Prenatal tobacco and alcohol exposure, white matter microstructure, and early language skills in toddlers from a South African birth cohort

Abstract

Introduction: Tobacco and alcohol are the two most common substances used during pregnancy, and both can disrupt neurodevelopment, resulting in cognitive and behavioral deficits including language difficulties. Previous studies show that children with prenatal substance exposure exhibit microstructural alterations in major white matter pathways, though few studies have investigated the impact of prenatal substance exposure on white matter microstructure and language skills during the toddler years.

Methods: In this study, 93 children (34 exposed to alcohol and/or tobacco) aged 23 years from the Drakenstein Child Health Study, South Africa, completed Expressive and Receptive Communication assessments from the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III) and underwent diffusion MRI scans. Diffusion images were preprocessed, and 11 major white matter tracts were isolated. Fractional anisotropy (FA) and mean diffusivity (MD) were extracted for each white matter tract. Linear regression was used to examine differences between the tobacco/alcohol exposed group and unexposed controls for FA, MD, and language scores, as well as relationships between brain metrics and language. There were no significant group differences in language scores or FA.

Results: Children with alcohol or tobacco exposure had lower average MD in the splenium of the corpus callosum compared to unexposed controls. Significant interactions between prenatal substance exposure and language scores were seen in 7 tracts but did not survive multiple comparisons correction.

Discussion: Our findings show that prenatal alcohol and/or tobacco exposure appear to alter the relationship between white matter microstructure and early language skills in this population of toddlers, potentially laying the basis of language deficits observed later in older children with prenatal substance exposure, which may have implications for learning and interventions.

Keywords: development; diffusion tensor imaging (DTI); early language; prenatal alcohol exposure (PAE); prenatal tobacco exposure.

Figure 1

Left sagittal view of isolated white matter tracts. Red, Uncinate Fasciculus (UF); Orange, Inferior Fronto-Occipital Fasciculus (IFOF); Yellow, Arcuate Fasciculus (AF); Green, Inferior Longitudinal Fasciculus (ILF); Blue, Genu of the Corpus Callosum (CC); Aqua, Body of the Corpus Callosum (CC); Purple, Splenium of the Corpus Callosum (CC).

Figure 2

Boxplot displaying mean FA values of the white matter tracts for prenatal substance exposure (PSE; defined as prenatal alcohol and/or tobacco exposure) and control groups. FA values were directly pulled from the white matter for each participant interest variables have not been regressed out. No group differences in FA were significant. ILF, inferior longitudinal fasciculus; UF, uncinate fasciculus; IFOF, inferior fronto-occipital fasciculus; AF, arcuate fasciculus; CC, corpus callosum.

Figure 3

Boxplot displaying mean MD values of the white matter tracts for prenatal substance exposure (PSE; defined as prenatal alcohol and/or tobacco exposure) and control groups. MD values were directly pulled from the white matter for each participant interest variables have not been regressed out. MD was significantly lower in the PSE group compared to controls in the left IFOF (p = 0.030; q = 0.33) and the splenium of the CC (p = 0.002, q = 0.034). ILF, inferior longitudinal fasciculus; UF, uncinate fasciculus; IFOF, inferior fronto-occipital fasciculus; AF, arcuate fasciculus; CC, corpus callosum.

Figure 4

There were significant interactions between Prenatal Substance Exposure (PSE; defined as prenatal alcohol and/or tobacco exposure) and FA for Expressive Communication scores in the left IFOF and body of the CC. PSE had an interaction with MD in the body of the CC for Receptive Communication scores. Interaction effects did not survive multiple comparisons correction (q > 0.05 for all). The top row of plots displays FAxPSE interactions for Expressive Communication while the bottom row shows MDxPSE interactions for Receptive Communication. PSE group (n = 34) can be visualized in blue, and the control group (n = 59) is shown in green. IFOF, Inferior Fronto-Occipital Fasciculus.

Figure 5

Prenatal Substance Exposure (PSE; defined as prenatal alcohol and/or tobacco exposure)had an interaction with MD in the bilateral ILF, right IFOF, genu and body of the CC, and right AF for Expressive Communication scores. None of these interaction effects survived multiple comparisons correction (q > 0.05 for all). PSE group (n = 34) can be visualized in blue, and the control group (n = 59) is shown in green. ILF, Inferior Longitudinal Fasciculus; IFOF, Inferior Fronto-Occipital Fasciculus; AF, Arcuate Fasciculus.