Maternal Inflammatory Proteins in Pregnancy and Neurodevelopmental Disorders at Age 10 Years

Tingting Wang; Parisa Mohammadzadeh; Jens Richardt Møllegaard Jepsen; Jonathan Thorsen; Julie Bøjstrup Rosenberg; Cecilie Koldbæk Lemvigh; Nicklas Brustad; Liang Chen; Mina Ali; Rebecca Vinding; Casper-Emil Tingskov Pedersen; María Hernández-Lorca; Birgitte Fagerlund; Birte Y Glenthøj; Niels Bilenberg; Jakob Stokholm; Klaus Bønnelykke; Bo Chawes; Bjørn H Ebdrup

doi:10.1001/jamapsychiatry.2025.0122

. 2025 Mar 12;82(5):514–525. doi: 10.1001/jamapsychiatry.2025.0122

Maternal Inflammatory Proteins in Pregnancy and Neurodevelopmental Disorders at Age 10 Years

Tingting Wang ¹, Parisa Mohammadzadeh ^1,², Jens Richardt Møllegaard Jepsen ^2,⁵, Jonathan Thorsen ¹, Julie Bøjstrup Rosenberg ^1,², Cecilie Koldbæk Lemvigh ², Nicklas Brustad ¹, Liang Chen ¹, Mina Ali ¹, Rebecca Vinding ¹, Casper-Emil Tingskov Pedersen ¹, María Hernández-Lorca ^1,², Birgitte Fagerlund ^5,⁶, Birte Y Glenthøj ^2,⁴, Niels Bilenberg ⁷, Jakob Stokholm ^1,^3,⁸, Klaus Bønnelykke ^1,⁴, Bo Chawes ^1,^4,^✉, Bjørn H Ebdrup ^2,⁴

¹COPSAC, Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of Copenhagen, Copenhagen, Denmark

²Center for Neuropsychiatric Schizophrenia Research (CNSR& Center for Clinical Intervention and Neuropsychiatric Schizophrenia Research (CINS), Mental Health Centre Glostrup, Copenhagen University Hospital – Mental Health Services CPH, Copenhagen, Denmark

³Department of Pediatrics, Slagelse Hospital, Slagelse, Denmark

⁴Faculty of Health and Medical Sciences, Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark

⁵Child and Adolescent Mental Health Center, Copenhagen University Hospital - Mental Health Services CPH, Copenhagen, Denmark

⁶Department of Psychology, University of Copenhagen, Copenhagen, Denmark

⁷Institute of Clinical Research, University of Southern Denmark, Odense, Denmark

⁸Department of Food Science, University of Copenhagen, Frederiksberg C, Denmark

Accepted for Publication: December 23, 2024.

Published Online: March 12, 2025. doi:10.1001/jamapsychiatry.2025.0122

^✉

Corresponding Author: Bo Chawes, MD, PhD, DMSc, Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of Copenhagen, Ledreborg Alle 34, Gentofte 2820, Denmark (chawes@copsac.com).

Author Contributions: Dr Chawes had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Wang and Mohammadzadeh are co–first authors and contributed equally to this work. Drs Chawe and Ebdrup are co–senior authors and contributed equally to this work.

Concept and design: Wang, Mohammadzadeh, Jepsen, Rosenberg, Chen, Fagerlund, Bilenberg, Bønnelykke, Chawes, Ebdrup.

Acquisition, analysis, or interpretation of data: Wang, Mohammadzadeh, Jepsen, Thorsen, Rosenberg, Lemvigh, Brustad, Chen, Ali, Vinding, Pedersen, Hernández-Lorca, Glenthoj, Stokholm, Bønnelykke, Chawes, Ebdrup.

Drafting of the manuscript: Wang, Mohammadzadeh, Pedersen, Ebdrup.

Critical review of the manuscript for important intellectual content: All authors.

Statistical analysis: Wang, Mohammadzadeh, Thorsen, Rosenberg, Chen, Ali, Pedersen.

Obtained funding: Fagerlund, Glenthoj, Bønnelykke, Chawes, Ebdrup.

Administrative, technical, or material support: Wang, Thorsen, Rosenberg, Hernández-Lorca, Bønnelykke.

Supervision: Jepsen, Thorsen, Lemvigh, Brustad, Vinding, Fagerlund, Bilenberg, Bønnelykke, Chawes, Ebdrup.

Conflict of Interest Disclosures: Dr Thorsen reported receiving grants from European Union (EU) H2020 HEDIMED during the conduct of the study and personal fees from AstraZeneca outside the submitted work. Dr Rosenberg reported receiving grants from the Lundbeck Foundation during the conduct of the study. Dr Lemvigh reported receiving grants from Mental Health Center, Glostrup during the conduct of the study. Dr Glenthøj reported receiving grants from The Lundbeck Foundation during the conduct of the study. Dr Bilenberg reported receiving grants from the NOVO Nordic foundation, the Lundbeck foundation, the Tryg foundation, and the Danish Research Fund; and personal fees from NOVO Nordic during the conduct of the study. Dr Ebdrup reported receiving personal fees from Boehringer Ingelheim, Lundbeck Pharma, and Otsuka Pharma Scandinavia outside the submitted work. No other disclosures were reported.

Funding/Support: All funding received by Copenhagen Prospective Studies on Asthma in Childhood (COPSAC) 2010 is listed on www.copsac.com. The Copenhagen Prospective Study on Neuro-Psychiatric Development Study was funded by The Lundbeck Foundation (grant No. R269-2017-5); The Ministry of Health (grant No. 903516); Danish Council for Strategic Research (grant No. 0603-00280B) and The Capital Region Research Foundation have provided core support to the COPSAC research center. TW, MA, and BC are supported by the European Research Council under the EU’s Horizon 2020 research and innovation program (grant No. 946228). The proteomics analyses were funded by the EU’s Horizon 2020 Research and Innovation Programme under grant No. 874864-HEDIMED.

Role of the Funder/Sponsor: 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.

Data Sharing Statement: See Supplement 3.

Additional Contributions: We thank the children and families of the COPSAC 2010 cohort study for all their support and commitment. We acknowledge and appreciate the efforts of the COPSAC research team.

^✉

Corresponding author.

PMCID: PMC11904801 PMID: 40072459

This cohort study evaluates whether maternal inflammatory proteins during pregnancy are associated with the risk of neurodevelopmental disorders and executive functions in middle childhood.

Key Points

Question

Are maternal inflammatory proteins during pregnancy associated with offspring risk of neurodevelopmental disorder in middle childhood?

Findings

In this 10-year follow-up cohort study of 555 mother-child pairs, maternal inflammatory protein profiles during pregnancy were associated with an increased risk of neurodevelopmental disorder in offspring by age 10 years. Key proteins identified included vascular endothelial growth factor A, C-C motif chemokine ligand 3, CD5, interleukin 12B, fibroblast growth factor-23, and monocyte chemoattractant protein-1.

Meaning

These findings suggest that identifying and targeting specific maternal inflammatory proteins during pregnancy could guide future preventive strategies to reduce the burden of childhood neurodevelopmental disorders.

Abstract

IMPORTANCE

Maternal inflammation during pregnancy has been associated with an increased risk of neurodevelopmental disorders (NDDs), such as attention-deficit/hyperactivity disorder (ADHD) and autism, and cognitive deficits in early childhood. However, little is known about the contributions of a wider range of inflammatory proteins to this risk.

OBJECTIVE

To determine whether maternal inflammatory proteins during pregnancy are associated with the risk of NDDs and executive functions (EF) in middle childhood and to identify protein patterns associated with NDDs and EF.

DESIGN, SETTING, AND PARTICIPANTS

This was a 10-year follow-up cohort study of the Danish Copenhagen Prospective Studies on Asthma 2010 mother-child birth cohort, using plasma samples collected at week 24 in pregnancy, where 92 inflammatory proteins were assessed. NDDs and EF were assessed in the offspring at age 10 years, between January 2019 and December 2021. Mother-offspring dyads with available maternal prenatal inflammatory proteins during pregnancy and offspring NDD psychopathology data at follow-up were included. Data analyses took place between December 2023 and August 2024.

EXPOSURES

Levels of 92 inflammatory proteins from panel collected at week 24 during pregnancy.

MAIN OUTCOMES AND MEASURES

Categorical and dimensional psychopathology of NDDs (primary outcome) and EF (secondary outcome).

RESULTS

A total of 555 mothers (mean [SD] age, 32.4 [4.3] years) and their children (285 male [51%]) were included. The principal component analysis showed that higher levels of maternal inflammatory proteins depicted in principal component 1 were associated with a higher risk of any NDD (OR, 1.49; 95% CI, 1.15-1.94; P = .003), particularly autism (OR, 2.76; 95% CI, 1.45-5.63; P = .003) and ADHD with predominantly inattentive presentation (OR, 1.57; 95% CI, 1.05-2.39; P = .03). The single protein analysis showed that 18 of 92 proteins reached false discovery rate (FDR) 5% significance after adjustment. Vascular endothelial growth factor A, C-C motif chemokine ligand, CD5, interleukin 12B, fibroblast growth factor-23, and monocyte chemoattractant protein-1 emerged as top proteins associated with risk of NDDs. The sparse partial least squares approach identified 34 proteins associated with any NDD, and 39 with ADHD with predominantly inattentive presentation. There were no associations with EF after FDR correction.

CONCLUSIONS AND RELEVANCE

The maternal inflammatory proteome during pregnancy was associated with NDDs risks in offspring at age 10 years. Further research is warranted to elucidate the specific pathways involving these proteins during pregnancy that could be targeted with prevention strategies to reduce risk of NDDs in children.

Introduction

Neurodevelopmental disorders (NDDs) rank among the most prevalent mental health disorders in children, with a cumulative prevalence ranging from 16.2% to 21.5%.^1,2 NDDs constitute a significant global health challenge, profoundly impacting the well-being and societal functioning of children and families. NDDs encompass a group of disorders including autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), Tourette syndrome (TS), disorders of intellectual development, and developmental speech and language disorders according to International Classification of Diseases, 11th Revision.³ These disorders typically manifest in early childhood and are commonly associated with impaired executive functions (EF), a transdiagnostic factor relevant to both NDDs and a wide range of psychiatric disorders.^4,5,6

Recent studies^7,8 show a rise in NDD prevalence, with genetics as a known risk factor and growing evidence for prenatal and perinatal environmental factors, especially in genetically susceptible individuals.^9,10 Maternal inflammation during pregnancy is linked to NDDs and EF impairments.^11,12 Elevated maternal interleukin 6 (IL-6) and monocyte chemoattractant protein-1 (MCP-1) levels are associated with higher ADHD risk in offspring.¹¹ Increased neonatal IL-6 and IL-8 levels are linked to higher ASD risk.^11,13 Our cohort previously found higher maternal high-sensitivity C-reactive protein (hs-CRP) levels in pregnancy linked to higher ADHD risk at age 10 years.¹⁴ Impaired EF is also linked to maternal inflammation (IL-6) during pregnancy.^12,15

Pregnancy increases inflammatory proteins,¹⁶ and the maternal immune activation (MIA) hypothesis proposes that intrauterine inflammation can adversely affect fetal brain development.^10,17 Cytokines synthesized by fetal microglia support brain development, and disturbances can increase NDD risk.^17,18,19,20

Previous studies on maternal inflammation and NDDs have mostly investigated single or limited proteins. Such an approach hampers the ability to identify potential patterns of biomarkers that may be crucial in fetal brain development. Advances in proteomics now allow for a more complete inflammatory profile. Emerging research highlights the prognostic potential of newly identified inflammatory proteins in mental health disorders.^21,22 The high comorbidity of NDDs also highlights the need to understand shared causal factors, including inflammation.⁴

In this study, we first evaluated the association between maternal inflammatory proteins during pregnancy and the risk of offspring NDDs, including categorical and dimensional psychopathology. Second, we investigated the association between maternal inflammatory proteins and EF. We hypothesized that specific inflammatory proteins and protein patterns during midgestation would be associated with the risk of NDDs and EF in offspring. We examined 555 mother-child pairs, collecting inflammatory protein data at week 24 of pregnancy and assessing NDDs and EF outcomes in children aged 10 years through association analysis and protein-NDD signature.

Methods

This cohort study was approved by the Scientific Ethics Committees for the Capital Region of Denmark, the Danish Data Protection Agency, and the Danish Health and Medicines Authority. Written and oral informed consent were obtained at enrollment. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline was used.

Study Population

The Copenhagen Prospective Studies on Asthma in Childhood 2010 birth cohort (COPSAC2010) is an ongoing Danish study of 700 deeply phenotyped mother-child pairs, aiming to reduce asthma, eczema, and allergy.^23,24 Study design details are available in the eMethods in Supplement 1.

The Copenhagen Prospective Study on Neuro-Psychiatric Development (COPSYCH) was a 10-year follow-up cohort study based on COPSAC2010 with extensive psychiatric and cognitive assessments of children aged 10 years conducted between January 2019 and December 2021.²⁵ Exclusion criteria in the present study included offspring with a birth weight under 1500 g, gestational age below 28 weeks, and twin B (second born twin) in pairs. The primary outcome of this study was NDDs, including categorical diagnoses and dimensional psychopathology measures, with categorical NDD diagnoses as the main outcome. The secondary outcome was EF.

Outcomes

NDD outcomes included categorical and dimensional psychopathology.²⁶ Research diagnoses followed the ICD-10,^27,28 based on clinical information from the Kiddie Schedule for Affective Disorders and Schizophrenia Present and Lifetime Version²⁹ and other available clinical information.³⁰ Study design details are available in the COPSYCH study protocol.²⁵

We focused on the most prevalent categorical NDD outcomes in this cohort: any NDDs, ADHD, and autism. Dimensional psychopathology of NDDs includes autistic and ADHD symptoms component scores, ADHD Rating Scale (ADHD-RS) and the Social Responsiveness Scale 2 (SRS-2). Parents completed the ADHD-RS and the SRS-2 to assess ADHD symptoms and autistic traits (eMethods in Supplement 1).^31,32,33

EF was assessed using the Behavior Rating Inventory of Executive Function, second edition (BRIEF-2) for everyday EF, and principal component analysis (PCA) on performance-based EF tests (see the eMethods in Supplement 1 for details).^34,35

Covariates

Thirteen covariates were included for the main analyses (eMethods in Supplement 1). Race was included and was defined by the investigators as either White or other based on parental reporting, surname, and visual appearance. We also performed a further stepwise covariate-adjusted analysis in which we added hs-CRP and maternal polygenic risk score (PRS) for ADHD and autism, respectively. The PRS was calculated from previous studies (detailed in eMethods in Supplement 1).^36,37

Inflammatory Protein Profiling

Plasma proteins at week 24 of pregnancy were assessed using the Olink Target 96 inflammation panels (Olink). The information on 92 inflammatory proteins with full names, mean (SD), limit of detection (LOD), percentage of samples below LOD, coefficient of variance, and mean differences between groups is detailed in eTable 1 and eTable 2 in Supplement 2. Details on sample preparation, quality control, and hs-CRP measurement are provided in the eMethods in Supplement 1.

Statistical Analysis

All statistical analyses were conducted using R version 4.2.1 (R Project for Statistical Computing).³⁸ The R functions cor and cor.test calculated Spearman correlation coefficients. The ComplexHeatmap R package visualized protein correlations.³⁹ R packages PCAtools,⁴⁰ caret and mixOmics^41,42 were used in PCA and modeling. We applied a false discovery rate (FDR) of 5% to account for multiple testing. A 2-sided P < .05 was considered significant.

The analysis strategy included PCA, single protein analysis, and sparse partial least squares (sPLS). First, we conducted a data-driven PCA using the principal components against offspring NDDs and EF outcomes. Second, linear and logistic regression were used to investigate associations between each single protein and NDDs and EF outcomes. All analyses were performed unadjusted and adjusted for covariates and additionally adjusted for ADHD and ASD PRS and hs-CRP levels. Third, a supervised sPLS model identified the optimal set of proteins using 10-fold cross-validation repeated 10 times based on the area under the curve (AUC) statistic (eMethods in Supplement 1).

We conducted a sensitivity analysis to determine if the maternal inflammatory profile was specifically associated with NDDs. Associations with non-NDD outcomes were also evaluated. Additionally, to assess the sensitivity of the sPLS model, we performed 1000 permutation tests and a nested 5-fold cross-validation (detailed in the eMethods in Supplement 1).

Results

Baseline Characteristics

Of the original cohort, 604 children (86.3%) participated in the 10-year clinical visit (mean [SD] age, 10.3 [0.36]; 313 male [51.8%]). Two children were excluded due to birth weight below 1500 g or gestational age less than 28 weeks, and 5 twin B children were also excluded. Of 604 children, 593 underwent a psychopathological assessment, and 38 were excluded due to incomplete maternal inflammatory protein profiling, resulting in a final sample size of 555 (mean [SD] age of mothers, 32.4 [4.3] years).

Of the 555 children (285 male [51%]), 86 (15.5%) met the criteria for at least 1 NDD: 64 children fulfilled criteria for an ADHD diagnosis (35 ADHD with combined presentation and 29 ADHD with predominantly inattentive presentation) and 16 for autism. A complete overview of all psychiatric diagnoses and EF outcomes is in eTables 3 and 4 in Supplement 1. Baseline characteristics are summarized in the Table and eResults in Supplement 1.

Table. Baseline Characteristics for Children With vs Without Neurodevelopmental Disorders (NDDs)^a.

Characteristics	Children, No. (%)^b			P value^c
Characteristics	Overall	Without NDDs (n = 469)	With any NDDs (n = 86)	P value^c
Child characteristics
Birth weight, kg, median (IQR)	3.58 (3.23 to 3.93)	3.55 (3.24 to 3.93)	3.65 (3.20 to 3.94)	.35
Gestational age, d, mean (IQR)	281 (274 to 287)	281 (274 to 287)	281 (274 to 288)	.66
Sex
Female	270 (49)	244 (52)	26 (30)	<.001
Male	285 (51)	225 (48)	60 (70)	<.001
Parental characteristics
Alcohol intake during pregnancy	80 (14)	70 (15)	10 (12)	.45
Antibiotics use during pregnancy	201 (36)	168 (36)	33 (38)	.66
Gestational diabetes during pregnancy	14 (2.5)	11 (2.3)	3 (3.5)	.46
Maternal age, y, mean (SD)	32.4 (4.3)	32.3 (4.2)	32.7 (4.6)	.53
Maternal prepregnancy BMI, median (IQR)^d	23.7 (21.7 to 26.6)	23.7 (21.6 to 26.5)	24.2 (21.9 to 27.9)	.23
Maternal smoking during pregnancy	18 (3.2)	13 (2.8)	5 (5.8)	.18
Preeclampsia in third trimester	25 (4.5)	22 (4.7)	3 (3.5)	.78
PRS, median (IQR)
Maternal ADHD PRS	−0.06 (−0.73 to 0.59)	−0.16 (−0.78 to 0.55)	0.33 (−0.33 to 0.87)	.001
Maternal ASD PRS	0.03 (−0.61 to 0.61)	−0.03 (−0.65 to 0.57)	0.18 (−0.45 to 0.79)	.04
Family income, kr^e
Low, <150 000	184 (33)	149 (32)	35 (41)	.19
Medium 150 000-250 000	287 (52)	245 (52)	42 (49)
High, >250 000	83 (15)	74 (16)	9 (10)
Maternal education level
Low	46 (8.3)	36 (7.7)	10 (12)	<.001
Medium	352 (63)	286 (61)	66 (77)
High	157 (28)	147 (31)	10 (12)
Paternal education level
Low	51 (9.4)	43 (9.4)	8 (9.5)	.07
Medium,	338 (63)	277 (61)	61 (73)
High	151 (28)	136 (30)	15 (18)
RCT intervention
Fish oil	289 (52)	251 (54)	38 (44)	.11
Vitamin D	232 (42)	196 (42)	36 (42)	.35
Race^f
White	531 (96)	446 (95)	85 (99)	.15
Other	24 (4)	23 (5)	1(1)	.15
Season of birth
Autumn	115 (21)	98 (21)	17 (20)	.29
Spring	156 (28)	130 (28)	26 (30)
Summer	115 (21)	92 (20)	23 (27)
Winter	169 (30)	149 (32)	20 (23)

Open in a new tab

Abbreviations: ADHD, attention-deficit/hyperactivity disorder; ASD, autism spectrum disorder; BMI, body mass index; PRS, polygenic risk score; RCT, randomized clinical trial.

^{^a}

Any NDD, defined as any neurodevelopmental disorder that was present in our cohort, including pervasive developmental disorders using International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (DF84.0, DF84.5, and DF84.8), other disorders of psychological development (DF88), unspecified disorders of psychological development (DF89), hyperkinetic disorders (DF90.0 and DF90.8), attention deficit disorder without hyperactivity (DF98.8), chronic motor or vocal tics (DF95.1), and Tourette syndrome (DF95.2).

^{^b}

Numbers missing were 1 for alcohol intake during pregnancy, 1 for antibiotic use during pregnancy, 3 for maternal prepregnancy BMI, 74 for maternal ADHD PRS, 74 for maternal ASD PRS, 1 for family income, and 15 for paternal education level.

^{^c}

Fisher exact test, Pearson χ² test, or Wilcoxon rank sum test.

^{^d}

Calculated as weight in kilograms divided by height in meters squared.

^{^e}

kr 1.00 is equal to US $0.14.

^{^f}

Specific racial or ethnic backgrounds were not recorded because of small numbers.

Inflammatory Proteins and Association With NDDs and EF

PCA-Based Inflammatory Protein Profiling and NDDs

The PCA model showed that 30.7% variance was explained by PC1 and 6.4% by PC2 (eResults in Supplement 1). eFigure 1 in Supplement 1 illustrates the proteins with the highest absolute loading scores contributing to PC1 and PC2 (eTable 5 in Supplement 2). PC1, reflecting higher levels of most inflammatory proteins (especially vascular endothelial growth factor A [VEGFA], CD244, CD40, tumor necrosis factor–related apoptosis-inducing ligand [TRAIL], and TNF-related weak inducer of apoptosis [TWEAK]), showed significant associations with higher risk of any NDDs (adjusted odds ratio [aOR], 1.49; 95% CI, 1.15-1.94; P = .003) and autism (aOR, 2.76; 95% CI, 1.45-5.63; P = .003). There was a no association with ADHD (aOR, 1.33; 95% CI, 1.00-1.78; P = .05), ADHD with predominantly inattentive presentation (aOR, 1.57; 95% CI, 1.05-2.39; P = .03), or ADHD with combined presentation (aOR, 1.08; 95% CI, 0.74-1.57; P = .70). Associations remained similar after adjusting for hs-CRP, ADHD PRS, and ASD PRS (Figure 1a; eTable 6 in Supplement 2).

PC1 was associated with the following dimensional psychopathology indices: autistic symptoms component score (adjusted β [aβ] = 0.01; 95% CI, 0.00 to 0.02; P = .02) and autistic traits from SRS-2 (aβ = 0.3; 95% CI, 0.01 to 0.58; P = .04). Additionally, PC1 was not associated with total ADHD symptom load from ADHD-RS (aβ = 0.12; 95% CI, −0.01 to 0.26; P = .08). After adjusting for hs-CRP and PRS, the significant associations remained significant (Figure 1B; eTable 7 in Supplement 2). Interaction analysis between PC1 and selected covariates is shown in eTable 8 in Supplement 2. The correlations between all proteins were shown in eFigure 2 in Supplement 1.

Identifying Proteins Associated With NDDs Using Adjusted Single Protein Analyses

For any NDDs, normalized protein expression (NPX) levels of 18 proteins were associated with a 41% to 200% higher risk (aORs from 1.41; 95% CI, 1.11-1.80; t0 3.0; 95% CI, 1.55-5.96; P < .001) (Figure 2A; eFigure 12 in Supplement 1) after the first step of 13 covariate adjustments and multiple testing correction (eTable 9 in Supplement 2). VEGFA, CD5, signaling lymphocytic activation molecule 1 (SLAMF1), IL-12B, and C-C motif chemokine ligand 3 (CCL3) exhibited the largest ORs. Associations were similar and remained FDR significant after adjustment for hs-CRP, ADHD PRS, and ASD PRS.

Figure 2. — A, Maternal plasma proteins (n tests = 92) associated with any NDDs in offspring. B, Plasma proteins associated with autism. All associations in Figure 2 were calculated using logistic regression adjusted for 13 covariates: prepregnancy maternal body mass index, household income level at birth, maternal and paternal education levels at birth, maternal age, child sex, birth weight, gestational age, smoking during pregnancy, antibiotic use, preeclampsia, maternal diabetes, and alcohol intake during pregnancy. All of the associations shown in Figure 2 are nominally significant, and the P values shown in the figures are FDR corrected. Error bars represent FDR-corrected P values less than .05. The panels shown are a partial representation of all associations in the COPSAC2010 cohort, abbreviated for space. The full Figure is available in eFigure 12 in Supplement 1. CCL indicates C-C motif chemokine ligand; CSF-1, colony stimulating factor-1; CXCL, C-X-C motif chemokine ligand; DNER, delta/notch like EGF repeat containing; FGF, fibroblast growth factor; FLT3L, Fms-related tyrosine kinase 3 ligand; GDNF, glial cell line–derived neurotrophic factor; IFN, interferon; IL, interleukin; LAP-TGF-beta-1, latency-associated peptide-transforming growth factor beta-1; MCP, monocyte chemoattractant protein; MMP, matrix metalloproteinase; NDD, neurodevelopmental disorders; PD-L1, programmed cell death ligand 1; SCF, stem cell factor; SLAMF1, signaling lymphocytic activation molecule 1; TNF, tumor necrosis factor; TNFSRF9, TNF-receptor superfamily 9; TRAIL, TNF–related apoptosis-inducing ligand; TRANCE, TNF-related activation-induced cytokine; TWEAK, TNF-related weak inducer of apoptosis; uPA, urokinase-type plasminogen activator; VEGFA, vascular endothelial growth factor A.

^aIndicates that the P value is nominally significant.

^bIndicates that the P value remains significant after adjusting for FDR 5% in multiple testing.

For overall ADHD and ADHD with predominantly inattentive and combined presentations, no proteins reached FDR-corrected significance in the adjusted analyses (Figure 2B; eFigure 12 in Supplement 1). For autism, 5 proteins including glial cell line–derived neurotrophic factor, C-X-C motif chemokine ligand-9, CCL11, fibroblast growth factor-23 (FGF-23), and FGF19 reached FDR significance after adjustment (aORs from 2.00; 95% CI, 1.30-3.09; to 6.17; 95% CI, 2.01-21.41). After further adjustment for hs-CRP and maternal PRS of ADHD and ASD, respectively, 7 and 25 proteins were FDR significant (eFigure 3 in Supplement 1). Due to large missingness of PRS, analyses without PRS adjustment (464 analyses) for autism were conducted and yielded similar results (eTable 9 in Supplement 2), suggesting this effect was mainly due to sample size reduction.

The levels of 20 FDR-significant proteins associated with NDD diagnoses are shown in Figure 3, grouped by any NDDs with biological interpretation in eTable 10 in Supplement 2. Associations between selected proteins and covariates are shown eFigure 4 in Supplement 1, with sex-stratified associations in eFigure 5 in Supplement 1.

Identifying Key Proteins Associated With NDDs Using sPLS

The sPLS model demonstrated adequate performance, indicating a cross-validated median repeat AUC of 0.59 (95% CI, 0.59-0.60) for risk of any NDDs, 0.63 (95% CI, 0.61-0.65) for ADHD with predominantly inattentive presentation, and 0.64 (95% CI, 0.58-0.66) for risk of autism. We did not achieve an acceptable AUC for ADHD (AUC, 0.51; 95% CI, 0.45-0.56). The model selected 34 proteins that were jointly associated with risk of any NDDs, 39 proteins associated with ADHD with predominantly inattentive presentation, and 49 proteins associated with autism (Figure 4; eFigure 13 in Supplement 1). Permutation testing gave the sPLS model for any NDDs a statistically significant P value of .04, while the ADHD inattentive model and the autism model were not significant. Figure 4 and eFigure 13 in Supplement 1 display the loading scores of selected proteins for each diagnosis.

Figure 4. — Representation of the proteins related to risk of any NDD (34) (A) or ADHD predominantly inattentive presentation (39) (B) contributing to the optimal with set of loadings for the 10-times repeated 10-fold cross-validation sPLS model (n = 100) using the entire sample (n = 555). Bars depict the median across repeats ± SD. Proteins are sorted from top to bottom based on loading scores. The panels shown have been abbreviated for space. The full Figure depicting representation of proteins related to risk of autism, as well as box plots representing the AUC from repeated 10-fold cross-validation of an sPLS model for risk of any NDDs, ADHD with predominantly inattentive presentation, or autism, can be found in eFigure 13 in Supplement 1. ADHD indicates attention-deficit/hyperactivity disorder; NGF, nerve growth factor; CCL, C-C motif chemokine ligand; CSF-1, colony stimulating factor-1; CXCL, C-X-C motif chemokine ligand; CX3CL, fractalkine; FGF, fibroblast growth factor; FLT3L, Fms-related tyrosine kinase 3 ligand; IFN, interferon; IL, interleukin; LAP-TGF-beta-1, latency-associated peptide-transforming growth factor beta-1; MCP, monocyte chemoattractant protein; MMP, matrix metalloproteinase; SLAMF1, signaling lymphocytic activation molecule 1; TGF-alpha, transforming growth factor-alpha; TNF, tumor necrosis factor; TNFSF, TNF superfamily; TNFSRF9, TNF-receptor superfamily 9; TRAIL, TNF–related apoptosis-inducing ligand; TRANCE, TNF-related activation-induced cytokine; TSLP, thymic stromal lymphopoietin; uPA, urokinase-type plasminogen activator; VEGFA, vascular endothelial growth factor A.

Sensitivity Analysis

We found no association between PC1 or single protein and nonneurodevelopmental psychiatric disorders or non-NDD-related outcomes such as fractures and molar incisor hypomineralization (eTable 11 Supplement 2 and eResults in Supplement 1). PC2 was not associated with any NDD outcomes (eFigure 6 in Supplement 1). The sPLS model yielded a comparable performance for any NDDs and ADHD with predominantly inattentive presentation using a nested 5-fold cross-validation model (eResults and eFigure 7 in Supplement 1). The similar associations between PC1 and NDDs, adjusted for plateID or excluding proteins below a certain LOD percentage are shown in eFigures 8 and 9 in Supplement 1.

Inflammatory Protein Profiling and Association With EF

Single protein analysis showed associations between proteins and all 4 everyday EF indices from BRIEF-2, but none were FDR significant. There was no association between PC1 and any EF outcomes (eFigure 10 in Supplement 1 and eTables 12-13 in Supplement 2)

Discussion

This study comprehensively analyzed the maternal inflammatory proteins at week 24 of pregnancy in relation to NDDs and EF in offspring aged 10 years among 555 mother-child pairs from the COPSAC2010 cohort. Associations between proteins and outcomes were examined using PCA, single protein analyses, and sPLS model. We observed consistent findings across different approaches. The maternal inflammatory protein profile depicted in PC1 was associated with a higher risk of NDDs, including autism and ADHD with predominantly inattentive presentation. VEGFA, CD5, IL-12B, CCL3, FGF-23, and MCP-1 were the top proteins associated with any NDDs in both single protein analyses and the sPLS model, where VEGFA and CCL3 also had the highest loadings in the PC1 in PCA. These data demonstrate an association between the maternal inflammatory proteins in midpregnancy and higher risk of NDDs in the offspring, contrasting no associations with childhood EF.

In our cohort, 11% of children met the criteria for ADHD, and 15% for any NDDs,¹⁴ which is higher than what has been shown in most population-based studies. This may be due to an increase in ADHD diagnoses during the COVID-19 pandemic.⁴³ However, a Danish study⁴⁴ with clinically assessed children aged 11 years reported a similar ADHD incidence (10.9%), and a meta-analysis⁴⁵ found a peak prevalence of 11.4%.

Maternal inflammatory protein profile, represented by PC1, was significantly associated with increased risk of NDDs by age 10, especially autism and ADHD with predominantly inattentive presentation. Our findings also showed that maternal inflammation during pregnancy was not associated with overall ADHD or ADHD symptom severity.

Single protein analysis identified several proteins, including VEGFA, CD5, SLAMF1, IL-12B, CCL3, latency-associated peptide-transforming growth factor beta-1 (LAP-TGF-beta-1, FGF-23, MCP-1, MCP-2, and CCL23, as greatly associated with any NDDs, ranked by their ORs. LAP-TGF-beta-1, MCP-1, IL-12B, MCP-2, and CCL3 participate in the cytokine-cytokine receptor interaction pathway; IL-12B and CCL3 also play a role in the toll-like receptor (TLR) signaling pathway.⁴⁶ LAP-TGF-beta-1, MCP-1, IL-12B, CD5, SLAMF1, and CCL3 play roles in leukocyte activation.⁴⁶ High plasma CD5 levels in children have been linked to ASD pathophysiology, potentially contributing to autoimmunity.^47,48 Increased maternal MCP-1 levels have been associated with a higher risk of ADHD in offspring aged 4 to 6 years (62 offspring).¹¹ Elevated MCP-1 in neonatal dried blood samples has also been associated with ASD in previous studies.⁴⁹ We found an FDR-significant association between maternal MCP-1 and any NDDs, but only nominally significant association with ADHD.

The supervised sPLS models identified optimal sets of inflammatory proteins linked to increased NDD risk. IL-6 was identified as the most important protein associated with any NDDs. Elevated IL-6 levels in the mother during pregnancy, in newborns, and in children have been linked to an increased risk of ASD and ADHD in offspring.^11,13,50 Additionally, IL-6 and hs-CRP are highly correlated, and our previous study found hs-CRP associated with ADHD.^14,51

VEGFA, CD5, IL-12B, CCL3, FGF-23, and MCP-1 emerged as top proteins associated with any NDDs in both sPLS (top 10) and single protein analysis (top 10 by OR). Except for MCP-1, previously associated with ADHD risk, our study is the first to show associations between these proteins during pregnancy and NDD outcomes in offspring. All FDR-significant proteins associated with any NDDs were also identified by the final sPLS models, indicating high concordance.

VEGFA and CCL3 were the top proteins associated with NDDs across all 3 approaches. VEGFA is crucial for angiogenesis and nervous tissue growth during embryonic development.^52,53 One study⁵⁴ shows low average VEGFA levels in healthy pregnant women, suggesting low VEGFA levels are sufficient for embryogenesis. Elevated VEGFA levels are observed in ASD brains and during acute episodes of major depressive and bipolar disorders.^54,55,56 A prior study⁵⁷ suggests VEGF mediates the adverse effects of maternal immune activation on fetal central nervous system (CNS) development in mice. Increased VEGFA expression is reported in smoking or women with high body mass index (BMI) during pregnancy.^16,58 CCL3, a chemokine crucial for immune cell activation and neurodevelopment,⁵⁹ induces VEGFA expression, leading to endothelial cell migration and tube formation.⁶⁰ CCL3 is proposed as a potential biomarker, with elevated levels found in autistic children (N = 77),⁵⁹ but we are the first we know of to show an association between maternal CCL3 levels and other NDD outcomes in children.

We used both unsupervised (PCA) and supervised (univariable single protein analysis and multivariable sPLS) approaches to reinforce our findings. We ultimately conclude that a broader inflammation profile, as depicted in PC1, is likely associated with a higher risk of NDDs, and VEGFA, CD5, IL-12B, CCL3, FGF-23, and MCP-1 are the top proteins identified through supervised approaches.

TLRs are pattern recognition receptors on immune and CNS cells, activated by pathogen- and damage-associated molecular patterns (DAMPs).⁶¹ TLRs represent a potential unified pathway that links maternal inflammation to immune disruptions in offspring brain development.⁶² Risk factors such as infections, obesity, preeclampsia, depression, and asthma could lead to an increased release of DAMPs.^63,64,65,66 Chronic inflammatory conditions like diabetes and preeclampsia may also stimulate TLRs.^64,65,67 These risk factors may converge on TLR pathways, affecting immunity in maternal blood, the placenta, and the fetal brain, increasing NDD risk in offspring. This study found associations between TLR pathway proteins (CCL3, IL-12B, and IL-6) and risk factors like maternal BMI. The cytokine-cytokine receptor interaction pathway and chemokine signaling pathway have previously been reported as associated with ASD in children.⁶⁸ Hence, long-term follow-up could be of interest for offspring of mothers with pregnancy-related inflammatory events to monitor for NDD signs.

NDDs, including ADHD and autism, are highly heritable yet genetically heterogeneous, with evidence suggesting a shared genetic cause from common variants.⁶⁹ Adjusting for PRSs for ADHD and ASD strengthened the associations, possibly reflecting the impact of inflammatory environments alongside genetic predispositions. However, due to the significant missingness in PRS scores, the observed effects may also be attributed to the sample size reduction, which could have decreased the power of the association analysis. Factors contributing to MIA include obesity, gestational diabetes, smoking, alcohol exposure, preeclampsia, depression, allergies, asthma, and stress.^9,62,70,71 CCL3, VEGFA, and colony stimulating factor-1 were significantly associated with maternal BMI in a previous study.¹⁶ Obesity and pregnancy both contribute to chronic inflammation, and their combined response can be particularly detrimental to mother and fetus.⁷²

Nominally significant results for EF were found in individual protein analyses but did not pass the 5% FDR. IL-6 has been associated with EF in the general population, but earlier studies focused on single proteins with fewer predictors, lowering the FDR threshold.¹⁵ Our measurements suggest maternal inflammation during pregnancy is not associated with childhood EF. Some causal models of ADHD have highlighted other cognitive deficits than EF deficits (ie, there is a neuropsychological heterogeneity in ADHD).^73,74 Thus, not all individuals with ADHD may exhibit EF deficits, which could impact the results.

Strengths and Limitations

The study benefits from a large participant pool from the general Danish population, ensuring broad representation and applicability. The comprehensive data collection and the high retention rate within the COPSAC cohort allows for robust adjustment of numerous covariates, controlling potential confounding factors and highlighting contributors to inflammation during pregnancy. Another strength is the thorough neuropsychiatric assessment using interviews, performance-based measures, and questionnaires.

However, several limitations should be considered. This study only used a single time point assessment of the maternal inflammatory protein profile during pregnancy (ie, we did not assess the maternal inflammatory proteins before or after pregnancy). The cohort’s sample size, predominantly White demographic, higher average education level, and missing data may limit the generalizability of the findings. Additionally, uncertainty in outcome variables, such as ADHD-RS, SRS-2, and EF items in the parent-completed questionnaire, should be approached with caution. We also estimated the genetic heritability of NDDs using parental PRSs but lacked data on parental mental health and cognitive functions, which is important given the heritability of NDDs and EF. PRSs explain only a small portion of the variance captured in family history of mental disorders. Furthermore, the sensitivity analysis using unbiased internal validation, which, though less robust than external replication, was the best available option for this study. Additionally, our findings are limited to the 92 inflammation-related proteins in the panel, potentially excluding other relevant proteins.

Conclusions

This study found an association between the maternal inflammatory plasma proteome during pregnancy and increased risks of NDDs in children by age 10 years. VEGFA, CD5, IL-12B, CCL3, FGF-23, and MCP-1 were identified as key proteins, suggesting that maternal inflammation negatively impacts prenatal brain development. Further research is required to pinpoint the specific inflammatory pathways involving these proteins, aiming to develop targeted prevention strategies that could enhance neurodevelopmental outcomes in children.

Supplement 1.

eMethods.

eResults.

eFigure 1. Principal Component Analysis of 92 Proteins Based on the NDD Diagnoses Within the Data Set

eFigure 2. Spearman Rank Correlation Heatmap of All 92 Proteins and hs-CRP

eFigure 3. Maternal Inflammatory Protein (Single Protein) Associations Adjusted by Different Covariates

eFigure 4. Association Between 6 Selected Proteins and Selected Covariates

eFigure 5. Association Analysis of Inflammatory Proteins With NDDs in Male and Female Offspring

eFigure 6. Association Between PC2 Scores and Risk of Psychiatric Disorders

eFigure 7. sPLS Loading Scores From the Nested 5-Fold Cross-Validation

eFigure 8. Association Between PC1 Scores and Risk of Psychiatric Disorders Adjusted for plateID

eFigure 9. Association Between PC1 Scores Derived From 72/69 Proteins and Risk of Psychiatric Disorders

eFigure 10. Maternal Inflammatory Protein Profile PC1 Associated With EF in Offspring

eFigure 11. Spearman Rank Correlation Between EF and ADHD Symptoms Severity

eFigure 12. Maternal single plasma protein associations in COPSAC2010 cohort (N = 534)

eFigure 13. sPLS models and NDDs

eTable 3. Prevalence of Research Diagnoses and Groups in the Cohort (Supplement 1)

eTable 4. Cognitive Outcomes (Supplement 1)

eReferences.

jamapsychiatry-e250122-s001.pdf^{(882.5KB, pdf)}

Supplement 2.

eTable 1. Full List of 92 Studied Inflammatory Proteins

eTable 2. Levels of proteins between different outcomes group

eTable 5. PCA Loading Scores of 92 Proteins

eTable 6. PC1 and NDD Categorical Psychopathology

eTable 7. Association Between PC1 and NDD Dimensional Psychopathology

eTable 8. Interaction Analysis Between Different Factors and PC1 in Relation to NDDs

eTable 9. Odds Ratios

eTable 10. Enrichment Analysis of Selected Proteins That Pass FDR in Single Protein Association

eTable 11. Sensitivity Analysis of Other Non-NDD Outcomes

eTable 12. Association Between Single Protein and EF

eTable 13. Association Between PC Score and EF

jamapsychiatry-e250122-s002.xlsx^{(266KB, xlsx)}

Supplement 3.

Data Sharing Statement

jamapsychiatry-e250122-s003.pdf^{(12.3KB, pdf)}

References

1.Zablotsky B, Black LI, Maenner MJ, et al. Prevalence and trends of developmental disabilities among children in the United States: 2009-2017. Pediatrics. 2019;144(4):e20190811. doi: 10.1542/peds.2019-0811 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Francés L, Ruiz A, Soler CV, et al. Prevalence, comorbidities, and profiles of neurodevelopmental disorders according to the DSM-5-TR in children aged 6 years old in a European region. Front Psychiatry. 2023;14:1260747. doi: 10.3389/fpsyt.2023.1260747 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.World Health Organization . International Classification of Diseases, 11th Revision (ICD-11) . World Health Organization; 2018. [Google Scholar]
4.Sokolova E, Oerlemans AM, Rommelse NN, et al. A causal and mediation analysis of the comorbidity between attention deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD). J Autism Dev Disord. 2017;47(6):1595-1604. doi: 10.1007/s10803-017-3083-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Morris-Rosendahl DJ, Crocq MA. Neurodevelopmental disorders-the history and future of a diagnostic concept. Dialogues Clin Neurosci. 2020;22(1):65-72. doi: 10.31887/DCNS.2020.22.1/macrocq [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Zelazo PD. Executive function and psychopathology: a neurodevelopmental perspective. Annu Rev Clin Psychol. 2020;16(1):431-454. doi: 10.1146/annurev-clinpsy-072319-024242 [DOI] [PubMed] [Google Scholar]
7.Visser SN, Bitsko RH, Danielson ML; Centers for Disease Control and Prevention (CDC) . Increasing prevalence of parent-reported attention-deficit/hyperactivity disorder among children—United States, 2003 and 2007. MMWR Morb Mortal Wkly Rep. 2010;59(44):1439-1443. [PubMed] [Google Scholar]
8.Rutter M. Incidence of autism spectrum disorders: changes over time and their meaning. Acta Paediatr. 2005;94(1):2-15. doi: 10.1111/j.1651-2227.2005.tb01779.x [DOI] [PubMed] [Google Scholar]
9.Carlsson T, Molander F, Taylor MJ, Jonsson U, Bölte S. Early environmental risk factors for neurodevelopmental disorders—a systematic review of twin and sibling studies. Dev Psychopathol. 2021;33(4):1448-1495. doi: 10.1017/S0954579420000620 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Knuesel I, Chicha L, Britschgi M, et al. Maternal immune activation and abnormal brain development across CNS disorders. Nat Rev Neurol. 2014;10(11):643-660. doi: 10.1038/nrneurol.2014.187 [DOI] [PubMed] [Google Scholar]
11.Gustafsson HC, Sullivan EL, Battison EAJ, et al. Evaluation of maternal inflammation as a marker of future offspring ADHD symptoms: a prospective investigation. Brain Behav Immun. 2020;89:350-356. doi: 10.1016/j.bbi.2020.07.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Rudolph MD, Graham AM, Feczko E, et al. Maternal IL-6 during pregnancy can be estimated from newborn brain connectivity and predicts future working memory in offspring. Nat Neurosci. 2018;21(5):765-772. doi: 10.1038/s41593-018-0128-y [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Heuer LS, Croen LA, Jones KL, et al. An exploratory examination of neonatal cytokines and chemokines as predictors of autism risk: the early markers for autism study. Biol Psychiatry. 2019;86(4):255-264. doi: 10.1016/j.biopsych.2019.04.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Rosenberg JB, Richardt Møllegaard Jepsen J, Mohammadzadeh P, et al. Maternal inflammation during pregnancy is associated with risk of ADHD in children at age 10. Brain Behav Immun. 2024;115:450-457. doi: 10.1016/j.bbi.2023.10.023 [DOI] [PubMed] [Google Scholar]
15.Graham AM, Rasmussen JM, Rudolph MD, et al. Maternal systemic interleukin-6 during pregnancy is associated with newborn amygdala phenotypes and subsequent behavior at 2 years of age. Biol Psychiatry. 2018;83(2):109-119. doi: 10.1016/j.biopsych.2017.05.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Hedman AM, Lundholm C, Andolf E, Pershagen G, Fall T, Almqvist C. Longitudinal plasma inflammatory proteome profiling during pregnancy in the Born into Life study. Sci Rep. 2020;10(1):17819. doi: 10.1038/s41598-020-74722-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Graham AM, Marr M, Buss C, Sullivan EL, Fair DA. Understanding vulnerability and adaptation in early brain development using network neuroscience. Trends Neurosci. 2021;44(4):276-288. doi: 10.1016/j.tins.2021.01.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Meyer U, Feldon J, Schedlowski M, Yee BK. Immunological stress at the maternal-foetal interface: a link between neurodevelopment and adult psychopathology. Brain Behav Immun. 2006;20(4):378-388. doi: 10.1016/j.bbi.2005.11.003 [DOI] [PubMed] [Google Scholar]
19.Bilbo SD, Block CL, Bolton JL, Hanamsagar R, Tran PK. Beyond infection—maternal immune activation by environmental factors, microglial development, and relevance for autism spectrum disorders. Exp Neurol. 2018;299(Pt A):241-251. doi: 10.1016/j.expneurol.2017.07.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Han VX, Patel S, Jones HF, Dale RC. Maternal immune activation and neuroinflammation in human neurodevelopmental disorders. Nat Rev Neurol. 2021;17(9):564-579. doi: 10.1038/s41582-021-00530-8 [DOI] [PubMed] [Google Scholar]
21.Sokolowska I, Ngounou Wetie AG, Wormwood K, Thome J, Darie CC, Woods AG. The potential of biomarkers in psychiatry: focus on proteomics. J Neural Transm (Vienna). 2015;122(S1)(suppl 1):S9-S18. doi: 10.1007/s00702-013-1134-6 [DOI] [PubMed] [Google Scholar]
22.Hsiao EY, Patterson PH. Placental regulation of maternal-fetal interactions and brain development. Dev Neurobiol. 2012;72(10):1317-1326. doi: 10.1002/dneu.22045 [DOI] [PubMed] [Google Scholar]
23.Bisgaard H, Vissing NH, Carson CG, et al. Deep phenotyping of the unselected COPSAC2010 birth cohort study. Clin Exp Allergy. 2013;43(12):1384-1394. doi: 10.1111/cea.12213 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Chawes BL, Bønnelykke K, Stokholm J, et al. Effect of vitamin D3 supplementation during pregnancy on risk of persistent wheeze in the offspring: a randomized clinical trial. JAMA. 2016;315(4):353-361. doi: 10.1001/jama.2015.18318 [DOI] [PubMed] [Google Scholar]
25.Mohammadzadeh P, Rosenberg JB, Vinding R, et al. Effects of prenatal nutrient supplementation and early life exposures on neurodevelopment at age 10: a randomised controlled trial - the COPSYCH study protocol. BMJ Open. 2022;12(2):e047706. doi: 10.1136/bmjopen-2020-047706 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Hagstrøm J, Nielsen T, Sørensen ME, et al. Dimensional profiling of psychopathology in children and adolescents based on the K-SADS-PL and an analysis of the construct validity of two ADHD symptom dimensions. Nord J Psychiatry. 2024;78(1):71-78. doi: 10.1080/08039488.2023.2262997 [DOI] [PubMed] [Google Scholar]
27.Bell CC. DSM-IV: diagnostic and statistical manual of mental disorders. JAMA. 1994;(272):828-829. doi: 10.1001/jama.1994.03520100096046 [DOI] [Google Scholar]
28.World Health Organization . The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Reprinted. World Health Organization; 2009. [Google Scholar]
29.Kaufman J, Birmaher B, Brent D, et al. Schedule for affective disorders and schizophrenia for school-age children-present and lifetime version (K-SADS-PL): initial reliability and validity data. J Am Acad Child Adolesc Psychiatry. 1997;36(7):980-988. doi: 10.1097/00004583-199707000-00021 [DOI] [PubMed] [Google Scholar]
30.Kaufman J, Schweder AE. The Schedule for Affective Disorders and Schizophrenia for School-Age Children: Present and Lifetime version (K-SADS-PL). In: Comprehensive Handbook of Psychological Assessment, Vol. 2: Personality Assessment. John Wiley & Sons, Inc; 2004:247-255. [Google Scholar]
31.DuPaul GJ, Power TJ, Anastopoulos AD, Reid R. ADHD Rating Scale—IV: Checklists, Norms, and Clinical Interpretation. The Guilford Press; 1998. [Google Scholar]
32.Makransky G, Bilenberg N. Psychometric properties of the parent and teacher ADHD Rating Scale (ADHD-RS): measurement invariance across gender, age, and informant. Assessment. 2014;21(6):694-705. doi: 10.1177/1073191114535242 [DOI] [PubMed] [Google Scholar]
33.Constantino J, Gruber C. Social Responsiveness Scale. 2nd ed. Western Psychological Services; 2012. [Google Scholar]
34.Toplak ME, West RF, Stanovich KE. Practitioner review: do performance-based measures and ratings of executive function assess the same construct? J Child Psychol Psychiatry. 2013;54(2):131-143. doi: 10.1111/jcpp.12001 [DOI] [PubMed] [Google Scholar]
35.Gioia GA, Isquith PK, Guy SC, Kenworthy L. Behavior Rating Inventory of Executive Function 2 (BRIEF2). Second Edition. 2015. Accessed February 5, 2025. https://www.parinc.com/products/pkey/24
36.Grove J, Ripke S, Als TD, et al. ; Autism Spectrum Disorder Working Group of the Psychiatric Genomics Consortium; BUPGEN; Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium; 23andMe Research Team . Identification of common genetic risk variants for autism spectrum disorder. Nat Genet. 2019;51(3):431-444. doi: 10.1038/s41588-019-0344-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Demontis D, Walters GB, Athanasiadis G, et al. ; ADHD Working Group of the Psychiatric Genomics Consortium; iPSYCH-Broad Consortium . Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains. Nat Genet. 2023;55(2):198-208. doi: 10.1038/s41588-022-01285-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.R: The R Project for Statistical Computing. Accessed January 10, 2019. Accessed February 5, 2025. https://www.r-project.org/
39.Gu Z. Complex heatmap visualization. Imeta. 2022;1(3):e43. doi: 10.1002/imt2.43 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Blighe K. kevinblighe/PCAtools. 2024; https://github.com/kevinblighe/PCAtools
41.Rohart F, Gautier B, Singh A, Lê Cao KA. mixOmics: an R package for 'omics feature selection and multiple data integration. PLoS Comput Biol. 2017;13(11):e1005752. doi: 10.1371/journal.pcbi.1005752 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Kuhn M. Building predictive models in R using the caret package. J Stat Softw. 2008;28:1-26. doi: 10.18637/jss.v028.i0527774042 [DOI] [Google Scholar]
43.Bliddal M, Rasmussen L, Andersen JH, et al. Psychotropic medication use and psychiatric disorders during the COVID-19 pandemic among danish children, adolescents, and young adults. JAMA Psychiatry. 2023;80(2):176-180. doi: 10.1001/jamapsychiatry.2022.4165 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Gregersen M, Søndergaard A, Brandt JM, et al. Mental disorders in preadolescent children at familial high-risk of schizophrenia or bipolar disorder—a four-year follow-up study: The Danish High Risk and Resilience Study, VIA 11. J Child Psychol Psychiatry. 2022;63(9):1046-1056. doi: 10.1111/jcpp.13548 [DOI] [PubMed] [Google Scholar]
45.Willcutt EG. The prevalence of DSM-IV attention-deficit/hyperactivity disorder: a meta-analytic review. Neurotherapeutics. 2012;9(3):490-499. doi: 10.1007/s13311-012-0135-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.STRING . functional protein association networks. 2024. Accessed February 5, 2025. https://string-db.org/
47.Desoky T, Hassan MH, Fayed HM, Sakhr HM. Biochemical assessments of thyroid profile, serum 25-hydroxycholecalciferol and cluster of differentiation 5 expression levels among children with autism. Neuropsychiatr Dis Treat. 2017;13:2397-2403. doi: 10.2147/NDT.S146152 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Halepoto DM, Alhowikan AM, Ayadhi LA. Cluster of Differentiation 5 (CD5) Levels in the Plasma of Children with Autism Spectrum Disorder (ASD). J Coll Physicians Surg–Pak JCPSP. 2017;27(3):149-152. [PubMed] [Google Scholar]
49.Zerbo O, Yoshida C, Grether JK, et al. Neonatal cytokines and chemokines and risk of autism spectrum disorder: the early markers for autism (EMA) study: a case-control study. J Neuroinflammation. 2014;11(1):113. doi: 10.1186/1742-2094-11-113 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Othman AB, Slama H, Cherif E, Azaiez M, Gharsallah H. Inflammatory cytokines dysfunction in autism spectrum disorder. Eur Psychiatry. 2021;64(S1):S214-S215. doi: 10.1192/j.eurpsy.2021.571 [DOI] [Google Scholar]
51.Hidayat F, Labeda I, Sampetoding S, et al. Correlation of interleukin-6 and C-reactive protein levels in plasma with the stage and differentiation of colorectal cancer: a cross-sectional study in East Indonesia. Ann Med Surg (Lond). 2021;62:334-340. doi: 10.1016/j.amsu.2021.01.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Palm M, Basu S, Larsson A, Wernroth L, Åkerud H, Axelsson O. A longitudinal study of plasma levels of soluble fms-like tyrosine kinase 1 (sFlt1), placental growth factor (PlGF), sFlt1: PlGF ratio and vascular endothelial growth factor (VEGF-A) in normal pregnancy. Acta Obstet Gynecol Scand. 2011;90(11):1244-1251. doi: 10.1111/j.1600-0412.2011.01186.x [DOI] [PubMed] [Google Scholar]
53.Rosenstein JM, Krum JM, Ruhrberg C. VEGF in the nervous system. Organogenesis. 2010;6(2):107-114. doi: 10.4161/org.6.2.11687 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Lee BH, Kim YK. Increased plasma VEGF levels in major depressive or manic episodes in patients with mood disorders. J Affect Disord. 2012;136(1-2):181-184. doi: 10.1016/j.jad.2011.07.021 [DOI] [PubMed] [Google Scholar]
55.Gnanasekaran A, Kelchen MN, Brogden NK, Smith RM. Vascular endothelial growth factor (VEGF) expression and neuroinflammation is increased in the frontopolar cortex of individuals with autism spectrum disorder. bioRXiv. Published online May 3, 2019. doi: 10.1101/627083 [DOI]
56.Tseng PT, Cheng YS, Chen YW, Wu CK, Lin PY. Increased levels of vascular endothelial growth factor in patients with major depressive disorder: A meta-analysis. Eur Neuropsychopharmacol. 2015;25(10):1622-1630. doi: 10.1016/j.euroneuro.2015.06.001 [DOI] [PubMed] [Google Scholar]
57.Arrode-Brusés G, Brusés JL. Maternal immune activation by poly I:C induces expression of cytokines IL-1β and IL-13, chemokine MCP-1 and colony stimulating factor VEGF in fetal mouse brain. J Neuroinflammation. 2012;9(1):83. doi: 10.1186/1742-2094-9-83 [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Shinjo A, Ventura W, Koide K, et al. Maternal smoking and placental expression of a panel of genes related to angiogenesis and oxidative stress in early pregnancy. Fetal Diagn Ther. 2014;35(4):289-295. doi: 10.1159/000357704 [DOI] [PubMed] [Google Scholar]
59.Shen Y, Ou J, Liu M, et al. Altered plasma levels of chemokines in autism and their association with social behaviors. Psychiatry Res. 2016;244:300-305. doi: 10.1016/j.psychres.2016.07.057 [DOI] [PubMed] [Google Scholar]
60.Liao YY, Tsai HC, Chou PY, et al. CCL3 promotes angiogenesis by dysregulation of miR-374b/ VEGF-A axis in human osteosarcoma cells. Oncotarget. 2016;7(4):4310-4325. doi: 10.18632/oncotarget.6708 [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Akira S, Takeda K. Toll-like receptor signalling. Nat Rev Immunol. 2004;4(7):499-511. doi: 10.1038/nri1391 [DOI] [PubMed] [Google Scholar]
62.Han VX, Patel S, Jones HF, et al. Maternal acute and chronic inflammation in pregnancy is associated with common neurodevelopmental disorders: a systematic review. Transl Psychiatry. 2021;11(1):71. doi: 10.1038/s41398-021-01198-w [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Sharma SK, Naidu G. The role of danger-associated molecular patterns (DAMPs) in trauma and infections. J Thorac Dis. 2016;8(7):1406-1409. doi: 10.21037/jtd.2016.05.22 [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Liu J, Buisman-Pijlman F, Hutchinson MR. Toll-like receptor 4: innate immune regulator of neuroimmune and neuroendocrine interactions in stress and major depressive disorder. Front Neurosci. 2014;8:309. doi: 10.3389/fnins.2014.00309 [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Afkham A, Eghbal-Fard S, Heydarlou H, Azizi R, Aghebati-Maleki L, Yousefi M. Toll-like receptors signaling network in pre-eclampsia: an updated review. J Cell Physiol. 2019;234(3):2229-2240. doi: 10.1002/jcp.27189 [DOI] [PubMed] [Google Scholar]
66.Di Candia L, Gomez E, Venereau E, et al. HMGB1 is upregulated in the airways in asthma and potentiates airway smooth muscle contraction via TLR4. J Allergy Clin Immunol. 2017;140(2):584-587.e8. doi: 10.1016/j.jaci.2016.11.049 [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Dasu MR, Devaraj S, Park S, Jialal I. Increased toll-like receptor (TLR) activation and TLR ligands in recently diagnosed type 2 diabetic subjects. Diabetes Care. 2010;33(4):861-868. doi: 10.2337/dc09-1799 [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Hughes HK, Rowland ME, Onore CE, Rogers S, Ciernia AV, Ashwood P. Dysregulated gene expression associated with inflammatory and translation pathways in activated monocytes from children with autism spectrum disorder. Transl Psychiatry. 2022;12(1):39. doi: 10.1038/s41398-021-01766-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Parenti I, Rabaneda LG, Schoen H, Novarino G. Neurodevelopmental disorders: from genetics to functional pathways. Trends Neurosci. 2020;43(8):608-621. doi: 10.1016/j.tins.2020.05.004 [DOI] [PubMed] [Google Scholar]
70.Furman D, Campisi J, Verdin E, et al. Chronic inflammation in the etiology of disease across the life span. Nat Med. 2019;25(12):1822-1832. doi: 10.1038/s41591-019-0675-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Shimizu Y, Sakata-Haga H, Saikawa Y, Hatta T. Influence of immune system abnormalities caused by maternal immune activation in the postnatal period. Cells. 2023;12(5):741. doi: 10.3390/cells12050741 [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Madan JC, Davis JM, Craig WY, et al. Maternal obesity and markers of inflammation in pregnancy. Cytokine. 2009;47(1):61-64. doi: 10.1016/j.cyto.2009.05.004 [DOI] [PubMed] [Google Scholar]
73.Sonuga-Barke EJS. Causal models of attention-deficit/hyperactivity disorder: from common simple deficits to multiple developmental pathways. Biol Psychiatry. 2005;57(11):1231-1238. doi: 10.1016/j.biopsych.2004.09.008 [DOI] [PubMed] [Google Scholar]
74.Sonuga-Barke E, Bitsakou P, Thompson M. Beyond the dual pathway model: evidence for the dissociation of timing, inhibitory, and delay-related impairments in attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry. 2010;49(4):345-355. doi: 10.1097/00004583-201004000-00009 [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials