Variability of multi-omics profiles in a population-based child cohort

A comprehensive review of machine learning techniques for multi-omics data integration: challenges and applications in precision oncology

Multi-omics data play a crucial role in precision medicine, mainly to understand the diverse biological interaction between different omics. Machine learning approaches have been extensively employed in this context over the years. This review aims to comprehensively summarize and categorize these advancements, focusing on the integration of multi-omics data, which includes genomics, transcriptomics, proteomics and metabolomics, alongside clinical data. We discuss various machine learning techniques and computational methodologies used for integrating distinct omics datasets and provide valuable insights into their application. The review emphasizes both the challenges and opportunities present in multi-omics data integration, precision medicine and patient stratification, offering practical recommendations for method selection in various scenarios. Recent advances in deep learning and network-based approaches are also explored, highlighting their potential to harmonize diverse biological information layers. Additionally, we present a roadmap for the integration of multi-omics data in precision oncology, outlining the advantages, challenges and implementation difficulties. Hence this review offers a thorough overview of current literature, providing researchers with insights into machine learning techniques for patient stratification, particularly in precision oncology. Contact:  anirban@klyuniv.ac.in.

Longitudinal multi-omics study reveals common etiology underlying association between plasma proteome and BMI trajectories in adolescent and young adult twins

BACKGROUND

The influence of genetics and environment on the association of the plasma proteome with body mass index (BMI) and changes in BMI remains underexplored, and the links to other omics in these associations remain to be investigated. We characterized protein-BMI trajectory associations in adolescents and adults and how these connect to other omics layers.

METHODS

Our study included two cohorts of longitudinally followed twins: FinnTwin12 (N = 651) and the Netherlands Twin Register (NTR) (N = 665). Follow-up comprised 4 BMI measurements over approximately 6 (NTR: 23-27 years old) to 10 years (FinnTwin12: 12-22 years old), with omics data collected at the last BMI measurement. BMI changes were calculated in latent growth curve models. Mixed-effects models were used to quantify the associations between the abundance of 439 plasma proteins with BMI at blood sampling and changes in BMI. In FinnTwin12, the sources of genetic and environmental variation underlying the protein abundances were quantified by twin models, as were the associations of proteins with BMI and BMI changes. In NTR, we investigated the association of gene expression of genes encoding proteins identified in FinnTwin12 with BMI and changes in BMI. We linked identified proteins and their coding genes to plasma metabolites and polygenic risk scores (PRS) applying mixed-effects models and correlation networks.

RESULTS

We identified 66 and 14 proteins associated with BMI at blood sampling and changes in BMI, respectively. The average heritability of these proteins was 35%. Of the 66 BMI-protein associations, 43 and 12 showed genetic and environmental correlations, respectively, including 8 proteins showing both. Similarly, we observed 7 and 3 genetic and environmental correlations between changes in BMI and protein abundance, respectively. S100A8 gene expression was associated with BMI at blood sampling, and the PRG4 and CFI genes were associated with BMI changes. Proteins showed strong connections with metabolites and PRSs, but we observed no multi-omics connections among gene expression and other omics layers.

CONCLUSIONS

Associations between the proteome and BMI trajectories are characterized by shared genetic, environmental, and metabolic etiologies. We observed few gene-protein pairs associated with BMI or changes in BMI at the proteome and transcriptome levels.

Multi-Omics Integration for the Design of Novel Therapies and the Identification of Novel Biomarkers

Multi-omics is a cutting-edge approach that combines data from different biomolecular levels, such as DNA, RNA, proteins, metabolites, and epigenetic marks, to obtain a holistic view of how living systems work and interact. Multi-omics has been used for various purposes in biomedical research, such as identifying new diseases, discovering new drugs, personalizing treatments, and optimizing therapies. This review summarizes the latest progress and challenges of multi-omics for designing new treatments for human diseases, focusing on how to integrate and analyze multiple proteome data and examples of how to use multi-proteomics data to identify new drug targets. We also discussed the future directions and opportunities of multi-omics for developing innovative and effective therapies by deciphering proteome complexity.

Integrating -omics approaches into population-based studies of endocrine disrupting chemicals: A scoping review

Health effects of endocrine disrupting chemicals (EDCs) are challenging to detect in the general population. Omics technologies become increasingly common to identify early biological changes before the apparition of clinical symptoms, to explore toxic mechanisms and to increase biological plausibility of epidemiological associations. This scoping review systematically summarises the application of omics in epidemiological studies assessing EDCs-associated biological effects to identify potential gaps and priorities for future research. Ninety-eight human studies (2004-2021) were identified through database searches (PubMed, Scopus) and citation chaining and focused on phthalates (34 studies), phenols (19) and PFASs (17), while PAHs (12) and recently-used pesticides (3) were less studied. The sample sizes ranged from 10 to 12,476 (median = 159), involving non-pregnant adults (38), pregnant women (11), children/adolescents (15) or both latter populations studied together (23). Several studies included occupational workers (10) and/or highly exposed groups (11) focusing on PAHs, PFASs and pesticides, while studies on phenols and phthalates were performed in the general population only. Analysed omics layers included metabolic profiles (30, including 14 targeted analyses), miRNA (13), gene expression (11), DNA methylation (8), microbiome (5) and proteins (3). Twenty-one studies implemented targeted multi-assays focusing on clinical routine blood lipid traits, oxidative stress or hormones. Overall, DNA methylation and gene expression associations with EDCs did not overlap across studies, while some EDC-associated metabolite groups, such as carnitines, nucleotides and amino acids in untargeted metabolomic studies, and oxidative stress markers in targeted studies, were consistent across studies. Studies had common limitations such as small sample sizes, cross-sectional designs and single sampling for exposure biomonitoring. In conclusion, there is a growing body of evidence evaluating the early biological responses to exposure to EDCs. This review points to a need for larger longitudinal studies, wider coverage of exposures and biomarkers, replication studies and standardisation of research methods and reporting.

Longitudinal multi-omics study reveals common etiology underlying association between plasma proteome and BMI trajectories in adolescent and young adult twins

Background

The influence of genetics and environment on the association of the plasma proteome with body mass index (BMI) and changes in BMI remain underexplored, and the links to other omics in these associations remain to be investigated. We characterized protein-BMI trajectory associations in adolescents and adults and how these connect to other omics layers.

Methods

Our study included two cohorts of longitudinally followed twins: FinnTwin12 (N=651) and the Netherlands Twin Register (NTR) (N=665). Follow-up comprised four BMI measurements over approximately 6 (NTR: 23-27 years old) to 10 years (FinnTwin12: 12-22 years old), with omics data collected at the last BMI measurement. BMI changes were calculated using latent growth curve models. Mixed-effects models were used to quantify the associations between the abundance of 439 plasma proteins with BMI at blood sampling and changes in BMI. The sources of genetic and environmental variation underlying the protein abundances were quantified using twin models, as were the associations of proteins with BMI and BMI changes. In NTR, we investigated the association of gene expression of genes encoding proteins identified in FinnTwin12 with BMI and changes in BMI. We linked identified proteins and their coding genes to plasma metabolites and polygenic risk scores (PRS) using mixed-effect models and correlation networks.

Results

We identified 66 and 14 proteins associated with BMI at blood sampling and changes in BMI, respectively. The average heritability of these proteins was 35%. Of the 66 BMI-protein associations, 43 and 12 showed genetic and environmental correlations, respectively, including 8 proteins showing both. Similarly, we observed 6 and 4 genetic and environmental correlations between changes in BMI and protein abundance, respectively. S100A8 gene expression was associated with BMI at blood sampling, and the PRG4 and CFI genes were associated with BMI changes. Proteins showed strong connections with many metabolites and PRSs, but we observed no multi-omics connections among gene expression and other omics layers.

Conclusions

Associations between the proteome and BMI trajectories are characterized by shared genetic, environmental, and metabolic etiologies. We observed few gene-protein pairs associated with BMI or changes in BMI at the proteome and transcriptome levels.

Childhood exposure to non-persistent endocrine disrupting chemicals and multi-omic profiles: A panel study

BACKGROUND

Individuals are exposed to environmental pollutants with endocrine disrupting activity (endocrine disruptors, EDCs) and the early stages of life are particularly susceptible to these exposures. Previous studies have focused on identifying molecular signatures associated with EDCs, but none have used repeated sampling strategy and integrated multiple omics. We aimed to identify multi-omic signatures associated with childhood exposure to non-persistent EDCs.

METHODS

We used data from the HELIX Child Panel Study, which included 156 children aged 6 to 11. Children were followed for one week, in two time periods. Twenty-two non-persistent EDCs (10 phthalate, 7 phenol, and 5 organophosphate pesticide metabolites) were measured in two weekly pools of 15 urine samples each. Multi-omic profiles (methylome, serum and urinary metabolome, proteome) were measured in blood and in a pool urine samples. We developed visit-specific Gaussian Graphical Models based on pairwise partial correlations. The visit-specific networks were then merged to identify reproducible associations. Independent biological evidence was systematically sought to confirm some of these associations and assess their potential health implications.

RESULTS

950 reproducible associations were found among which 23 were direct associations between EDCs and omics. For 9 of them, we were able to find corroborating evidence from previous literature: DEP - serotonin, OXBE - cg27466129, OXBE - dimethylamine, triclosan - leptin, triclosan - serotonin, MBzP - Neu5AC, MEHP - cg20080548, oh-MiNP - kynurenine, oxo-MiNP - 5-oxoproline. We used these associations to explore possible mechanisms between EDCs and health outcomes, and found links to health outcomes for 3 analytes: serotonin and kynurenine in relation to neuro-behavioural development, and leptin in relation to obesity and insulin resistance.

CONCLUSIONS

This multi-omics network analysis at two time points identified biologically relevant molecular signatures related to non-persistent EDC exposure in childhood, suggesting pathways related to neurological and metabolic outcomes.

Systemic interindividual epigenetic variation in humans is associated with transposable elements and under strong genetic control

BACKGROUND

Genetic variants can modulate phenotypic outcomes via epigenetic intermediates, for example at methylation quantitative trait loci (mQTL). We present the first large-scale assessment of mQTL at human genomic regions selected for interindividual variation in CpG methylation, which we call correlated regions of systemic interindividual variation (CoRSIVs). These can be assayed in blood DNA and do not reflect interindividual variation in cellular composition.

RESULTS

We use target-capture bisulfite sequencing to assess DNA methylation at 4086 CoRSIVs in multiple tissues from each of 188 donors in the NIH Gene-Tissue Expression (GTEx) program. At CoRSIVs, DNA methylation in peripheral blood correlates with methylation and gene expression in internal organs. We also discover unprecedented mQTL at these regions. Genetic influences on CoRSIV methylation are extremely strong (median R²=0.76), cumulatively comprising over 70-fold more human mQTL than detected in the most powerful previous study. Moreover, mQTL beta coefficients at CoRSIVs are highly skewed (i.e., the major allele predicts higher methylation). Both surprising findings are independently validated in a cohort of 47 non-GTEx individuals. Genomic regions flanking CoRSIVs show long-range enrichments for LINE-1 and LTR transposable elements; the skewed beta coefficients may therefore reflect evolutionary selection of genetic variants that promote their methylation and silencing. Analyses of GWAS summary statistics show that mQTL polymorphisms at CoRSIVs are associated with metabolic and other classes of disease.

CONCLUSIONS

A focus on systemic interindividual epigenetic variants, clearly enhanced in mQTL content, should likewise benefit studies attempting to link human epigenetic variation to the risk of disease.

Untargeted Metabolomics Based on Liquid Chromatography-Mass Spectrometry for the Analysis of Plasma and Erythrocyte Samples in Childhood Obesity

The circulating metabolome of human peripheral blood provides valuable information to investigate the molecular mechanisms underlying the development of diseases and to discover candidate biomarkers. In particular, erythrocytes have been proposed as potential systemic indicators of the metabolic and redox status of the organism. To accomplish wide-coverage metabolomics analysis, the combination of complementary analytical techniques is necessary to manage the physicochemical complexity of the human metabolome. Herein, we describe an untargeted metabolomics method to capture the plasmatic and erythroid metabolomes based on ultrahigh-performance liquid chromatography coupled to high-resolution mass spectrometry, combining reversed-phase liquid chromatography and hydrophilic interaction liquid chromatography. The method provides comprehensive metabolomics fingerprinting of plasma and erythrocyte samples, thereby enabling the elucidation of the distinctive metabolic disturbances behind childhood obesity and associated comorbidities, such as insulin resistance.

Genome-wide distribution of histone trimethylation reveals a global impact of bisphenol A on telomeric binding proteins and histone acetyltransferase factors: a pilot study with human and in vitro data

OBJECTIVE

To assess the genetic and epigenetic effects promoted by Bisphenol A (BPA) exposure in adolescent males from the Spanish INMA-Granada birth cohort, and in human cells.

METHODS

DNA methylation was analysed using MEDIP. Repeat number variation in genomic DNA was evaluated, along with the analysis of H3K4me3 by using chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq). Analyses were performed with material extracted from whole blood of the adolescents, complemented by in vitro assessments of human (HeLa) cells exposed to 10 nM BPA, specifically, immunofluorescence evaluation of protein levels, gene expression analysis and ChIP‒qPCR analysis.

RESULTS

Adolescents in the high urinary BPA levels group presented a higher level of Satellite A (SATA) repetitive region copy numbers compared to those in the low BPA group and a tendency towards increase in telomere length. We also observed decreased DNA methylation at the promoters of the imprinted genes H19, KCNQ1, and IGF2; at LINE1 retroelements; and at the ARID2, EGFR and ESRRA and TERT genes. Genome-wide sequencing revealed increased H3K4me3 occupancy at the promoters of genes encoding histone acetyltransferases, telomeric DNA binding factors and DNA repair genes. Results were supported in HeLa cells exposed to 10 nM BPA in vitro. In accordance with the data obtained in blood samples, we observed higher H3K4me3 occupancy and lower DNA methylation at some specific targets in HeLa cells. In exposed cells, changes in the expression of genes encoding DNA repair factors (ATM, ARID2, TRP53) were observed, and increased expression of several genes encoding telomeric DNA binding factors (SMG7, TERT, TEN1, UPF1, ZBTB48) were also found. Furthermore, an increase in ESR1/ERa was observed in the nuclei of HeLa cells along with increased binding of ESR1 to KAT5, KMT2E and TERF2IP promoters and decreased ESR1 binding at the RARA promoter. The DNA damage marker p53/TP53 was also increased.

CONCLUSION

In this pilot study, genome-wide analysis of histone trimethylation in adolescent males exposed to BPA revealed a global impact on the expression of genes encoding telomeric binding proteins and histone acetyltransferase factors with similar results in HeLa cells. Nevertheless, larger studies should confirm our findings.

The clinical significance of integrin subunit alpha V in cancers: from small cell lung carcinoma to pan-cancer

BACKGROUND

Little is known about the relationship between integrin subunit alpha V (ITGAV) and cancers, including small cell lung cancer (SCLC).

METHODS

Using large sample size from multiple sources, the clinical roles of ITGAV expression in SCLC were explored using differential expression analysis, receiver operating characteristic curves, Kaplan-Meier curves, etc. RESULTS: Decreased mRNA (SMD = - 1.05) and increased protein levels of ITGAV were detected in SCLC (n = 865). Transcription factors-ZEB2, IK2F1, and EGR2-may regulate ITGAV expression in SCLC, as they had ChIP-Seq (chromatin immunoprecipitation followed by sequencing) peaks upstream of the transcription start site of ITGAV. ITGAV expression made it feasible to distinguish SCLC from non-SCLC (AUC = 0.88, sensitivity = 0.78, specificity = 0.84), and represented a risk role in the prognosis of SCLC (p < 0.05). ITGAV may play a role in cancers by influencing several immunity-related signaling pathways and immune cells. Further, the extensive pan-cancer analysis verified the differential expression of ITGAV and its clinical significance in multiple cancers.

CONCLUSION

ITGAV served as a potential marker for prognosis and identification of cancers including SCLC.

Predict ovarian cancer by pairing serum miRNAs: Construct of single sample classifiers

Objective

The accuracy of CA125 or clinical examination in ovarian cancer (OVC) screening is still facing challenges. Serum miRNAs have been considered as promising biomarkers for clinical applications. Here, we propose a single sample classifier (SSC) method based on within-sample relative expression orderings (REOs) of serum miRNAs for OVC diagnosis.

Methods

Based on the stable REOs within 4,965 non-cancer serum samples, we developed the SSC for OVC in the training cohort (GSE106817: OVC = 200, non-cancer = 2,000) by focusing on highly reversed REOs within OVC. The best diagnosis is achieved using a combination of reversed miRNA pairs, considering the largest evaluation index and the lowest number of miRNA pairs possessed according to the voting rule. The SSC was then validated in internal data (GSE106817: OVC = 120, non-cancer = 759) and external data (GSE113486: OVC = 40, non-cancer = 100).

Results

The obtained 13-miRPairs classifier showed high diagnostic accuracy on distinguishing OVC from non-cancer controls in the training set (sensitivity = 98.00%, specificity = 99.60%), which was reproducible in internal data (sensitivity = 98.33%, specificity = 99.21%) and external data (sensitivity = 97.50%, specificity = 100%). Compared with the published models, it stood out in terms of correct positive predictive value (PPV) and negative predictive value (NPV) (PPV = 96.08% and NPV=95.16% in training set, and both above 99% in validation set). In addition, 13-miRPairs demonstrated a classification accuracy of over 97.5% for stage I OVC samples. By integrating other non-OVC serum samples as a control, the obtained 17-miRPairs classifier could distinguish OVC from other cancers (AUC&gt;92% in training and validation set).

Conclusion

The REO-based SSCs performed well in predicting OVC (including early samples) and distinguishing OVC from other cancer types, proving that REOs of serum miRNAs represent a robust and non-invasive biomarker.

Clinical significance of cyclin-dependent kinase inhibitor 2C expression in cancers: from small cell lung carcinoma to pan-cancers

Background

Cyclin-dependent kinase inhibitor 2C (CDKN2C) was identified to participate in the occurrence and development of multiple cancers; however, its roles in small cell lung carcinoma (SCLC) remain unclear.

Methods

Differential expression analysis of CDKN2C between SCLC and non-SCLC were performed based on 937 samples from multiple centers. The prognosis effects of CDKN2C in patients with SCLC were detected using both Kaplan–Meier curves and log-rank tests. Using receiver-operating characteristic curves, whether CDKN2C expression made it feasible to distinguish SCLC was determined. The potential mechanisms of CDKN2C in SCLC were investigated by gene ontology terms and signaling pathways (Kyoto Encyclopedia of Genes and Genomes). Based on 10,080 samples, a pan-cancer analysis was also performed to determine the roles of CDKN2C in multiple cancers.

Results

For the first time, upregulated CDKN2C expression was detected in SCLC samples at both the mRNA and protein levels (p of Wilcoxon rank-sum test < 0.05; standardized mean difference = 2.86 [95% CI 2.20–3.52]). Transcription factor FOXA1 expression may positively regulate CDKN2C expression levels in SCLC. High CDKN2C expression levels were related to the poor prognosis of patients with SCLC (hazard ratio > 1, p < 0.05) and showed pronounced effects for distinguishing SCLC from non-SCLC (sensitivity, specificity, and area under the curve ≥ 0.95). CDKN2C expression may play a role in the development of SCLC by affecting the cell cycle. Furthermore, the first pan-cancer analysis revealed the differential expression of CDKN2C in 16 cancers (breast invasive carcinoma, etc.) and its independent prognostic significance in nine cancers (e.g., adrenocortical carcinoma). CDKN2C expression was related to the immune microenvironment, suggesting its potential usefulness as a prognostic marker in immunotherapy.

Conclusions

This study identified upregulated CDKN2C expression and its clinical significance in SCLC and other multiple cancers, suggesting its potential usefulness as a biomarker in treating and differentiating cancers.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12890-022-02036-5.

Pro-inflammatory and (Epi-)genetic markers in saliva for disease risk in childhood obesity

BACKGROUND AND AIM

Childhood obesity is an emerging problem often leading to earlier onset of non-communicable diseases in later life. Biomarkers to identify individual risk scores are insufficient in routine clinical practice, which is related to the need for easily sampled, non-invasive survey methods in children. We aimed to investigate and strengthen possible pro-inflammatory markers and epigenetic risk factors in saliva of obese children compared to lean controls.

METHODS AND RESULTS

19 overweight/obese (OC, 10.1 ± 1.9 years, BMI 27.7 ± 3.2 kg/m²) and 19 lean control children (CC, 9.7 ± 2.5 years, BMI 16.4 ± 1.8 kg/m²) participated in this explorative pilot study. Anthropometric measures, saliva and cheek swab samples were taken. Saliva profiles were examined for acute phase proteins (CRP and neopterin) and pro-inflammatory cytokines (IL-17a/IL-1β/IL-6). Cheek swabs were analyzed to investigate DNA methylation differences with subsequent hierarchical cluster and principal component analyses (PCA). Saliva analysis showed significant increased CRP concentrations in OC compared to CC (p < 0.001). There were no significant differences, but high intra-individual values in neopterin, IL-17a, IL-1β and IL-6. An unsupervised PCA of CpG loci with high variance (σ/σ_max > 0.2) clearly separated OC and CC according to their methylation pattern. Furthermore, a supervised approach revealed 7125 significantly differentially methylated loci, whose corresponding genes were significantly enriched for genes playing roles in e.g., cellular signalling, cytoskeleton organization and cell motility.

CONCLUSIONS

CRP and methylation status determinations in saliva are suitable as non-invasive methods for early detection of risks for non-communicable diseases in children/adolescents and might be a useful supplementary approach in the routine clinical practice/monitoring.

Epigenetic landscape in the kick-and-kill therapeutic vaccine BCN02 clinical trial is associated with antiretroviral treatment interruption (ATI) outcome

BACKGROUND

The BCN02-trial combined therapeutic vaccination with a viral latency reversing agent (romidepsin, RMD) in HIV-1-infected individuals and included a monitored antiretroviral pause (MAP) as an efficacy read-out identifying individuals with an early or late (< or > 4weeks) viral-rebound. Integrated -omics analyses were applied prior treatment interruption to identify markers of virus control during MAP.

METHODS

PBMC, whole-genome DNA methylation and transcriptomics were assessed in 14 BCN02 participants, including 8 Early and 4 Late viral-rebound individuals. Chromatin state, histone marks and integration analysis (histone-3 acetylation (H3Ac), viral load, proviral levels and HIV-specific T cells responses) were included. REDUC-trial samples (n = 5) were included as a control group for RMD administration alone.

FINDINGS

DNA methylation imprints after receiving the complete intervention discriminated Early versus Late viral-rebound individuals before MAP. Also, differential chromatin accessibility and histone marks at DNA methylation level were detected. Importantly, the differential DNA methylation positions (DMPs) between Early and Late rebounders before MAP were strongly associated with viral load, proviral levels as well as the HIV-specific T-cell responses. Most of these DMPs were already present prior to the intervention and accentuated after RMD infusion.

INTERPRETATION

This study identifies host DNA methylation profiles and epigenetic cascades that are predictive of subsequent virus control in a kick-and-kill HIV cure strategy.

FUNDING

European Union Horizon 2020 Framework Programme for Research and Innovation under Grant Agreement N°681137-EAVI2020 and N°847943-MISTRAL, the Ministerio de Ciencia e Innovación (SAF2017_89726_R), and the National Institutes of Health-National Institute of Allergy and Infectious Diseases Program Grant P01-AI131568.

BDNF as a potential mediator between childhood BPA exposure and behavioral function in adolescent boys from the INMA-Granada cohort

BACKGROUND

Bisphenol A (BPA) exposure has been linked to altered behavior in children. Within the European Human Biomonitoring Initiative (HBM4EU), an adverse outcome pathway (AOP) network was constructed supporting the mechanistic link between BPA exposure and brain-derived neurotrophic factor (BDNF).

OBJECTIVE

To test this toxicologically-based hypothesis in the prospective INMA-Granada birth cohort (Spain).

METHODS

BPA concentrations were quantified by LC-MS/MS in spot urine samples from boys aged 9-11 years, normalized by creatinine and log-2 transformed. At adolescence (15-17 years), blood and urine specimens were collected, and serum and urinary BDNF protein levels were measured using immunoassays. DNA methylation levels at 6 CpGs in Exon IV of the BDNF gene were also assessed in peripheral blood using bisulfite-pyrosequencing. Adolescent's behavior was parent-rated using the Child Behavior Checklist (CBCL/6-18) in 148 boys. Adjusted linear regression and mediation models were fit.

RESULTS

Childhood urinary BPA concentrations were longitudinally and positively associated with thought problems (β = 0.76; 95% CI: 0.02, 1.49) and somatic complaints (β = 0.80; 95% CI: -0.16, 1.75) at adolescence. BPA concentrations were positively associated with BDNF DNA methylation at CpG6 (β = 0.21; 95% CI: 0.06, 0.36) and mean CpG methylation (β = 0.10; 95% CI: 0.01, 0.18), but not with total serum or urinary BDNF protein levels. When independent variables were categorized in tertiles, positive dose-response associations were observed between BPA-thought problems (p-trend = 0.08), BPA-CpG6 (p-trend ≤ 0.01), and CpG6-thought problems (p-trend ≤ 0.01). A significant mediated effect by CpG6 DNA methylation was observed (β = 0.23; 95% CI: 0.01, 0.57), accounting for up to 34% of the BPA-thought problems association.

CONCLUSIONS

In line with toxicological studies, BPA exposure was longitudinally associated with increased BDNF DNA methylation, supporting the biological plausibility of BPA-behavior relationships previously described in the epidemiological literature. Given its novelty and preliminary nature, this effect biomarker approach should be replicated in larger birth cohorts.