Longitudinal Assessment of Prenatal, Perinatal, and Early-Life Aflatoxin B1 Exposure in 828 Mother-Child Dyads from Bangladesh and Malawi

Abstract

Background: In utero or early-life exposure to aflatoxin, which contaminates staple crops in disadvantaged settings, may compromise pregnancy and infant outcomes, but investigations into the extent, persistence, and determinants of aflatoxin exposure at these life stages have lacked longitudinal data collection and broad geographic representation.

Objectives: Aflatoxin exposure and selected determinants thereof were characterized in mother-child dyads with serial plasma/serum samples in prenatal, perinatal, and early life in Malawi and Bangladesh.

Methods: Circulating aflatoxin B1 (AFB1)-lysine albumin adducts were measured in dyads from Bangladesh (n = 573; maternal first and third trimester, 3 mo postpartum, cord blood, infant 24 mo) and Malawi (n = 255; maternal second and third trimester, 6 mo postpartum, infant 6 and 18 mo) with isotope dilution mass spectrometry. We examined AFB1-lysine adduct magnitude, persistence, seasonality, and associations with infant feeding, and estimated daily AFB1 intake.

Results: Maternal AFB1-lysine was higher in Malawi (98% detectable; median: 0.469, IQR: 0.225-1.027 pg/µL) than in Bangladesh (59%; 0.030, nondetectable [nd]-0.077 pg/µL). Although estimated dietary exposure in Malawi was temporally stable (648 ng AFB1/day), estimated intake in Bangladesh was reduced by 94% between rainy and winter seasons (98 to 6 ng/day). AFB1-lysine was low in cord blood from Bangladesh (15% detectable; 0.045, 0.031-0.088 pg/µL among detectable) and in Malawian infants at 6 mo of age (0.072, nd-0.236 pg/µL), but reached maternal concentrations by 18 or 24 mo (Bangladesh: 0.034, nd-0.063 pg/µL; Malawi: 0.370, 0.195-0.964 pg/µL). In Malawian infants, exclusive breastfeeding at 3 mo was associated with 58% lower AFB1-lysine concentrations at 6 mo compared with other feeding modes (P = 0.010).

Conclusions: Among pregnant women, aflatoxin exposure was persistently high in Malawi, while lower and seasonal in Bangladesh. Infants were partially protected from exposure in utero and with exclusive breastfeeding, but exposures reached adult levels by 18-24 mo of age. The Bangladesh and Malawi trials are registered at clinicaltrials.gov as NCT00860470 and NCT01239693.

Keywords: aflatoxin; breastfeeding; cord blood; diet; infancy; mass spectrometry; pregnancy; seasonality; toxicology.

© The Author(s) 2022. Published by Oxford University Press on behalf of the American Society for Nutrition.

Figures

FIGURE 1

Number and percentage of samples with detectable AFB1-lysine adducts. (A) Counts of samples with detectable and nondetectable AFB1-lysine adducts. (B) Percentage of samples with detectable and nondetectable AFB1-lysine adducts. AFB1, aflatoxin B1; C-Cord, child's cord blood at delivery; C-6mo, child at 6 mo old; C-18mo, child at 18 mo old; C-24mo, child at 24 mo old; Detect, detectable; M-1TM, mother at first trimester; M-2TM, mother at second trimester; M-3TM, mother at third trimester; M-3mo, mother at 3 mo postpartum; M-6mo, mother at 6 mo postpartum; Non Detect, non-detectable.

FIGURE 2

Distributions of AFB1-lysine (lys) adduct levels within Bangladeshi and Malawian mother–child dyads. Box plots of AFB1-lysine adduct concentration with whiskers extended to 5th and 95th percentiles. Dotted red line indicates limit of detection (0.01 pg/µL); nondetectable values are imputed at 0.005 pg/µL. Horizontal bars indicate statistical significance in pairwise comparisons: ****P < 0.0001 by Kruskal-Wallis test and Dwass, Steel, Critchlow-Fligner adjustment for multiple comparisons. AFB1, aflatoxin B1; C-Cord, child's cord blood at delivery; C-6mo, child at 6 mo old; C-18mo, child at 18 mo old; C-24mo, child at 24 mo old; M-1TM, mother at first trimester; M-2TM, mother at second trimester; M-3TM, mother at third trimester; M-3mo, mother at 3 mo postpartum; M-6mo, mother at 6 mo postpartum.

FIGURE 3

Subject-specific trajectories of AFB1-lysine (lys) adduct concentrations within Bangladeshi and Malawian women through pregnancy and into the postpartum period. Data as spaghetti plots of AFB1-lysine adduct concentration in Bangladeshi (A) and Malawian (B) women. Gray lines connect samples from a single individual. Dotted red lines indicate limit of detection (0.01 pg/µL); nondetectable values are imputed at 0.005 pg/µL. AFB1, aflatoxin B1; M-1TM, mother at first trimester; M-2TM, mother at second trimester; M-3TM, mother at third trimester; M-3mo, mother at 3 mo postpartum; M-6mo, mother at 6 mo postpartum.

FIGURE 4

Seasonality of AFB1-lysine (lys) adduct exposure in Bangladesh and Malawi. (A) Boxplot of AFB1-lysine adduct concentrations in Bangladeshi mothers, for samples collected during the winter, dry, or rainy seasons. Whiskers extend to the 5th and 95th percentiles. Dotted red line indicates limit of detection (0.01 pg/µL); nondetectable values are imputed at 0.005 pg/µL. (B) AFB1-lysine adduct concentrations in Bangladeshi mothers, plotted by date of sample collection. LOWESS curves were fit to data from each maternal sample type. (C) Boxplot of AFB1-lysine adduct concentrations in Malawian mothers and children, for samples collected during either the rainy or dry seasons. Data are plotted as in panel A. (D) AFB1-lysine adduct concentrations in Malawian mothers and children, plotted by date of sample collection. LOWESS curves were fit to data from each sample type. Horizontal bars in boxplots indicate statistical significance pairwise comparisons by Kruskal-Wallis test, using the Dwass, Steel, Critchlow-Fligner adjustment for multiple comparisons; **P < 0.01, ****P < 0.0001. AFB1, aflatoxin B1; C, 6mo, child at 6 mo old; C, 18mo, child at 18 mo old; LOWESS, locally weighted scatterplot smoothing; M, 1TM, mother at first trimester; M, 2TM, mother at second trimester; M, 3TM, mother at third trimester; M, 3mo, mother at 3 mo postpartum; M, 6mo, mother at 6 mo postpartum.

FIGURE 5

Rates of seasonal change in AFB1-lysine (Lys) internal dose in Bangladeshi women. Censored linear regression of AFB1-lysine adduct concentration (log10-transformed) compared with day of sample collection during the dry and rainy seasons (left) or winter season (right), across all maternal sample types (1TM, 3TM, 3mo) and years of sample collection (2008–2011). Blue shading indicates the timing of the rainy reason. Units of the independent variable was the number of elapsed days relative to 15 February in the dry and rainy seasons (242 days total) or relative to 15 October in the winter season (123 days total). Censored regression best-fit lines and equations are shown for dry + rainy and winter seasons separately. Both slopes are significantly different from 0 (P < 0.0001). Dotted red line indicates limit of detection (0.01 pg/µL); nondetectable values were imputed at 0.005 pg/µL. The daily proportion of samples with nondetectable AFB1-lysine concentrations is shown in the band plots at the bottom of the graph; data represent LOWESS curves fit to daily values. AFB1, aflatoxin B1; LOWESS, locally weighted scatterplot smoothing; n.d., not detected; 1TM, first trimester; 3mo, 3 mo postpartum; 3TM, third trimester.