Exposure to pesticides in utero impacts the fetal immune system and response to vaccination in infancy

Indoor residual spraying (IRS) of insecticides is a key malaria control strategy, yet the biological effects of many pesticides on pregnant women and fetuses are poorly understood. Carbamate pesticides can modulate human immune function. This study investigated whether prenatal exposure to bendiocarb—a commonly used carbamate for IRS—is systemically absorbed by pregnant women, transferred transplacentally to fetuses, persists in early infancy, and alters fetal immune homeostasis and function. The work also examined whether prenatal exposure affects infant responses to routine measles vaccination, addressing potential long-term immunologic consequences of IRS-related exposure.

Prior studies indicate that pesticides, including organophosphates and organochlorines, can adversely affect fetal growth and neurodevelopment and may influence infant infection risk. Carbamates and related dithiocarbamates have immunomodulatory properties, including effects on T cell differentiation, NF-κB signaling, and cytokine production; some act via the aryl hydrocarbon receptor. Earlier human studies reported trace bendiocarb in maternal and cord blood, presumed due to household pesticide use, but at levels over 100-fold lower than observed here and without evaluation of fetal immune effects. Bendiocarb is thought to undergo rapid renal elimination, but environmental persistence and ongoing exposure may occur post-spraying.

Design and participants: A cohort of 300 HIV-negative pregnant women in Tororo District, Uganda, were enrolled at 12–20 weeks’ gestation and randomized to dihydroartemisinin-piperaquine (DP) or sulfadoxine-pyrimethamine (SP) for intermittent preventive treatment of malaria (NCT02163447). A total of 294 children were born (Oct 2014–May 2015). After birth, infants were randomized to DP monthly or every 12 weeks from 8 weeks to 24 months and followed clinically to 3 years. All infants received measles vaccine at 9 months; measles IgG was measured at 56 weeks. Exposure assessment: IRS with bendiocarb occurred Dec 2014–Feb 2015; dates of spraying were recorded for each home. Additional IRS rounds occurred Jun–Jul 2015, Nov–Dec 2015, Jun–Jul 2016, and Jul–Aug 2017. Malaria assessments: Mothers were screened every 4 weeks by LAMP; febrile episodes were confirmed by microscopy and treated. Placental malaria was assessed by microscopy, LAMP, and histopathology. Infants were categorized as malaria-exposed in utero if any maternal infection during pregnancy/delivery or placental malaria was present; active placental malaria was defined by positive placental LAMP. Bendiocarb quantification: Plasma bendiocarb was measured by LC/MS/MS (UPLC with RRHD C8 column; Sciex 6500) after solid-phase extraction. Linear regression standard curve 10–1000 pg/mL; LLOQ 10 pg/mL; deuterated bendiocarb internal standard. Immunologic assays: Cord blood mononuclear cells (CBMCs) and maternal PBMCs were phenotyped by flow cytometry (FoxP3 Fix/Perm kit; multiple antibody panels) for CD4 subsets (including Tregs: CD25highFoxP3+CD127lo), CD8 T cells, activation/proliferation marker Ki67, and memory/naïve markers. Dendritic cells (myeloid: Lin–HLA-DR+CD11c+CD123–; plasmacytoid: Lin–HLA-DR+CD11c–CD123+) and absolute CD4 counts were enumerated using TruCount tubes or counting beads. Cytokines/chemokines: Twenty analytes (14 cytokines, 6 chemokines) were quantified in cord plasma by Luminex; samples run in duplicate with specified limits of detection; values below detection set to LOD. T cell function: CBMCs were stimulated with PMA/ionomycin for 5 h to measure intracellular cytokines (IL-2, IFN-γ, TNFα; IL-8 assessed in total CD4). Antigen-specific proliferation was assessed by CFSE dilution after 6-day culture with P. falciparum schizont extract (PfSE) versus uninfected RBCs. Measles IgG: Plasma at 56 weeks was tested in duplicate by ELISA (Abcam ab108750); positivity called per kit thresholds (positive/equivocal/negative). Statistics: Nonparametric tests (Wilcoxon rank-sum, Wilcoxon signed-rank) for group comparisons; Spearman correlations; multivariate linear regression for infant immune parameters adjusted for prenatal malaria exposure, maternal chemoprevention, congenital CMV, infant sex, maternal gravidity, prematurity, and delivery mode; maternal models adjusted for relevant covariates; vaccine response models further adjusted for child chemoprevention and malaria episodes before 56 weeks. Measles IgG positivity compared by chi-squared; malaria incidence by negative binomial regression. Two-sided p<0.05 considered significant; cytokine multiple testing addressed by Bonferroni in figure-level analysis; multivariate models not adjusted for multiple comparisons.

Exposure and pharmacology: Bendiocarb was detected at high levels post-IRS and transferred transplacentally. Detection above LLOQ (10 pg/mL): 0/12 (0%) unexposed maternal plasma at delivery; 143/167 (86%) exposed maternal plasma at delivery; 51/109 (47%) exposed cord plasma; 66/114 (58%) exposed infant plasma at 8 weeks (prior to repeat IRS). Maternal plasma concentration at delivery correlated with gestational week of exposure (Spearman). Cord blood immune homeostasis: Exposed infants had higher non-naive CD4 T cell frequency (p=0.001) and lower Treg frequency (p=0.0025). Myeloid and plasmacytoid dendritic cell absolute counts were reduced (overall p=0.026). Cord plasma cytokines/chemokines were increased in exposed infants: TNF-RI, IL-2, IL-5, IL-7, MDC, IP-10, and others (unadjusted p-values included TNF-RI p<0.0001, IL-2 p<0.0001, IL-4 p=0.037, IL-5 p<0.0001, IL-7 p<0.0001, MDC p<0.0001, IP-10 p=0.0009, MIP-1α p=0.01); after Bonferroni correction, all remained significant except IL-4 and MIP-1α. Cord T cell functional responses: Upon PMA/ionomycin stimulation, exposed infants had higher frequencies of cytokine-producing T cells: non-naive CD4 producing IL-2 (p<0.0001), IFN-γ, and TNFα (p=0.0002); total CD4 producing IL-8 increased (p=0.004); non-naive CD8 producing IL-2 increased (p<0.0001). These differences were independent of malaria exposure and correlated with maternal bendiocarb plasma levels (dose–response; Table 1). Proliferation: Maternal CD4 Ki67% was lower with exposure, indicating reduced in vivo proliferation (p=0.0005). Antigen-driven proliferation of cord T cells to PfSE was inhibited in exposed infants. Overall, CD8 CFSElow was reduced with exposure (p=0.038) and CD4 CFSElow showed a trend (p=0.056). In stratified analyses, proliferation to PfSE was robust among unexposed infants with active placental malaria (CD4 p=0.0004; CD8 p=0.004) but was ablated among exposed infants (CD4 p=0.035; CD8 p=0.037 when stratified). Dose–response and multivariable models: Maternal bendiocarb levels at delivery were associated with lower cord Tregs (p=0.011), higher cord plasma IL-2 (p=0.046), IL-5 (p=0.015), IL-7 (p=0.007), TNF-RI (p=0.019), and higher stimulated T cell cytokines (e.g., CD4+IL-2 p=0.009, CD4+IL-8 p=0.010, CD8+IL-2 p=0.013). In adjusted models (Table 2), prenatal exposure remained associated with: lower Tregs (coef -0.49; p=0.016), lower myeloid DCs (coef -31.74; p=0.018), higher non-naive CD4 (coef 1.85; p=0.004), higher cord cytokines (e.g., IL-2 coef 31.58; p<0.0001; IL-5 coef 106.13; p<0.0001; IL-7 coef 4.01; p<0.0001; MDC coef 230.61; p<0.0001; TNF-R1 coef 2318.01; p<0.0001; IP-10 coef 13.61; p=0.038), higher stimulated cytokines (e.g., CD4+IL-2 coef 17.35; p<0.0001), lower CD8 proliferation to PfSE (coef -3.71; p=0.007), and lower maternal CD4 Ki67 (coef -0.58; p=0.005). Vaccine response: At 1 year, measles IgG levels were higher among exposed infants (p<0.0001), and IgG positivity was more frequent (91.3% vs 77.8%; p=0.008). Association with measles titers persisted after adjustment (coef 3.28; p=0.001). Measles IgG positivity did not differ by in utero malaria exposure, sex, gravidity, or number of malaria episodes in the first year. Overall: Prenatal bendiocarb exposure is linked to systemic absorption and fetal exposure, shifts fetal immune homeostasis toward inflammation with reduced regulatory T cells, inhibits antigen-driven proliferation, and enhances measles vaccine antibody responses at 1 year.

The study demonstrates that bendiocarb used in IRS is systemically absorbed by pregnant women, crosses the placenta, and is detectable in infants, indicating meaningful fetal exposure. Prenatal exposure was associated with a shift in fetal immune homeostasis toward an inflammatory phenotype—characterized by decreased FoxP3+ regulatory CD4 T cells, increased circulating cytokines/chemokines, heightened T cell cytokine production, and reduced T cell proliferation. These coordinated changes imply altered regulation of fetal immunity that could influence maternofetal tolerance and postnatal immune development. Mechanisms may involve aryl hydrocarbon receptor signaling, NF-κB pathway modulation, apoptosis, or direct antiproliferative effects observed with carbamates, though precise targets remain undefined. Despite reduced antigen-driven proliferation, exposed infants exhibited higher measles vaccine IgG titers at one year, suggesting durable modulation of functional immunity that is not uniformly detrimental. No effect on malaria incidence during infancy was detected, possibly masked by postnatal IRS and chemoprevention. The findings underscore that vector control pesticides can have unintended immunologic effects on the fetus and infant, necessitating careful risk–benefit evaluation and mechanistic investigation.

This work provides evidence that prenatal exposure to the carbamate pesticide bendiocarb results in systemic maternal absorption, transplacental transfer, and detectable infant levels, and is associated with dose-dependent alterations in fetal immune homeostasis and function. These include reduced regulatory T cells, increased inflammatory cytokines, diminished antigen-driven T cell proliferation, and enhanced measles vaccine antibody responses at one year. The study highlights the need to balance malaria control benefits of IRS against potential immunologic risks to fetuses and infants. Future research should define bendiocarb pharmacokinetics in pregnancy and infancy, clarify cellular and molecular mechanisms of immune modulation, assess exposure earlier in gestation, evaluate breast milk contributions, and determine clinical consequences and long-term health outcomes.

- Exposure timing: All participants were exposed during the second half of pregnancy; effects of first-trimester or earlier exposure were not assessed. - Potential confounding: Postnatal exposure (subsequent IRS rounds) and environmental contact/breastfeeding could contribute to infant bendiocarb levels; breast milk was not collected. - Pharmacokinetics: Lack of detailed PK data in pregnant women/infants limits interpretation of persistence and clearance. - Multiple comparisons: Multivariable models were not adjusted for multiple comparisons, increasing risk of type I error; although Bonferroni correction was applied in some cytokine analyses, not all analyses were corrected. - Generalizability: Findings from a holoendemic malaria region with concurrent malaria interventions (chemoprevention, IRS) may not generalize to other settings. - Clinical outcomes: While immune alterations were observed, clear impacts on clinical endpoints (e.g., infection incidence, adverse pregnancy outcomes) were not demonstrated within this study.