-19-uninfected (COVID-) mothers. Children of COVID+ mothers were more likely to be born during 2021 and be members of KPNC longer than children of COVID− mothers. COVID+ mothers were more likely to be obese before pregnancy than COVID− mothers (Table 1).

Table 1.

Demographic and clinical characteristics of newborns and their mothers in the IMPaCT COVID-19 study, by maternal COVID-19 infection status during pregnancy

Maternal COVID-19 infection during pregnancy, No. (%) P-valuea
Characteristic Yes
(n = 460)		 No
(n = 85)
Child sex, n (%) 0.45
 Female 237 (52) 40 (47)
 Male 223 (48) 45 (53)
Birth season and yearb, n (%) <0.0001
 Winter 2020 33 (7) 8 (9)
 Spring 2020 4 (1) 19 (22)
 Summer 2020 30 (7) 17 (20)
 Fall 2020 54 (12) 15 (18)
 Winter 2021 83 (18) 4 (5)
 Spring 2021 119 (26) 6 (7)
 Summer 2021 112 (24) 11 (13)
 Fall 2021 25 (5) 5 (6)
Birth weight (g), mean (SD) 3388.2 (502.1) 3351.9 (467.6) 0.40
Age (hrs) at bloodspot collection, mean (SD) 23 (7.8) 24.8 (7.6) 0.32
Child insurance type, n (%) 0.22
 Commercial 399 (87) 73 (86)
 Government 56 (12) 9 (11)
 Unknown 5 (1) 3 (4)
Child insurance length (mos), mean (SD) 17.6 (6.8) 19.7 (6.5) <0.0001
Maternal race, n (%) 0.40
 Hispanic 204 (44) 40 (47)
 White 150 (33) 31 (37)
 Otherc 106 (23) 14 (16)
Maternal age (yrs), mean (SD) 34.4 (6) 32.2 (5.6) 0.23
Breastfeeding status, n (%) 0.30
Breastfeeding only 162 (35) 36 (42)
Breastfeeding and formula feeding 296 (64) 48 (56)
Formula feeding only 2 (1) 1 (1)
Delivery method, n (%) 0.30
Natural 343 (75) 68 (80)
Cesarean section 116 (25) 17 (20)
Missing 1 (0) 0 (0)
Gestational age (wks), mean (SD) 38.9 (1.4) 38.6 (2.2) 0.44
Maternal asthma during pregnancy, n (%) 72 (16) 12 (14) 0.72
Maternal gestational diabetes, n (%) 48 (10) 8 (9) 0.78
Maternal pre-pregnancy obesity, n (%) 161 (35) 19 (22) 0.02
Maternal allergy, n (%) 85 (18) 17 (20) 0.74
Maternal autoimmunity during pregnancy, n (%) 41 (9) 3 (4) 0.09
Maternal non-COVID-19 infection during pregnancy, n (%) 301 (65) 47 (55) 0.07
Open in a new tab
a

Group differences were assessed using Chi-squared tests for categorical variables and Wilcoxon’s rank-sum tests for continuous variables.

b

Season defined as: Winter = December to February, Spring = March to May, Summer = June to August, Fall = September to November.

c

Other race includes Asian, Black/African American, Multi-racial, and Pacific Islander/American Indian.

The association of COVID-19 exposure in utero and neonatal immune molecule profiles

Before correcting for multiple comparisons, based on the unadjusted model, children of COVID+ mothers had significantly lower newborn levels of IL-1β (estimate [est.] = −0.41, 95% Cl −0.78, −0.03, p= 0.03) than those of COVID− mothers (Table 2). After adjusting for child sex, birth season and year, age at bloodspot collection, maternal race, pre-pregnancy obesity status, and Luminex plate number (Model 1), children of COVID+ mothers had significantly higher neonatal levels of the growth factor GM-CSF (~31% increase, est.= 0.27, 95% Cl 0.09, 0.46, p= 0.004) and IL-22 (~17% increase, est.= 0.16, 95% Cl 0.01, 0.3, p= 0.04) (Fig. 1, Table 2) compared with children of COVID− mothers. A similar increasing trend was seen for IL-2 (est.= 0.47, 95% Cl −0.02, 0.97, p= 0.06) and IL-4 (est.= 0.09, 95% Cl −0.004, 0.60, p= 0.06) (Fig. 1, Table 2). Further adjustment for additional maternal gestational conditions associated with COVID-19 infection such as maternal gestational diabetes, prepregnancy obesity, asthma, non-COVID-19 infection, maternal autoimmunity, breastfeeding status, and delivery method during pregnancy showed similar results to Model 1 (GM-CSF, ~30% increase, est.= 0.26, 95% Cl 0.07, 0.44; IL-22, ~17% increase, est.= 0.16, 95% Cl 0.01, 0.3) (Model 2, Supplementary Table 3). None of these results survived correction for multiple comparisons (all corrected p> 0.16).

Table 2.

Unadjusted and adjusted associations between neonatal cytokine/chemokine/growth factor concentrations and maternal COVID-19 infection status during pregnancy (n= 545)

Cytokine/Chemokine/growth factor COVID-19 exposed (COVID+) COVID-19 unexposed (COVID−) COVID+ vs COVID−
unadjusted difference		 COVID+ vs COVID−
adjusted difference
Mean (SD)a Mean (SD)a Estimateb (95% Cl) P-value Estimatec (95% Cl) P-value
Pro-inflammatory/pro-inflammatory mediator
sCD40L 489.6 (253.7) 532.1 (357.1) −0.09 (−0.26, 0.08) 0.28 −0.04 (−0.23, 0.14) 0.70
Eotaxin 25.4 (11.0) 25.5 (9.5) −0.01 (−0.1, 0.07) 0.74 0.04 (−0.04, 0.13) 0.36
GROα 40.8 (23.8) 40.0 (22.6) 0.02 (−0.09, 0.13) 0.70 0.06 (−0.06, 0.18) 0.31
IFNγ 1.2 (0.8) 1.1 (0.7) −0.01 (−0.21, 0.19) 0.93 0.13 (−0.08, 0.33) 0.22
IL-1α 227.6 (280.5) 237.4 (226.1) −0.1 (−0.34, 0.15) 0.44 0.01 (−0.27, 0.29) 0.92
IL-1β 2.2 (2.5) 2.9 (3.0) −0.41 (−0.78, −0.03) 0.03 −0.02 (−0.39, 0.35) 0.90
IL-2 0.2 (0.2) 0.1 (0.2) 0.29 (−0.21, 0.79) 0.25 0.47 (−0.02, 0.97) 0.06
IL-3 25.2 (15.5) 24.3 (13.1) −0.01 (−0.18, 0.16) 0.93 0.12 (−0.06, 0.3) 0.20
IL-5 0.3 (0.5) 0.4 (0.5) −0.25 (−0.62, 0.11) 0.17 −0.16 (−0.57, 0.24) 0.43
IL-6 6.5 (46.5) 8.8 (60.0) 0.05 (−0.3, 0.39) 0.78 0.09 (−0.26, 0.45) 0.61
IL-8 45.5 (36.2) 42.8 (29.8) 0.03 (−0.12, 0.18) 0.70 0.08 (−0.08, 0.24) 0.33
IL-12p70 15.3 (8.0) 16.1 (7.3) −0.1 (−0.28, 0.08) 0.26 −0.02 (−0.22, 0.18) 0.85
IL-15 5.8 (2.7) 5.6 (2.3) 0.02 (−0.1, 0.13) 0.80 0.01 (−0.11, 0.13) 0.88
IL-17E/25 99.7 (52.0) 103.1 (61.3) −0.01 (−0.14, 0.12) 0.86 0.05 (−0.09, 0.2) 0.46
IL-18 329.2 (156.4) 343.6 (198.3) −0.02 (−0.15, 0.1) 0.70 0.04 (−0.1, 0.19) 0.56
IP-10 15.0 (7.4) 15.0 (5.9) −0.02 (−0.12, 0.08) 0.68 0.03 (−0.08, 0.14) 0.60
MIP-1b 13.6 (5.8) 14.4 (7.8) −0.04 (−0.14, 0.07) 0.51 0.01 (−0.1, 0.13) 0.80
MCP-1 345.6 (274.2) 344.4 (179.2) −0.04 (−0.2, 0.11) 0.60 0.04 (−0.12, 0.2) 0.62
MCP-3 62.8 (26.0) 63.1 (22.1) −0.01 (−0.1, 0.07) 0.73 0.04 (−0.05, 0.12) 0.40
MDC 68.0 (31.2) 66.7 (23.5) −0.02 (−0.13, 0.09) 0.77 −0.02 (−0.14, 0.1) 0.77
MIG 61.1 (30.0) 57.9 (22.0) 0.01 (−0.1, 0.12) 0.83 0.06 (−0.06, 0.17) 0.33
RANTES 5521.2 (2221.1) 5492.0 (2073.4) −0.01 (−0.13, 0.11) 0.86 0.04 (−0.09, 0.18) 0.50
TNFα 7.6 (4.4) 7.3 (3.2) 0.01 (−0.1, 0.12) 0.83 −0.003 (−0.12, 0.11) 0.96
Growth/proliferation factor
IL-7 6.7 (3.2) 6.5 (2.9) 0.01 (−0.09, 0.11) 0.84 0.06 (−0.05, 0.17) 0.26
EGF 24.9 (20.7) 26.6 (20.8) −0.13 (−0.47, 0.22) 0.47 −0.03 (−0.42, 0.35) 0.87
FGF-2 59.8 (51.6) 66.1 (74.8) −0.09 (−0.28, 0.09) 0.30 −0.09 (−0.28, 0.11) 0.38
FLT-3L 4.2 (1.9) 4.3 (2.2) −0.03 (−0.14, 0.08) 0.59 0.05 (−0.06, 0.15) 0.39
G-CSF 36.4 (407.1) 12.1 (13.3) −0.01 (−0.42, 0.4) 0.96 −0.01 (−0.44, 0.41) 0.95
GM-CSF 39.9 (24.2) 38.7 (23.6) 0.08 (−0.09, 0.26) 0.35 0.27 (0.09, 0.46) 0.004
M-CSF 95.9 (57.1) 91.8 (53.7) 0.04 (−0.11, 0.18) 0.62 0.06 (−0.09, 0.21) 0.43
PDGF AA 847.3 (359.2) 870.3 (444.5) −0.01 (−0.12, 0.1) 0.84 0.03 (−0.08, 0.14) 0.60
PDGF AB/BB 4262.5 (1938.8) 4220.7 (1998.5) 0 (−0.12, 0.11) 0.96 0.01 (−0.11, 0.14) 0.82
TGFα 1.7 (1.2) 1.7 (1.2) −0.02 (−0.21, 0.18) 0.85 0.06 (−0.14, 0.27) 0.55
VEGF 82.8 (43.9) 77.5 (42.0) 0.06 (−0.06, 0.18) 0.30 0.09 (−0.03, 0.22) 0.14
Anti-inflammatory
IL-1RA 177.3 (109.8) 171.5 (121.2) 0.04 (−0.12, 0.21) 0.60 0.03 (−0.12, 0.18) 0.70
Pro- and anti-inflammatory
IL-9 19.2 (8.0) 18.7 (6.8) 0.01 (−0.09, 0.11) 0.79 0.08 (−0.03, 0.18) 0.14
IL-22 45.4 (22.0) 44.4 (20.6) 0.05 (−0.09, 0.19) 0.44 0.16 (0.01, 0.3) 0.04
IL-27 40.8 (41.8) 41.3 (35.9) 0.07 (−0.58, 0.73) 0.83 0.24 (−0.44, 0.88) 0.48
IFNα2 29.9 (14.2) 33.7 (16.7) −0.11 (−0.22, 0) 0.06 −0.01 (−0.12, 0.11) 0.91
Immuno-regulatory/mediator
IL-4 6.9 (2.9) 6.9 (2.8) 0.01 (−0.09, 0.1) 0.92 0.09 (0, 0.19) 0.06
IL-12p40 10.4 (9.3) 9.5 (7.3) 0.06 (−0.25, 0.37) 0.72 0.28 (−0.05, 0.6) 0.10
IL-13 5.0 (4.2) 5.3 (5.0) −0.03 (−0.38, 0.32) 0.88 0.07 (−0.26, 0.39) 0.67
Open in a new tab

Abbreviations: SD, standard deviation; Cl, confidence interval.

a

Means and SDs are reported for raw (non-transformed, normalized to total protein defined by BCA assay; pg/mg) cytokines/chemokines/growth factors.

b

Estimates represent unadjusted differences between newborns born to mothers infected with COVID-19 (n=460) and those born to mothers not infected with COVID-19 (n=85) during pregnancy from simple linear regression models fitted to natural log-transformed neonatal cytokine/chemokine/growth factor concentrations that included a term for maternal COVID-19 infection status during pregnancy.

c

Estimates represent adjusted differences between newborns born to mothers infected with COVID-19 (n=460) and those born to mothers not infected with COVID-19 (n=85) during pregnancy from multiple linear regression models fitted to natural log-transformed neonatal cytokine/chemokine/growth factor concentrations that included a term for maternal COVID-19 infection status during pregnancy and were adjusted for child sex, birth season and year, age at bloodspot collection, maternal race and pre-pregnancy obesity status, and Luminex plate number.

d

Bolding indicates values for which the group differences significantly differed from 0 (p< 0.05 without correcting for multiple comparisons). None of the results survive multiple comparisons adjustment (all p> 0.16).

Figure 1. Differences in neonatal cytokines/chemokines/growth factor levels between newborns exposed to COVID-19 in utero (COVID+) and those who was not exposed to COVID-19 in utero (COVID-).

Figure 1.

Open in a new tab

Estimated effect sizes for newborn differences (COVID+ vs COVID-) in cytokine/chemokine/growth factor concentrations using multiple regression models adjusted for child sex, birth season and year, age at bloodspot collection, maternal race and pre-pregnancy obesity status, and Luminex plate number (Model 1). Bars represent Cohen’s d standardized effect sizes for the COVID+ (n= 460) vs. COVID− (n= 85) difference in concentrations calculated after accounting for covariates and imbalance in the groups. Neonatal cytokines/chemokines/growth factors with higher concentrations in COVID+ newborns are shown on the right of 0; those with higher concentrations in COVID− newborns are shown on the left of 0. Neonatal cytokines/chemokines/growth factors with significant differences between the two groups are marked with an asterisk (unadjusted p< 0.05).

Demographics of newborns and their mothers in the COVID+ group, stratified by NDD diagnosis

A total of 73 children (47 males, 26 females) of the 460 (233 males, 237 females) born to COVID+ mothers were later diagnosed with an NDD. The average follow-up time after birth was 20.7 months among children diagnosed with an NDD (standard deviation [SD]= 7.2, range= 1–38 months) and 27.7 months among children not diagnosed with an NDD (SD= 14.8, range= 3–49 months). Compared to the non-NDD group, children with an NDD were more likely to be male (p= 0.003), have higher birth weight (p= 0.045), and longer KPNC membership (p <0.0001) (Table 3).

Table 3.

Demographic and clinical characteristics of COVID-19 infected mothers during pregnancy and their newborns by child diagnosis of neurodevelopmental disorder

Child diagnosis, No. (%) P-valuea
Characteristic NDD Non-NDD
(n= 73) (n= 387)
Follow-up in monthsb, mean (SD) 20.7 (7.2) 27.7 (14.8) <0.0001
COVID-19 infection trimester, n (%) 0.88
 Trimester 1 15 (20) 88 (23)
 Trimester 2 21 (29) 102 (26)
 Trimester 3 37 (51) 197 (51)
Child sex, n (%) 0.003
 Female 26 (36) 211 (55)
 Male 47 (64) 176 (45)
Birth season and yearc, n (%) 0.72
 Winter 2020 6 (8) 27 (7)
 Spring 2020 1 (1) 3 (1)
 Summer 2020 7 (9) 23 (6)
 Fall 2020 9 (12) 45 (11)
 Winter 2021 16 (22) 67 (17)
 Spring 2021 15 (20) 104 (27)
 Summer 2021 17 (23) 95 (25)
 Fall 2021 2 (3) 23 (6)
Birth weight (g), mean (SD) 3264.3 (574.6) 3411.6 (484.5) 0.045
Age (hrs) at bloodspot collection, mean (SD) 25.1 (8.2) 24.3 (7.4) 0.52
Child insurance type, n (%) 0.57
 Commercial 63 (86) 336 (87)
 Government 10 (14) 46 (12)
 Unknown 0 (0) 5 (1)
Child insurance length (mos), mean (SD) 21.1 (2.9) 16.9 (7.4) <0.0001
Maternal race, n (%) 0.87
 Hispanic 33 (45) 171 (44)
 White 22 (30) 128 (33)
 Otherd 18 (25) 88 (23)
Maternal age (yrs), mean (SD) 32.3 (5.7) 31.4 (5.1) 0.22
Brestfeeding status, n (%) 0.16
Breastfeeding only 19 (26) 143 (37)
Breastfeeding and formula feeding 54 (74) 242 (62)
Formula feeding only 0 (0) 2 (1)
Delivery method, n (%) 0.87
Natural 54 (74) 289 (75)
Cesarean section 19 (26) 97 (25)
Missing 0 (0) 1 (0)
Gestational age (wks), mean (SD) 38.8 (2.1) 39.3 (1.5) 0.07
Maternal pre-pregnancy asthma, n (%) 11 (15) 61 (16) 0.88
Maternal gestational diabetes, n (%) 10 (13) 38 (10) 0.32
Maternal pre-pregnancy obesity, n (%) 26 (36) 135 (35) 0.90
Maternal pre-pregnancy allergy, n (%) 13 (23) 72 (18) 0.33
Maternal autoimmunity, n (%) 8 (11) 33 (9) 0.50
Maternal infection defined as not COVID-19 during pregnancy, n (%) 50 (68) 251 (65) 0.55
Open in a new tab

Abbreviations: NDD, neurodevelopmental disorder diagnosis; SD, standard deviation.

a

Group differences were assessed using Chi-square tests for categorical variables and Wilcoxon’s rank-sum tests for continuous variables.

b

Children were followed from birth to an NDD diagnosis or a censoring event (leaving the KPNC health system or June 30, 2023, whichever occurred first).

c

Season defined as: Winter = December to February, Spring = March to May, Summer = June to August, Fall = September to November.

d

Other race includes Asian, Black/African American, Multi-racial, and Pacific Islander/American Indian.

Association of neonatal immune profiles and NDD diagnosis among newborns of COVID+ mothers

Next, we assessed associations between the neonatal immune profile and NDD risk only among children of COVID+ mothers. Before correcting for multiple comparisons, unadjusted Cox proportional hazards models showed that higher levels of neonatal sCD40L ([HR= 0.71, 95% Cl 0.59, 0.92, p= 0.01), IP-10 (HR= 0.51, 95% Cl 0.3, 0.88, p= 0.01), MIG (HR= 0.61, 95% Cl 0.38, 0.99, p= 0.04), EGF (HR= 0.84, 95% Cl 0.55, 0.97, p= 0.02), FLT-3L (HR= 0.59, 95% Cl 0.35, 0.99, p= 0.045), VEGF (HR= 0.57, 95% Cl 0.36, 0.92, p= 0.02), and IL-22 (HR= 0.63, 95% Cl 0.46, 0.9, p= 0.01) were associated with lower risk of a later diagnosis of an NDD (Table 4). After adjusting for child birth season and year, maternal age, gestational age, and Luminex plate number, higher levels of select cytokines/chemokines/growth factors remained associated with lower risk of a later diagnosis of an NDD (Table 4, Fig. 2). These included neonatal sCD40L (HR= 0.7, 95% Cl 0.57, 0.9, p= 0.009), IP-10 (HR= 0.46, 95% Cl 0.25, 0.83, p= 0.009), MIG (HR= 0.52, 95% Cl 0.3, 0.9, p= 0.02), FLT-3L (HR= 0.45, 95% Cl 0.24, 0.83, p= 0.01), PDGF AB/BB (HR= 0.56, 95% Cl 0.36, 0.99, p= 0.046), VEGF (HR= 0.57, 95% Cl 0.34, 0.98, p= 0.04), IL-22 (HR= 0.49, 95% Cl 0.33, 0.75, p= 0.001), and IL-4 (HR= 0.48, 95% Cl 0.26, 0.93, p= 0.03) (Table 4, Fig. 2). While not reaching statistical significance, similar directional associations were observed for eotaxin (HR= 0.45, 95% Cl 0.2, 1, p= 0.051), IFNγ (HR= 0.74, 95% Cl 0.54, 1.01, p= 0.06), IL-17E/25 (HR= 0.67, 95 Cl 0.44, 1.03, p= 0.06), MCP-3 (HR= 0.51, 95% Cl 0.24, 1.09, p= 0.08), TNFα (HR= 0.61, 95% Cl 0.35, 1.06, p= 0.08), and EGF (HR= 0.86, 95% CI 0.74, 1, p= 0.053, where higher levels of these cytokines/chemokines/growth factors were associated with reduced risk of a later diagnosis of NDD (Table 4, Fig. 2). After correcting for multiple comparisons, in adjusted analyses, only IL-22 remained significantly associated with lower NDD risk (corrected p=0.04).

Table 4.

Unadjusted and adjusted associations between neonatal cytokine/chemokine/growth factor concentrations and child NDD diagnosis among children born to mothers infected with COVID-19 during pregnancy (n= 460)

Cytokine/chemokine/growth factor NDD
Mean (SD)a 		Non-NDD
Mean (SD)a 		NDD vs Non-NDD
HR (95% Cl)b 		P-value NDD vs Non-NDD
aHR (95% Cl)c 		P-valued
Pro-inflammatory/pro-inflammatory mediator
sCD40L 442.89 (255.87) 498.42 (252.63) 0.71 (0.59, 0.92) 0.01 0.7 (0.57, 0.9) 0.009
Eotaxin 24.28 (8.52) 25.63 (11.45) 0.72 (0.36, 1.41) 0.34 0.45 (0.2, 1) 0.051
GROα 37.65 (16.41) 41.4 (24.95) 0.66 (0.4, 1.11) 0.12 0.69 (0.39, 1.22) 0.20
IFNγ 1.07 (0.8) 1.17 (0.8) 0.79 (0.61, 1.04) 0.10 0.74 (0.54, 1.01) 0.06
IL-1α 185.27 (148.13) 235.59 (298.42) 0.88 (0.72, 1.09) 0.24 0.91 (0.72, 1.16) 0.45
IL-1β 2.36 (3.18) 2.15 (2.38) 0.97 (0.84, 1.12) 0.65 0.91 (0.78, 1.08) 0.25
IL-2 0.14 (0.19) 0.17 (0.19) 0.92 (0.83, 1.02) 0.11 0.9 (0.79, 1.03) 0.13
IL-3 23.34 (11.36) 25.52 (16.2) 1.01 (0.79, 1.41) 0.97 1.03 (0.78, 1.52) 0.85
IL-5 0.29 (0.33) 0.34 (0.52) 1.04 (0.9, 1.21) 0.62 1.03 (0.89, 1.21) 0.68
IL-6 3.09 (4.45) 7.14 (50.68) 0.97 (0.83, 1.13) 0.65 0.97 (0.83, 1.15) 0.75
IL-8 43.38 (28.18) 45.89 (37.54) 0.96 (0.66, 1.39) 0.83 1 (0.67, 1.47) 0.98
IL-12p70 14.83 (8.42) 15.34 (7.94) 0.85 (0.7, 1.13) 0.24 0.81 (0.64, 1.11) 0.18
IL-15 5.38 (2.39) 5.87 (2.8) 0.79 (0.54, 1.21) 0.26 0.74 (0.49, 1.2) 0.21
IL-17E/25 94.92 (58.25) 100.65 (50.82) 0.76 (0.52, 1.13) 0.17 0.67 (0.44, 1.03) 0.06
IL-18 315.16 (133.51) 331.83 (160.42) 0.84 (0.53, 1.37) 0.49 0.75 (0.47, 1.27) 0.28
IP-10 13.55 (6.21) 15.33 (7.63) 0.51 (0.3, 0.88) 0.01 0.46 (0.25, 0.83) 0.009
MIP-1b 13.03 (5.11) 13.74 (5.95) 0.73 (0.42, 1.27) 0.26 0.76 (0.42, 1.39) 0.37
MCP-1 309.6 (147.13) 352.35 (291.62) 0.81 (0.57, 1.16) 0.24 0.75 (0.49, 1.17) 0.20
MCP-3 59.92 (20.66) 63.37 (26.86) 0.67 (0.34, 1.31) 0.25 0.51 (0.24, 1.09) 0.08
MDC 64.19 (23.15) 68.71 (32.46) 0.78 (0.47, 1.31) 0.34 0.73 (0.44, 1.24) 0.24
MIG 56.23 (28.14) 61.96 (30.33) 0.61 (0.38, 0.99) 0.04 0.52 (0.3, 0.9) 0.02
RANTES 5212.65 (1609.81) 5579.38 (2315.44) 0.69 (0.48, 1.28) 0.22 0.63 (0.44, 1.15) 0.12
TNFα 7.52 (5.59) 7.61 (4.19) 0.72 (0.43, 1.2) 0.21 0.61 (0.35, 1.06) 0.08
Growth/proliferation factor
IL-7 6.38 (2.7) 6.71 (3.32) 0.88 (0.55, 1.48) 0.63 0.67 (0.4, 1.18) 0.16
EGF 20.64 (19.27) 25.69 (20.84) 0.84 (0.74, 0.97) 0.02 0.86 (0.74, 1) 0.053
FGF-2 57.48 (40.9) 60.24 (53.44) 0.92 (0.7, 1.23) 0.59 0.88 (0.67, 1.18) 0.39
FLT-3L 3.87 (1.79) 4.22 (1.94) 0.59 (0.35, 0.99) 0.045 0.45 (0.24, 0.83) 0.01
G-CSF 20.53 (54.55) 39.42 (443.25) 0.97 (0.86, 1.11) 0.69 0.94 (0.82, 1.08) 0.38
GM-CSF 38.83 (22.36) 40.12 (24.6) 0.95 (0.71, 1.32) 0.76 0.89 (0.64, 1.28) 0.51
M-CSF 94.03 (51.38) 96.23 (58.14) 0.85 (0.59, 1.25) 0.40 0.77 (0.51, 1.2) 0.24
PDGF AA 816.09 (312.85) 853.19 (367.35) 0.75 (0.43, 1.33) 0.32 0.6 (0.34, 1.13) 0.11
PDGF AB/BB 3965.54 (1714.67) 4318.52 (1975.3) 0.63 (0.41, 1.07) 0.08 0.56 (0.36, 0.99) 0.046
TGFα 1.48 (1) 1.68 (1.23) 0.85 (0.67, 1.12) 0.24 1.01 (0.77, 1.36) 0.94
VEGF 74.27 (36.82) 84.38 (44.94) 0.57 (0.36, 0.92) 0.02 0.57 (0.34, 0.98) 0.04
Anti-inflammatory
IL-1RA 163.72 (108.93) 179.88 (109.91) 0.82 (0.6, 1.15) 0.25 0.87 (0.58, 1.34) 0.53
Pro- and anti-inflammatory
IL-9 18.92 (7.32) 19.27 (8.17) 0.9 (0.58, 1.57) 0.70 0.79 (0.51, 1.44) 0.41
IL-22 40.82 (20.35) 46.22 (22.23) 0.63 (0.46, 0.9) 0.01 0.49 (0.33, 0.75) 0.001
IL-27 42.71 (40.09) 40.47 (42.2) 1 (0.92, 1.1) 0.99 1 (0.9, 1.12) 0.97
IFNα2 28.76 (13.05) 30.07 (14.41) 0.86 (0.55, 1.38) 0.52 0.77 (0.45, 1.34) 0.34
Immuno-regulatory/mediator
IL-4 6.53 (2.46) 7.01 (3.02) 0.63 (0.38, 1.1) 0.10 0.48 (0.26, 0.93) 0.03
IL-12p40 9.84 (8.16) 10.51 (9.56) 1 (0.85, 1.2) 0.97 1.05 (0.88, 1.27) 0.58
IL-13 4.32 (3.91) 5.1 (4.28) 0.9 (0.78, 1.04) 0.14 0.89 (0.76, 1.06) 0.18
Open in a new tab

Abbreviations: NDD, neurodevelopmental disorder; SD, standard deviation; HR, hazard ratio; aHR, adjusted hazard ratio; CI, confidence interval.

a

Means and SDs are reported for raw (non-transformed, normalized to their total protein defined by BCA assay; pg/mg).

b

From Cox proportional hazards models using each natural log-transformed neonatal cytokine/chemokine/growth factor concentration as an independent variable and using profile likelihood confidence limits. Children were followed from birth to an NDD diagnosis or a censoring event (leaving the KPNC health system or June 30, 2023, whichever occurred first).

c

From Cox proportional hazards models adjusted for child sex, child birth season and year, maternal age, gestational age, and Luminex plate number using each natural log-transformed neonatal cytokine/chemokine/growth factor concentration as an independent variable and using profile likelihood confidence limits. Children were followed from birth to an NDD diagnosis or a censoring event (leaving the KPNC health system or June 30, 2023, whichever occurred first).

d

Bolding indicates values for which the group differences significantly differed from 0 (p< 0.05 without correcting for multiple comparisons). Only IL-22 survives multiple comparisons adjustment (p= 0.04).

Figure 2. Adjusted associations between cytokines/chemokines/growth factors and later diagnosis of a neurodevelopmental disorder for COVID-19 in utero exposed (n=460).

Figure 2.

Open in a new tab

Hazard ratios and 95% confidence intervals were calculated for a 1 standard deviation increase in each analyte.

Discussion

We investigated prenatal exposure to COVID-19 in relation to levels of immune molecules measured in neonatal bloodspots and associations between neonatal immune profiles and later risk of NDD among newborns with prenatal exposure to COVID-19. Our findings suggest that maternal COVID-19 during pregnancy may be associated with a distinct neonatal immune profile at birth and that a specific immune molecule pattern is linked to a higher risk of an NDD diagnosis among COVID-19-exposed infants.

Although we were not able to discern specific neonatal cytokines/chemokines/growth factors associated to maternal COVID-19 infection during pregnancy after correcting for multiple comparisons, our results prior to correction indicate a potential association between maternal gestational COVID-19 infection and elevated neonatal levels of the growth/immunomodulatory factor GM-CSF, as well as effector T cell-stimulating growth/development factor IL-2, and antibody-regulating IL-4. These results are largely in contrast to the hypothesis that exposure to COVID-19 in utero would result in a largely pro-inflammatory response involving the IL-1 family (e.g., IL-1α, IL-1β, and IL-18), IL-6, TNFα, and IFNγ (Hasanvand, 2022). Instead, our findings suggest that additional molecules associated with immune cell development and function could be equally important for newborns to develop an appropriate immune response following exposure to COVID-19 in utero, providing additional perspectives on this topic. The combination of cytokines observed in our findings prior to correction for multiple comparisons align well with the context of viral infection. For example, GM-CSF plays a key role in inflammation that promotes immune cell maturation and has been frequently implicated in COVID-19 infection (Huang et al., 2020; Lang et al., 2020; Leavis et al., 2022). Recently, GM-CSF has been utilized in clinical trials as vaccine adjuvant to promote immunity against COVID-19 infection (Petrina et al., 2021). While not specifically related to viral infections, GM-CSF is critical in treating and preventing bacterial infections in newborns and infants as it stimulates the production and anti-bacterial function of neutrophils and monocytes (Carr et al., 2003). Thus, an elevation in this critical growth factor could have lasting impacts on neonatal immune function and development. IL-22 is also essential for host immunity in response to viral infections (Albayrak et al., 2022) and increased serum levels of IL-22 have been observed in COVID-19 patients aged between 2 and 16 months (Ahmed Mostafa et al., 2022). The immune molecules measured in the neonatal bloodspots in the current study primarily reflect newborn immune status. However, considering the timeframe (72 hours) of when the bloodspots were collected, the bloodspots may also reflect some aspects of maternal immunity. Therefore, the increased levels of GM-CSF and IL-22 could represent a strategy for both the fetus and the pregnant mother to mount appropriate immunity against COVID-19 infection. Of note, because neonatal immunity at birth is primarily innate and typically characterized by more chemokines than cytokines (Yu et al., 2018), it is interesting to observe the potentially increased levels of the cytokines IL-22, IL-2, and IL-4 in the neonatal bloodspots. One possible explanation may be that fetal “priming” of the immune system in utero may lead to elevated levels of cytokines at birth, whereby maternal infection induces generalized and persistent changes to the fetal immune system with the potential to have life-long consequences (Barrat et al., 2019; Csaba, 2020; Dominguez-Andres et al., 2023; Eades et al., 2022; Levy and Wynn, 2014; Netea et al., 2020; Netea et al., 2016). For example, in a recent study of mothers with COVID-19 infection and their neonates within two weeks of delivery, higher IP-10 levels were observed in neonates born to mothers with higher plasma IP-10 levels (Gee et al., 2021). Moreover, while the current study cannot further address how the neonatal levels observed at birth relate to maternal immune status during COVID-19 infection during pregnancy, Gee et al. further suggests that elevated neonatal inflammatory cytokines/chemokines appear to be more related to the COVID-19 infection status of their mother as the authors observed lower inflammatory markers in neonates born to mothers who had already recovered from the infection (Gee et al., 2021).

Recent studies have reported on the impact of maternal COVID-19 infection on offspring neurodevelopment (Edlow et al., 2023; Firestein et al., 2023; Hessami et al., 2022; Shook et al., 2022). Our study expands upon prior research by demonstrating how maternal COVID-19 infection during pregnancy could potentially influence neurodevelopment through alterations in the neonatal profile. In contrast to several studies suggesting an association between elevated inflammatory-related cytokines with higher risk of NDD (Aureli et al., 2014; Croen et al., 2024; Jiang et al., 2018), our findings suggest rather that elevated levels of cytokine are protective against later development of an NDD among newborns exposed to COVID-19 in utero. This includes pro- and anti-inflammatory IL-22. The pro-inflammatory and/or pro-inflammatory mediators sCD40L, MIG, growth and/or proliferation factors FLT-3L, PDGF AB/BB, VEGF, and immuno-regulatory/mediator IL-4 could also potentially be included, although these associations did not survive correction for multiple comparisons. Our results are not the first to observe that neonatal levels of cytokines are inversely related to NDD outcome (Emanuele et al., 2010; Kim et al., 2022). However, to date literature on IL-22 and NDD is limited to young children with autism (a type of NDD) rather than neonates, where these children are reported to have higher expression of IL-22 and IL-22-expressing CD40+ cells compared to typically developing controls (Ahmad et al., 2017; Aldossari et al., 2023). Elevated peripheral IL-22 has the potential to disrupt the integrity of the blood-brain barrier and to activate glial cells (Chen et al., 2022). Together with our findings, the data suggest that the role of IL-22 in neurodevelopment and brain function is complex and context-dependent. In fact, IL-22 can promote self-renewal of the neural stem cells in the ventricular-subventricular zone (Coronas et al., 2023) and can function as neuroprotective and regenerative in the brain (Mattapallil et al., 2019).

Of importance, our findings do not provide direct evidence for a three-way relationship between COVID-19 infection in utero, higher levels of IL-22, and lower risk of an NDD. This is likely due to the absence of a comparison group (COVID-) with respect to the latter finding. The very limited number of NDD cases among the COVID− newborns precluded examination of the association between COVID-19 exposure in utero, neonatal immunity, and NDD outcome. Nevertheless, our results imply that neonatal immunity may be impacted by maternal COVID-19 infection during pregnancy, and that adequate levels of certain immune molecules are likely necessary for healthy neurodevelopment. This was especially true for newborns exposed to COVID-19 in utero. Further studies with a case-control matched study population are warranted to expand the current findings and to better understand the complex relationship between COVID-19 exposure in utero and newborn immunity as well as the association with a later diagnosis of an NDD.

Interestingly, the prevalence of NDD cases among newborns with COVID-19 exposure in utero in the current study (~ 16%) is higher than recent studies with smaller study populations and shorter follow-up of infants. Ayed et al. reported approximately 10% out of a total of 298 infants born to mothers with COVID-19 infection during pregnancy showed developmental delays at 10–12 months post birth (Ayed et al., 2022). In a study of 57 mother-infant dyads, approximately 13% of children exposed to COVID-19 in utero showed overall developmental delay at 3 months of age (Wu et al., 2021). Finally, in a study with a sample population size similar to that of the current study that included 555 infants with in utero COVID-19 exposure, approximately 8% received an NDD diagnosis during the first 18 months of life (Edlow et al., 2023). A meta-analysis review including the aforementioned studies along with others showed that an average of 12% of infants with gestational exposure to COVID-19 were at risk of an NDD diagnosis (Hessami et al., 2022). Our current findings of NDD risk are based on follow-up of children to 51 months of age, and the prevalence of 16% is consistent with the expected rate of 1 in 6 children with a developmental disability(Prevention, 2024).

Limitations and Strengths

While our study is among the first to examine a large array of neonatal immune molecules following prenatal exposure to COVID-19 as well as demonstrating a higher risk of a later diagnosis of an NDD with reduced neonatal levels of certain cytokines/chemokines/growth factors among children born to COVID+ mothers, several study limitations should be noted. The usage of neonatal bloodspots taken within 72 hours of birth is advantageous for exploration of the immune system in utero and at birth. However, it is challenging to define the origin of the immune molecules detected in neonatal bloodspots as they may contain maternally derived immune molecules. Particularly with respect to breastfeeding and delivery status, both routes could be a source of maternally derived adaptive immunity-related cytokines in the post-natal period. However, one study observed no difference in the cytokine profile in breastmilk of COVID-infected mothers compared to COVID-uninfected mothers (Trofin et al., 2022). Furthermore, neonatal bloodspot immune profiles can only directly define the peripheral immune status and not that of the central nervous system (CNS). Thus, the associations observed herein can only suggest a connection between the peripheral immune system and the brain. Finally, our definition of COVID+ relied on the results of a PCR test only; rapid test results were not available. Therefore, it is possible that some of the COVID− cases were misclassified.

Although we adjusted for major confounding factors related to maternal and newborn immunity, we were unable to investigate several other important factors that could impact findings, such as severity of maternal COVID-19 infection, infection length, infection frequency, the variant of SARS-CoV-2, and level of antibody against COVID-19. Additionally, the relatively small number of children diagnosed with NDD by the end of the follow-up period precluded the investigation of trimester of in utero exposure to COVID-19. In addition, due to the relatively modest number of COVID− study subjects, we had limited statistical power to investigate the association between neonatal immune profiles and NDD risk among children with no in utero exposure to COVID-19. Further, the study was not originally powered to control for multiple comparisons and the statistical power was limited. Thus, we elected to present both raw p-values and the results after corrections for multiple comparisons, offering a comprehensive view of the statistical results. As a result, we recognize the potential for an increased risk of Type I error and advise caution when interpreting p-values that are not corrected for multiple comparisons.

While we recognize the aforementioned limitations, we had the opportunity not only to analyze multiple immune molecules, ranging from cytokines/chemokines to growth factors in a large sample of NBS collected from children with in utero exposure to COVID-19, but also to analyze the same set of immune molecules in a smaller sample of NBS collected from unexposed newborns, allowing us to compare the effects of prenatal COVID-19 exposure on the newborn immune profile. Importantly, even with a limited number of unexposed samples, we were able to suggest differences in immune profiles of newborns born to mothers who tested positive for COVID-19 during pregnancy compared to those born to COVID-19-negative mothers. Further, the diagnoses of NDD were prospectively recorded in medical records, with consideration given to both the duration of child follow-up and the date of the first NDD diagnosis.

Conclusions

In this cohort study, examination of the association between maternal COVID-19 infection during pregnancy and newborn immune profiles and neurodevelopmental outcomes led to two interesting findings: 1) in utero exposure to COVID-19 may be associated with changes in neonatal immune markers, observable at birth, and 2) the immune status among the neonates with in utero COVID-19 exposure could be associated with NDD. Future studies focused on mother-child dyads with inclusion of gestational blood samples from the mothers, placental samples, and neonatal bloodspots would provide further understanding of the relationship between maternal response to COVID-19 infection during pregnancy, neonatal immune signaling molecules and growth factors, and risk of neurodevelopmental disorders in children exposed to viral infection in utero.

Supplementary Material

supplementary tables

Acknowledgement

We would like to thank all the participants in the IMPaCT-COVID study who made this unique and informative study possible. This study was supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD) of the National Institutes of Health under Award Number R01HD095128 and the NICHD-funded IDDRC P50 (P50HD103526). Statistical support was provided by the Biostatistics Core of the NIMH-funded UC Davis Conte Center (P50MH106438). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The biospecimens and/or data used in this study were obtained from the California Biobank Program, (SIS request number 1759). The California Department of Public Health is not responsible for the results or conclusions drawn by the authors of this publication.

Footnotes

Conflict of interest or disclosures

None reported.

Data Sharing Statement

The complete raw cytokine/chemokine/growth factor and NDD status data set is available from the corresponding author upon reasonable request.

References

  1. Ahmad SF, Nadeem A, Ansari MA, Bakheet SA, Attia SM, Zoheir KM, Al-Ayadhi LY, Alzahrani MZ, Alsaad AM, Alotaibi MR, Abd-Allah AR, 2017. Imbalance between the anti- and pro-inflammatory milieu in blood leukocytes of autistic children. Mol Immunol 82, 57–65. [DOI] [PubMed] [Google Scholar]
  2. Ahmed Mostafa G, Mohamed Ibrahim H, Al Sayed Shehab A, Mohamed Magdy S, AboAbdoun Soliman N, Fathy El-Sherif D, 2022. Up-regulated serum levels of interleukin (IL)-17A and IL-22 in Egyptian pediatric patients with COVID-19 and MIS-C: Relation to the disease outcome. Cytokine 154, 155870. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Albayrak N, Orte Cano C, Karimi S, Dogahe D, Van Praet A, Godefroid A, Del Marmol V, Grimaldi D, Bondue B, Van Vooren JP, Mascart F, Corbiere V, 2022. Distinct Expression Patterns of Interleukin-22 Receptor 1 on Blood Hematopoietic Cells in SARS-CoV-2 Infection. Front Immunol 13, 769839. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Aldossari AA, Ansari MA, Nadeem A, Attia SM, Bakheet SA, Al-Ayadhi LY, Alanazi MM, Shahid M, Alwetaid MY, Hussein MH, Ahmad SF, 2023. Upregulation of Inflammatory Mediators in Peripheral Blood CD40(+) Cells in Children with Autism Spectrum Disorder. Int J Mol Sci 24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ames JL, Ferrara A, Avalos LA, Badon SE, Greenberg MB, Hedderson MM, Kuzniewicz MW, Qian Y, Young-Wolff KC, Zerbo O, Zhu Y, Croen LA, 2021. COVID-19 prevalence, symptoms, and sociodemographic disparities in infection among insured pregnant women in Northern California. PLoS One 16, e0256891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Aureli A, Sebastiani P, Del Beato T, Marimpietri AE, Graziani A, Sechi E, Di Loreto S, 2014. Involvement of IL-6 and IL-1 receptor antagonist on intellectual disability. Immunol Lett 162, 124–131. [DOI] [PubMed] [Google Scholar]
  7. Ayed M, Embaireeg A, Kartam M, More K, Alqallaf M, AlNafisi A, Alsaffar Z, Bahzad Z, Buhamad Y, Alsayegh H, Al-Fouzan W, Alkandari H, 2022. Neurodevelopmental outcomes of infants born to mothers with SARS-CoV-2 infections during pregnancy: a national prospective study in Kuwait. BMC Pediatr 22, 319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Ayesa-Arriola R, Castro Quintas A, Ortiz-Garcia de la Foz V, Miguel Corredera M, San Martin Gonzalez N, Murillo-Garcia N, Neergaard K, Fananas Saura L, de Las Cuevas-Teran I, 2023. Exploring the impact of COVID-19 on newborn neurodevelopment: a pilot study. Sci Rep 13, 2983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Barrat FJ, Crow MK, Ivashkiv LB, 2019. Interferon target-gene expression and epigenomic signatures in health and disease. Nat Immunol 20, 1574–1583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Benesova O, 1994. [Perinatal drug administration and risks of functional teratogenic defects]. Cas Lek Cesk 133, 67–70. [PubMed] [Google Scholar]
  11. Boss J, Mukherjee B, Ferguson KK, Aker A, Alshawabkeh AN, Cordero JF, Meeker JD, Kim S, 2019. Estimating Outcome-Exposure Associations when Exposure Biomarker Detection Limits vary Across Batches. Epidemiology 30, 746–755. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Carr R, Modi N, Dore C, 2003. G-CSF and GM-CSF for treating or preventing neonatal infections. Cochrane Database Syst Rev 2003, CD003066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Chen W, Wang J, Yang H, Sun Y, Chen B, Liu Y, Han Y, Shan M, Zhan J, 2022. Interleukin 22 and its association with neurodegenerative disease activity. Front Pharmacol 13, 958022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Coronas V, Arnault P, Jegou JF, Cousin L, Rabeony H, Clarhaut S, Harnois T, Lecron JC, Morel F, 2023. IL-22 Promotes Neural Stem Cell Self-Renewal in the Adult Brain. Stem Cells 41, 252–259. [DOI] [PubMed] [Google Scholar]
  15. Croen LA, Ames JL, Qian Y, Alexeeff S, Ashwood P, Gunderson EP, Wu YW, Boghossian AS, Yolken R, Van de Water J, Weiss LA, 2024. Inflammatory Conditions During Pregnancy and Risk of Autism and Other Neurodevelopmental Disorders. Biol Psychiatry Glob Open Sci 4, 39–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Cron RQ, Goyal G, Chatham WW, 2023. Cytokine Storm Syndrome. Annu Rev Med 74, 321–337. [DOI] [PubMed] [Google Scholar]
  17. Csaba G, 2020. Possible contribution of trained immunity in faulty hormonal imprinting and DOHaD: Review and hypothesis. Acta Microbiol Immunol Hung 67, 143–147. [DOI] [PubMed] [Google Scholar]
  18. Dominguez-Andres J, Dos Santos JC, Bekkering S, Mulder WJM, van der Meer JWM, Riksen NP, Joosten LAB, Netea MG, 2023. Trained immunity: adaptation within innate immune mechanisms. Physiol Rev 103, 313–346. [DOI] [PubMed] [Google Scholar]
  19. Eades L, Drozd M, Cubbon RM, 2022. Hypoxia signalling in the regulation of innate immune training. Biochem Soc Trans 50, 413–422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Edlow AG, Castro VM, Shook LL, Haneuse S, Kaimal AJ, Perlis RH, 2023. Sex-Specific Neurodevelopmental Outcomes Among Offspring of Mothers With SARS-CoV-2 Infection During Pregnancy. JAMA Netw Open 6, e234415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Edlow AG, Castro VM, Shook LL, Kaimal AJ, Perlis RH, 2022. Neurodevelopmental Outcomes at 1 Year in Infants of Mothers Who Tested Positive for SARS-CoV-2 During Pregnancy. JAMA Netw Open 5, e2215787. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Emanuele E, Orsi P, Barale F, di Nemi SU, Bertona M, Politi P, 2010. Serum levels of vascular endothelial growth factor and its receptors in patients with severe autism. Clin Biochem 43, 317–319. [DOI] [PubMed] [Google Scholar]
  23. Firestein MR, Shuffrey LC, Hu Y, Kyle M, Hussain M, Bianco C, Hott V, Hyman SP, Kyler M, Rodriguez C, Tejeda Romero M, Tzul Lopez H, Alcantara C, Amso D, Austin J, Bain JM, Barbosa J, Battarbee AN, Bruno A, Ettinger S, Factor-Litvak P, Gilboa S, Goldman S, Gyamfi-Bannerman C, Maniatis P, Marsh R, Morrill T, Mourad M, Muhle R, Newes-Adeyi G, Noble KG, O’Reilly KC, Penn AA, Reichle L, Sania A, Semenova V, Silver WG, Smotrich G, Tita AT, Tottenham N, Varner M, Welch MG, Zork N, Garey D, Fifer WP, Stockwell MS, Monk C, Dawood F, Dumitriu D, 2023. Assessment of Neurodevelopment in Infants With and Without Exposure to Asymptomatic or Mild Maternal SARS-CoV-2 Infection During Pregnancy. JAMA Netw Open 6, e237396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Gee S, Chandiramani M, Seow J, Pollock E, Modestini C, Das A, Tree T, Doores KJ, Tribe RM, Gibbons DL, 2021. The legacy of maternal SARS-CoV-2 infection on the immunology of the neonate. Nat Immunol 22, 1490–1502. [DOI] [PubMed] [Google Scholar]
  25. Gumusoglu SB, Stevens HE, 2019. Maternal Inflammation and Neurodevelopmental Programming: A Review of Preclinical Outcomes and Implications for Translational Psychiatry. Biol Psychiatry 85, 107–121. [DOI] [PubMed] [Google Scholar]
  26. Han VX, Patel S, Jones HF, Dale RC, 2021. Maternal immune activation and neuroinflammation in human neurodevelopmental disorders. Nat Rev Neurol 17, 564–579. [DOI] [PubMed] [Google Scholar]
  27. Hasanvand A, 2022. COVID-19 and the role of cytokines in this disease. Inflammopharmacology 30, 789–798. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hessami K, Norooznezhad AH, Monteiro S, Barrozo ER, Abdolmaleki AS, Arian SE, Zargarzadeh N, Shekerdemian LS, Aagaard KM, Shamshirsaz AA, 2022. COVID-19 Pandemic and Infant Neurodevelopmental Impairment: A Systematic Review and Meta-analysis. JAMA Netw Open 5, e2238941. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, Zhang L, Fan G, Xu J, Gu X, Cheng Z, Yu T, Xia J, Wei Y, Wu W, Xie X, Yin W, Li H, Liu M, Xiao Y, Gao H, Guo L, Xie J, Wang G, Jiang R, Gao Z, Jin Q, Wang J, Cao B, 2020. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet 395, 497–506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Jiang NM, Cowan M, Moonah SN, Petri WA Jr., 2018. The Impact of Systemic Inflammation on Neurodevelopment. Trends Mol Med 24, 794–804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kim DH, Krakowiak P, Meltzer A, Hertz-Picciotto I, Van de Water J, 2022. Neonatal chemokine markers predict subsequent diagnosis of autism spectrum disorder and delayed development. Brain Behav Immun 100, 121–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lang FM, Lee KM, Teijaro JR, Becher B, Hamilton JA, 2020. GM-CSF-based treatments in COVID-19: reconciling opposing therapeutic approaches. Nat Rev Immunol 20, 507–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Leavis HL, van de Veerdonk FL, Murthy S, 2022. Stimulating severe COVID-19: the potential role of GM-CSF antagonism. Lancet Respir Med 10, 223–224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Levy O, Wynn JL, 2014. A prime time for trained immunity: innate immune memory in newborns and infants. Neonatology 105, 136–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mattapallil MJ, Kielczewski JL, Zarate-Blades CR, St Leger AJ, Raychaudhuri K, Silver PB, Jittayasothorn Y, Chan CC, Caspi RR, 2019. Interleukin 22 ameliorates neuropathology and protects from central nervous system autoimmunity. J Autoimmun 102, 65–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. McCulloch CE, Searle SR, Neuhaus JM, 2008. Generalized, Linear, and Mixed Models, 2nd Edition. Wiley. [Google Scholar]
  37. Nakagawa S, Cuthill IC, 2007. Effect size, confidence interval and statistical significance: a practical guide for biologists. Biol Rev Camb Philos Soc 82, 591–605. [DOI] [PubMed] [Google Scholar]
  38. Netea MG, Dominguez-Andres J, Barreiro LB, Chavakis T, Divangahi M, Fuchs E, Joosten LAB, van der Meer JWM, Mhlanga MM, Mulder WJM, Riksen NP, Schlitzer A, Schultze JL, Stabell Benn C, Sun JC, Xavier RJ, Latz E, 2020. Defining trained immunity and its role in health and disease. Nat Rev Immunol 20, 375–388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Netea MG, Joosten LA, Latz E, Mills KH, Natoli G, Stunnenberg HG, O’Neill LA, Xavier RJ, 2016. Trained immunity: A program of innate immune memory in health and disease. Science 352, aaf1098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Petrina M, Martin J, Basta S, 2021. Granulocyte macrophage colony-stimulating factor has come of age: From a vaccine adjuvant to antiviral immunotherapy. Cytokine Growth Factor Rev 59, 101–110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Prevention C.f.D.C.a., 2024. Developmental Disabilities [Internet]. National Environmental Public Health Tracking Network. [Google Scholar]
  42. Shimizu Y, Sakata-Haga H, Saikawa Y, Hatta T, 2023. Influence of Immune System Abnormalities Caused by Maternal Immune Activation in the Postnatal Period. Cells 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Shook LL, Sullivan EL, Lo JO, Perlis RH, Edlow AG, 2022. COVID-19 in pregnancy: implications for fetal brain development. Trends Mol Med 28, 319–330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Trofin F, Dorneanu OS, Constantinescu D, Nastase EV, Lunca C, Iancu LS, Andrioaie IM, Duhaniuc A, Cianga CM, Pavel-Tanasa M, Anton-Paduraru DT, Cianga P, 2022. Cytokines and Chemokines in Breastmilk of SARS-CoV-2 Infected or COVID-19 Vaccinated Mothers. Vaccines (Basel) 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wu T, Chen L, Wang Y, Shi H, Niu J, Yin X, Li M, Tan C, Jiang H, Zheng D, Wei Y, Zhao Y, Wang X, Qiao J, 2021. Effects of SARS-CoV-2 Infection During Late Pregnancy on Early Childhood Development: A Prospective Cohort Study. Front Pediatr 9, 750012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Yu JC, Khodadadi H, Malik A, Davidson B, Salles E, Bhatia J, Hale VL, Baban B, 2018. Innate Immunity of Neonates and Infants. Front Immunol 9, 1759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zaretsky MV, Alexander JM, Byrd W, Bawdon RE, 2004. Transfer of inflammatory cytokines across the placenta. Obstet Gynecol 103, 546–550. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplementary tables
NIHMS2042041-supplement-supplementary_tables.docx (43.8KB, docx)

Data Availability Statement

The complete raw cytokine/chemokine/growth factor and NDD status data set is available from the corresponding author upon reasonable request.

RESOURCES