Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 15.
Published in final edited form as: Alcohol Clin Exp Res (Hoboken). 2024 May 2;48(6):1050–1062. doi: 10.1111/acer.15332

Lower resting state functional connectivity partially mediates adverse effects of prenatal alcohol exposure on arithmetic performance in children

Jia Fan 1,2, Keri J Woods 1,2, Joseph L Jacobson 3,4, Paul A Taylor 5, Jadrana T F Toich 1, Christopher D Molteno 6, Sandra W Jacobson 3,4,6, Ernesta M Meintjes 1,2,7
PMCID: PMC11998671  NIHMSID: NIHMS2053142  PMID: 38697927

Abstract

Background:

Fetal alcohol spectrum disorders (FASD) include a range of neurocognitive and behavioral impairments resulting from prenatal alcohol exposure (PAE). Among the PAE-related cognitive deficits, number processing is particularly affected. This study examines alterations in number processing networks and whether changes in functional connectivity mediate the adverse effects of PAE on arithmetic performance.

Methods:

Magnetic resonance imaging (MRI) was acquired in 57 children (mean (SD) age = 11.3 (+0.9) yr), 38 with FASD (19 fetal alcohol syndrome (FAS) or partial FAS (PFAS), 19 heavily exposed (HE)) and 19 controls. Whole-brain correlation analyses were performed from five seeds located in regions involved in number processing.

Results:

Children with FAS/PFAS showed dose-dependent reductions in resting state functional connectivity between the seed in the right (R) posterior superior parietal lobule and a cluster in the left (L) inferior frontal gyrus, and between a seed in the R horizontal intraparietal sulcus and clusters in the R precentral gyrus and L cerebellar lobule VI. HE children showed lower resting state functional connectivity in a subset of these regions. Lower functional connectivity in the two fronto-parietal connections partially mediated the adverse effects of PAE on arithmetic performance.

Conclusion:

This study demonstrates PAE-related functional connectivity impairments in functional networks involved in number processing. The weaker connectivity between the R posterior superior parietal lobule and the L inferior frontal gyrus suggests that impaired verbal processing and visuospatial working memory may play a role in number processing deficits, while weaker connectivity between the R intraparietal sulcus and the R precentral gyrus points to poorer finger-based numerical representation, which has been linked to arithmetic computational skills.

Keywords: arithmetic, fetal alcohol spectrum disorders, number processing, prenatal alcohol exposure, resting-state functional connectivity, resting-state functional MRI

INTRODUCTION

Fetal alcohol spectrum disorders (FASD) describe the spectrum of neurocognitive and behavioral impairments caused by prenatal alcohol exposure (PAE), including low IQ (Jacobson et al., 2004), deficits in attention (Burden et al., 2005; Kalberg et al., 2006), verbal learning and memory (Kaemingk et al., 2003; Lewis et al., 2015), working memory (Kingdon et al., 2016), executive function (Mattson et al., 2011; Pei et al., 2011), spatial navigation (Dodge et al., 2019), and eyeblink conditioning (Jacobson et al., 2008; Jacobson, Dodge, et al., 2011). Among the PAE-related cognitive deficits, arithmetic (Jacobson, Dodge, et al., 2011) and number processing have been found to be among the most sensitive outcomes and more impaired than reading or spelling (Howell et al., 2006), and deficits in numerosity have been reported as early as infancy (S Jacobson, Stanton, et al., 2011). Notably, these mathematical deficits remained after controlling for IQ (Burden et al., 2005; Chiodo et al., 2004; Jacobson et al., 2004).

Fetal alcohol syndrome (FAS) is the most severe form of FASD and is characterized by small head circumference, growth retardation, and a distinctive pattern of facial anomalies, including short palpebral fissures, a thin upper lip and a flat or smooth philtrum. These facial features have been associated with brain damage that occurs between the 10th and 20th weeks of pregnancy (Renwick & Asker, 1983). Partial fetal alcohol syndrome (PFAS) is the term applied to a child who shows at least two of the three alcohol-related facial anomalies and either a small head circumference, growth retardation (<10th percentile), or neurocognitive or behavioral deficits (Hoyme et al., 2005). Despite these major alcohol-related problems, heavy drinking during pregnancy continues to be prevalent in the United States (May et al., 2018) and endemic areas, including the Western Cape province of South Africa.

Brain lesion and neuroimaging studies have identified two distinct functional neural networks related to number processing: (1) a core quantity system, involving a non-verbal abstract representation of numerical quantity (magnitude and distance) (Dehaene et al., 2003), and (2) mental calculation, involving manipulation of verbally encoded numbers and verbally stored knowledge (e.g., arithmetic facts) (Zago et al., 2001). Within the core quantity system, magnitude comparison occurs bilaterally in the anterior portion of the horizontal segment of the intraparietal sulcus; verbal processing of numbers in the left (L) angular gyrus; and spatial and non-spatial attentional processes contributing to the visual processing of quantity in the bilateral posterior superior parietal lobules. Mental calculation involves the recruitment of a fronto-parietal number processing network that includes the intraparietal sulcus (Simon et al., 2002) and other parietal areas, such as the L angular gyrus, in interaction with an executive brain system that is not specific to number processing (Zago et al., 2008). The executive brain system mediates the integration and management of numerical operations in working memory, response decision and execution, and error monitoring (Gruber et al., 2001).

Functional connectivity changes in the context of mathematical cognition have also been reported. For example, Smith et al. (2009) demonstrated task-related changes in functional connectivity within the fronto-parietal network during mathematical problem-solving; Rosenberg-Lee et al. (2011) reported altered functional connectivity of the intraparietal sulcus during arithmetic compared to rest, and Abramov et al. (2019) demonstrated task-specific changes in functional connectivity within the brain’s attention network following prolonged arithmetic learning.

Several functional MRI (fMRI) studies have reported that PAE alters brain activation in parietal and frontal regions linked to number processing. Adults with fetal alcohol-related dysmorphology exhibited poorer performance on a subtraction task compared to controls, with lower activations in the right (R) inferior and L superior parietal regions and medial frontal gyrus during arithmetic processing (Santhanam et al., 2009). Meintjes et al. (2010) found that children with FAS or PFAS showed greater activation in the L angular gyrus than controls in a proximity judgment task, in which magnitude comparison is assessed by asking the child to determine which of two numbers is closest in magnitude to a third number displayed on the screen. In a follow-up study, Woods et al. (2015) investigated the effects of PAE on brain activation in the five parietal regions identified by Dehaene et al. (2003). An inverse correlation between level of PAE and brain activation was found in the R intraparietal sulcus on the proximity judgment task and a single-digit addition task, and children in the FAS/PFAS group showed greater activation in the L angular gyrus during the proximity judgment task compared to controls and heavily exposed (HE) children. Although these studies found PAE-associated alterations in specific brain regions while performing number processing tasks, no previous study has directly examined whether PAE alters functional connectivity within the parieto-frontal network known to mediate arithmetic calculation.

Resting state functional MRI (rs-fMRI) provides a useful tool to quantify resting state functional connectivity (RSFC) in functional networks. Notably, many of the regions within functional networks correspond to regions that are activated during task-based studies of the relevant functional domain. The “task-free” nature of the technique makes it ideal for pediatric studies since it reduces the potential for task-induced motion and removes the task performance and attention requirements that may obscure task-based neural activation in children.

Seed-based correlation analysis (SCA) is one of several statistical and mathematical methods that have been applied to quantify RSFC. It requires a priori selection of a seed based on previous literature, a hypothesis, or functional activation maps. The seed can be a voxel, cluster, or even an atlas region. The temporal correlations of the blood oxygen level-dependent (BOLD) signals of all other voxels in the brain to this seed provide a measure of the strength of the functional connectivity of every other voxel in the brain to this seed. These correlations are typically represented using color maps, with hot colors showing regions where the BOLD signals are strongly temporally correlated to the seed above some pre-selected threshold, i.e., functionally connected.

The current study tested the hypothesis that the parieto-frontal resting state network (RSN) mediating arithmetic processing is altered by PAE. The specific aims were (1) to perform voxelwise group comparisons of whole-brain SCA maps to identify regions showing PAE-related changes in RSFC to parietal number processing seeds, (2) to examine dose dependence of RSFC in regions showing PAE-related alterations, and (3) to examine the degree to which effects of PAE on arithmetic performance are mediated by alterations in RSFC.

MATERIALS AND METHODS

Participants

Participants were a subset of the mothers and children from our prospectively recruited Cape Town longitudinal cohort. This cohort comprises mothers who were recruited during pregnancy from a historically disadvantaged community in Cape Town, South Africa, between 1999 and 2002 (Jacobson et al., 2008). In this community, the incidence of heavy drinking and FASD is high due to poor socioeconomic circumstances, a history of working on grape and fruit farms, and the dop system, in which workers were paid in part with wine, which together contributed to a tradition of heavy recreational weekend drinking (May et al., 2007).

Mothers were interviewed at recruitment regarding the amount of alcohol they consumed on a day-by-day basis during a typical 2-week period, both around the time of conception and at the time of recruitment, using the timeline follow-back interview (TLFB; Jacobson et al., 2002, 2008). The TLFB interview was repeated in mid-pregnancy and at 1 month postpartum to provide information about drinking in late pregnancy. Volume was recorded for each type of beverage consumed daily, converted to ounces (oz) absolute alcohol (AA), and summarized using three measures: oz AA consumed/day (AA/day), oz AA consumed/drinking occasion, and frequency of drinking days/week. Pregnant women who reported drinking 1.0 oz AA/day or more (≈14 standard drinks/week) or binge drinking (≥5 drinks/occasion) at least twice/month during the first trimester of pregnancy were invited to participate in the study. The next woman presenting at the clinic who abstained or drank only minimally during pregnancy was invited to participate as a control. Exclusionary criteria included maternal age <18 years and chronic medical problems, such as diabetes, epilepsy, or cardiac problems requiring treatment. Infants from multiple births and those with major chromosomal anomalies, seizures, and neural tube defects were also excluded.

A total of 147 infants were enrolled in the Cape Town Longitudinal cohort, of whom 140 (95.2%) were retained through 5-year follow-up; there were two deaths and five relocations away from Cape Town. An additional 37 children were added to the cohort at the time of the 5-year follow-up. These children were born to mothers recruited during pregnancy for two ancillary studies using consistent recruitment interviews and criteria. By the time of the 5-year follow-up, the cohort therefore comprised 177 participants. At the subsequent 10-year follow-up, 162 individuals (91.5% of the 177) were assessed; there was 1 death, 12 relocations, and 2 participants were deemed untestable due to a severe non-alcohol-related intellectual disability. In the present study, rs-fMRI data were acquired around age 11 years in 75 of these children.

Children born to women enrolled in the Cape Town Longitudinal Cohort were assessed for growth and FAS dysmorphology at FASD diagnostic clinics conducted in Cape Town in 2005, 2009, 2013, and 2015 (Hoyme et al., 2005; Jacobson et al., 2008, 2021). Children who did not meet the criteria for FAS or PFAS were classified as either heavily exposed (HE) nonsyndromal or controls, depending on maternal alcohol history. Postnatal lead exposure was obtained from blood samples acquired at 5.1 ± 0.2 years.

In the present sample, data on alcohol use during pregnancy were missing for five women and were estimated from retrospective reports obtained at 5 years postpartum for one child in the FAS/PFAS group and three children in the HE group; maternal alcohol consumption for the mother of one child who met diagnostic criteria for FAS but denied drinking, was estimated as the median of the mothers of the other children with FAS. Smoking during pregnancy was reported as cigarettes smoked/day; cocaine, heroin, methaqualone (“mandrax”), and marijuana use in days/week. Demographic background, including maternal age at delivery and years of education were also obtained.

Ethics

We obtained approval for human research from the ethics committees of Wayne State University and the University of Cape Town (UCT) Faculty of Health Sciences (FHS). Informed consent was obtained from mothers at recruitment and again for each follow-up study. Children provided assent. The children were given a small gift, and the mothers received a photo of their child and compensation consistent with guidelines from the UCT ethics committee.

Cognitive testing

The Wechsler Intelligence Scale for Children-IV (WISC-IV) IQ test was administered at an earlier visit (10.2 ± 0.8 years) to our UCT Child Development Research Laboratory (see Lewis et al., 2015); number processing data were obtained from the WISC-IV Arithmetic subtest.

Neuroimaging protocol

Mothers and children were transported to the Cape Universities Brain Imaging Centre (CUBIC) in our research van. After familiarization with the scanning procedures on a mock scanner, the children were imaged using a 3T Allegra MRI (Siemens, Erlangen, Germany) using a single channel head coil. Rs-fMRI data were acquired using an echo planar imaging (EPI) sequence (resolution = 3.1 × 3.1 × 3.0 mm3, FOV = 200 × 200 × 152 mm3, 34 slices, 180 volumes, TR = 2000 ms, TE = 30 ms, flip angle 90°). T1-weighted structural images were acquired in the sagittal orientation using a 3D EPI-navigated multiecho magnetization prepared rapid gradientecho (MPRAGE) sequence (resolution = 1.3 × 1.0 × 1.3 mm3, FOV = 256 × 256 × 167 mm3, 192 × 256; 128 slices, TR 2530 ms, TI 1100 ms, TE’s 1.53/3.21/4.89/6.57 ms, flip angle 7°) (Tisdall et al., 2012; van der Kouwe et al., 2008).

Pre-processing and statistical analysis

Rs-fMRI data were analyzed using afni_proc.py in AFNI (v18.1.09); the exact command is provided in Data S1 and briefly described here. The first four volumes were discarded to allow for signal stabilization. Processing of the remaining time points included motion correction, alignment to standard space, regression, and blurring. For all subjects, the maximum displacement (as per Power et al., 2012) was below 1.4 mm. The linear affine transform was calculated between subject EPI and anatomical volumes, and the nonlinear warp from subject anatomical to Talairach-Tournoux (TT) standard space was calculated using 3dQwarp. These transforms were combined with motion correction into a single warp, which was applied to the EPI data to bring it into TT space, with a final resolution of 3 × 3 × 3 mm3. Mean white matter (WM) and cerebrospinal fluid (CSF) signals were regressed, as were their derivatives. Spatial smoothing was applied using a Gaussian kernel with a full width at half maximum (FWHM) of 6.0 mm, and the time series was band-pass filtered between 0.01 and 0.1 Hz to reduce the physiological contributions of respiratory and cardiovascular components.

Seed-based correlation maps were calculated from the pre-processed datasets. Seed regions were spherical regions of interest (ROIs; radius = 6 mm) derived in a meta-analysis by Dehaene et al. (2003) that identified peak coordinates of the five parietal regions involved in number processing (Figure 1). For each seed, whole-brain seed-based correlation analysis (SCA) was performed to identify brain regions where the BOLD signal was temporally correlated to that of the seed. Correlation values were Fisher transformed to z-scores to generate SCA maps. Then, using FSL-randomize (v6.0; Smith et al., 2004), we performed voxelwise pairwise group comparisons (FAS/PFAS vs. controls; HE vs. controls) to identify clusters where RSFC to the seeds differed between groups (p < 0.001). AFNI’s 3dFWHMx and 3dClustSim with two-sided thresholding were used to calculate the minimum volume of clusters for significance. These tools, which employ mixed autocorrelation function modeling (ACF) to account for non-Gaussianity in the spatial noise distribution (Cox et al., 2017), indicated that activation clusters of at least 810 mm3 were significant at voxelwise p = 0.001 and clusterwise α = 0.05.

FIGURE 1.

FIGURE 1

Open in a new tab

Regions identified in Dehaene’s meta-analysis that were used as seed regions of interest (adapted from Woods et al., 2015). A, anterior; L, left; P, posterior; R, right.

Further statistical analyses were performed using SPSS software (v. 25; IBM, Armonk, NY). We first examined the variables for normality of distribution. The continuous alcohol measure of oz AA/day was skewed and was, therefore, log transformed (ln (X + 1)). To examine dose-dependence in clusters where the diagnostic groups showed RSFC differences to the seeds, the mean z-scores, fALFF, and ReHo were extracted in each cluster and correlated with the average amount of alcohol their mothers consumed per day across pregnancy (i.e., mean oz AA/day). Notably, fractional ALFF (fALFF) reflects the relative contribution of the Low-Frequency Fluctuation (LFF) band to the entire observed frequency spectrum, while regional homogeneity (ReHo) assesses the similarity or synchronization between the time series of a given voxel and its nearest neighbors. We considered six control variables as potential confounders: child sex, age at scan, postnatal lead exposure and maternal age at delivery, years of education, and cigarettes smoked/day during pregnancy. Control variables also included total gray matter (GM) volume and the maximum displacement during the scan. Any missing sociodemographic data were estimated using the median for the relevant exposure group (see Table 1). Any control variables that were associated at least weakly with the cluster’s mean RSFC z-score (at p < 0.10) were included in a linear regression of that cluster’s RSFC z-score on AA/day to see if the effect of AA/day on the RSFC z-score remained significant after control for confounding. We subsequently reanalyzed the data omitting 5 children whose mothers reported using methaqualone or marijuana during pregnancy, to see if the effects persisted.

TABLE 1.

Sample characteristics (N = 57).

FAS/PFAS HE Control
N 19 19 19 F or χ2
Child characteristics
 Sex (% male) 47 42 26 1.93
 Age at rs-fMRI scan (year) 11.0 (1.0) 11.4 (0.7) 11.6 (1.0) 2.03
 Lead exposure (μg/L)a 9.9 (4.4) 8.0 (2.7) 7.5 (2.7) 2.91†
 WISC-IV Full-Scale IQ 66.2 (9.5) 76.6 (15.1) 76.9 (13.1) 4.31*
 WISC-IV Arithmetic subtest 6.5 (2.6) 8.3 (2.7) 8.9 (2.8) 4.04*
 Total gray matter volume (×105 mm3)b 6.6 (0.5) 7.2 (0.7) 6.8 (0.7) 3.57*
 Maximum displacement (mm) 0.7 (0.2) 0.7 (0.3) 0.8 (0.3) 0.82
Maternal characteristics
 Maternal age at delivery (year) 29.1 (7.6) 24.8 (4.7) 26.2 (4.6) 2.72†
 Education (year)c 8.2 (2.7) 9.0 (2.1) 10.1 (1.7) 3.51*
 oz absolute alcohol consumed/day across pregnancy (AA/day)d 0.9 (0.8) 0.8 (0.9) 0.0 (0.0) 10.10***
 oz absolute alcohol consumed/occasion across pregnancy (AA/occasion)d 4.0 (2.0) 3.9 (3.2) 0.1 (0.3) 20.01***
 Number of drinking days/week across pregnancy (frequency)d 1.5 (1.0) 1.2 (0.9) 0.0 (0.2) 20.41***
 Cigarettes smoked/day during pregnancye 6.9 (6.1) 5.4 (3.7) 3.8 (9.9) 4.1*
Open in a new tab

Note: FAS/PFAS group: 10 children with fetal alcohol syndrome (FAS) and 9 with partial FAS (PFAS). 1 oz AA ≈ 2 standard drinks. Values are mean (SD). Abbreviations: AA, absolute alcohol; HE, nonsyndromal heavily exposed group; WISC-IV, Wechsler Intelligence Scale for Children, 4th Edition.

a

Missing value estimated at group median for one child with FAS/PFAS.

b

One (control) outlier (3 SD beyond the mean) was recoded to 1 point lower than the lowest observed non-outlier group value (Winer, 1971).

c

One (FAS/PFAS) outlier was recoded to 1 point lower than the lowest observed non-outlier group value.

d

Maternal alcohol consumption during pregnancy was estimated from retrospective data obtained at 5 years postpartum for one mother with a child in the FAS/PFAS group and three with children in the HE group; alcohol consumption for one mother who denied drinking but had a child who met criteria for FAS was estimated at the median of the other mothers of the children with FAS.

e

One outlier (control) was recoded to 1 point higher than the highest observed non-outlier group value (Winer, 1971).

†

p < 0.10,

*

p < 0.05,

***

p < 0.001.

Finally, we tested the hypothesis that the lower RSFC to parietal seeds involved in number processing mediate the effect of PAE on number processing outcomes in a series of hierarchical multiple regressions. We entered AA/day during pregnancy in the first step and the RFSC z-score in the second step of each regression analysis. Mediation was inferred if the addition of the RSFC z-score substantially reduced the magnitude of the regression coefficient for AA/day. We used the Difference in Coefficients Test (Clogg et al., 1992) to test whether the reduction in the magnitude of the regression coefficient was statistically significant. In a Monte Carlo study comparing 14 different methods to test the statistical significance of mediation hypotheses, the Clogg test was found to be one of two with the greatest power (MacKinnon et al., 2002).

RESULTS

Sample characteristics

Neuroimaging data from 18 children were excluded due to ghosting artifacts. The mean age of the remaining 57 children was 11.3 ± 0.9 years (mean age ± SD). The sample was comprised of 10 children with FAS, 9 with PFAS, 19 HE, and 19 controls; the data from children with FAS or PFAS were pooled in the analyses to generate comparable size groups. Handedness assessed on the Edinburgh Behavioral Handedness Inventory (EHI; Oldfield, 1971) did not differ between groups (F(2, 54) < 0.26, p = 0.769). The sample characteristics are summarized in Table 1.

By design, there were no between-group differences in children’s age at scan. Sex, motion during the scan, and lead exposure were also similar. Mothers in the FAS/PFAS consumed more cigarettes during pregnancy. The FAS/PFAS group had lower WISC-IV IQ and arithmetic scores than the nonsyndromal HE and control groups. The total GM volume of the children in the FAS/PFAS group was lower compared to that of the HE children but not the controls. There was a tendency for lead exposure to be slightly more elevated in children in the FAS/PFAS group than controls. Mothers of children in the FAS/PFAS group also tended to be somewhat older than HE and control mothers. As expected, maternal alcohol consumption was higher in the FAS/PFAS and HE groups compared to controls. Except for one mother who reported drinking 1 drink on 4 occasions, the rest of the control mothers reported abstaining from alcohol use during pregnancy. One mother of a child with FAS used methaqualone; mothers of two children with PFAS used marijuana or methaqualone; while mothers of two HE children used both.

Group comparisons of resting state functional connectivity maps from SCA

To assess whether PAE alters functional connectivity to the parietal seeds, we performed voxelwise pairwise group comparisons of the generated SCA maps. These comparisons revealed significantly lower RSFC between the seed in the right (R) posterior superior parietal lobule and a cluster in the left (L) inferior frontal gyrus in children with FAS/PFAS compared to controls (Figure 2). Lower RSFC was also seen in children with FAS/PFAS compared to controls between the seed in the R horizontal intraparietal sulcus and clusters in both the R precentral gyrus and L cerebellar lobule VI. The sizes and peak coordinates of each of these clusters, as well as group mean RSFCs within these clusters and Pearson correlations of RSFC measures with the average daily alcohol consumption of mothers across pregnancy, are shown in Table 2 and Figure S1. Since z-scores showed stronger associations with levels of prenatal alcohol exposure than both fALFF and ReHo, only z-scores were used in subsequent analyses.

FIGURE 2.

FIGURE 2

Open in a new tab

Whole-brain seed-based correlation analysis maps. Hot colors show all regions with significant functional connectivity (thresholded at z > 2.3) to the relevant parietal number processing seed. Blue clusters are regions where children with FAS/PFAS demonstrate lower functional connectivity to the relevant seed than controls. Cross-hairs indicate the peak coordinates. (A) A cluster in the left inferior frontal gyrus shows lower connectivity to a seed in the right posterior superior parietal lobule. Clusters in (B) right precentral gyrus and (C) left cerebellum both show lower connectivity to a seed in the right horizontal intraparietal sulcus. A, anterior; L, left; P, posterior; R, right.

TABLE 2.

Cluster sizes and peak coordinates (in TT standard space) of regions where children with FAS/PFAS demonstrate lower RSFC in parietal seeds than controls. We report group means of RSFC z-scores, fALFF and ReHo within these clusters, as well as Pearson correlations of RSFC measures with the average amount of alcohol mothers consumed across pregnancy.

Seed Mean (SD) of RSFC measures within each cluster
Location of cluster Size (mm3) Peak coordinates (RPI coordinate; mm) RSFC FAS/PFAS HE Controls r
R posterior superior parietal lobule
 L inferior frontal gyrusa 1296 −19.5, 37.5, −15.5 z-score 0.06 (0.14) 0.06 (0.15) 0.31 (0.20) −0.37**
fALFF 0.48 (0.05) 0.51 (0.06) 0.57 (0.08) −0.30*
ReHo 0.52 (0.09) 0.53 (0.09) 0.61 (0.08) −0.29*
R horizontal intraparietal sulcus
 R precentral gyrusb 1431 10.5, −31.5, 68.5 z-score 0.10 (0.15) 0.04 (0.16) 0.33 (0.22) −0.34**
fALFF 0.50 (0.04) 0.53 (0.04) 0.56 (0.04) −0.27*
ReHo 0.52 (0.07) 0.53 (0.06) 0.60 (0.06) −0.24†
 L cerebellar lobule VI 864 −22.5, −67.5, −24.5 z-score 0.01 (0.11) 0.02 (0.10) 0.19 (0.10) −0.43**
fALFF 0.46 (0.03) 0.50 (0.05) 0.50 (0.39) −0.29*
ReHo 0.51 (0.06) 0.56 (0.05) 0.56 (0.04) −0.27*
Open in a new tab

Note: FAS/PFAS: 10 children with fetal alcohol syndrome (FAS) and 9 with partial FAS (PFAS).

Abbreviations: AA/day, oz absolute alcohol consumed/day across pregnancy; HE, nonsyndromal heavily exposed; L, left; R, right; RSFC, resting-state functional connectivity; TT, Talairach-Tournoux.

a

A similar cluster (1269 mm3; peak coordinates: −13.5, 37.5, −12.5) showed lower RSFC in HE compared to control children.

b

A similar cluster (891 mm3; peak coordinates: 4.5, −28.5, 71.5) showed lower RSFC in HE compared to control children.

†

p < 0.10

*

p < 0.05

**

p < 0.01.

The less affected non-syndromal HE children demonstrated lower RSFC compared to controls between the selected parietal seeds and two clusters: a cluster in the L inferior frontal gyrus demonstrated lower RSFC with the seed in the R superior parietal lobule, and a cluster in the R precentral gyrus with the seed in the R horizontal intraparietal sulcus. Since both of these clusters overlapped with those seen in the more severely affected children with FAS/PFAS, they were not investigated separately further.

Re-running the analyses with maximum displacement included as a predictor in FSL_randomize did not significantly alter the results. Results were also essentially unchanged when the analyses were re-run omitting the children whose mothers reported using marijuana and/or methaqualone, or the children for whom maternal drinking histories during pregnancy were missing. Cluster sizes and peak coordinates from these analyses are presented in Tables S1 and S2. The largest difference in cluster size was 54 mm3, which corresponds to only 2 voxels (voxel size: 3 × 3 × 3 mm3). Additionally, peak coordinates differed by just 3 mm (1 voxel) in any direction.

Table 3 presents the correlations of each of the control variables with the mean z-scores in the three clusters that demonstrated lower RSFC for parietal seeds in the FAS/PFAS group. As shown in Table 2 and Figure 3, the mean RSFC z-scores in all three clusters were associated with the amount of alcohol the mother consumed during pregnancy; associations remained significant after adjustment for potential confounding.

TABLE 3.

Correlation of each of the eight control variables with mean RSFC z-scores in the clusters showing alcohol-related reductions in connectivity to parietal seeds.

Seed
Child age Lead Maximum Total GM Maternal Smoking during
Location of cluster Child sex at scan exposure displacement volume Maternal age education pregnancy
R posterior superior parietal lobule
 L inferior frontal gyrus 0.01 0.01 −0.05 0.35** 0.20 0.02 0.04 −0.05
R horizontal intraparietal sulcus
 R precentral gyrus −0.02 −0.06 −0.02 0.36** 0.17 0.02 −0.01 −0.01
 L cerebellar lobule VI 0.07 0.12 −0.12 0.05 −0.04 0.02 0.24† −0.22†
Open in a new tab

Note: Values are Pearson r’s.

Abbreviations: GM, gray matter; L, left; R, right.

†

p < 0.10

**

p < 0.01.

FIGURE 3.

FIGURE 3

Open in a new tab

Plots showing mean RSFC z-scores as a function of ounces absolute alcohol consumed per day (AA/day) across pregnancy for the three clusters where we found prenatal alcohol exposure-related connectivity decreases to parietal seeds. r is the simple Pearson correlation and β is the standardized regression coefficient after adjustment for potential confounding by amaximum displacement and bmaternal education and smoking, respectively. FAS/PFAS, combined fetal alcohol syndrome (FAS) and partial FAS (PFAS) group; HE, nonsyndromal heavily exposed group.

Mediation of effects of PAE on WISC-IV arithmetic

To examine whether functional connectivity impairment mediates the effects of PAE on arithmetic performance, we performed mediation analyses. The results of the analyses for each of the three clusters showing functional connectivity impairment are presented in Figure 4. Increasing levels of PAE were related to lower WISC-IV arithmetic scores; however, the correlation fell just below conventional levels of significance after controlling for Full Scale IQ (β = −0.16, p = 0.075). Higher RSFC between the R posterior superior parietal lobule and L inferior frontal gyrus and between the R horizontal intraparietal sulcus and R precentral gyrus were significantly associated with higher WISC-IV Arithmetic scores. By contrast, RSFC between the R horizontal intraparietal sulcus and L cerebellar lobule VI was unrelated to WISC-IV Arithmetic. In our mediation analysis, the standardized regression coefficient for AA/day in relation to WISC-IV Arithmetic (τ = −0.26) decreased significantly when either RSFC between the R posterior superior parietal lobule and the L inferior frontal gyrus (τ′ = −0.12) or between the R horizontal intraparietal sulcus and R precentral gyrus (τ′ = −0.15) were entered into the regression. Thus, lower connectivity between these regions partially mediated the effect of PAE on arithmetic performance.

FIGURE 4.

FIGURE 4

Open in a new tab

Path model illustrating mediation of the effect of daily absolute alcohol consumption across pregnancy on WISC-IV Arithmetic by lower resting state functional connectivity to parietal seeds involved in number processing.

DISCUSSION

We previously identified localized dose-dependent RSFC deficits using independent components and dual regression analysis in children with PAE in five resting state networks—default mode, salience, ventral attention, dorsal attention, and right executive control (Fan et al., 2017). In view of number processing being a key deficit in FASD, we wanted to directly examine the effects of PAE on functional connectivity in and between regions known to be involved in number processing. We therefore performed SCA for seeds in five key parietal number processing regions, namely the bilateral posterior superior parietal lobules, horizontal intraparietal sulci, and left angular gyrus. Voxelwise group comparisons revealed lower RSFC between the seed in the R posterior superior parietal lobule and the L inferior frontal gyrus, and between the seed in the R horizontal intraparietal sulcus and clusters in both the R precentral gyrus and L cerebellar lobule VI, in children in the FAS/PFAS group compared to controls. Notably, HE children showed lower RSFC compared to controls in the same two fronto-parietal connections.

The bilateral posterior superior parietal lobules, which are activated during counting (Piazza et al., 2002) and visual–spatial tasks, are believed to support the engagement of attention during number processing (Dehaene et al., 2003). Here, we found reduced RSFC of the R superior parietal lobule to the L inferior frontal gyrus. The inferior frontal gyri have been shown to be activated during approximate numerosity judgment (Kovas et al., 2009), addition (Molko et al., 2003), approximate calculation (Kucian et al., 2006), and to demonstrate a numeric distance effect (greater activation when the numbers compared are closer together) (Gullick et al., 2011; Kucian et al., 2011). Notably, decreased activation has been reported in both the posterior superior parietal lobule and the inferior frontal gyrus during number processing in children with PAE compared to controls (Santhanam et al., 2009; Woods et al., 2018). Given that activation of the inferior frontal gyrus during number processing has been attributed to visual–spatial working memory (Kucian et al., 2011) and language processing (Fulbright et al., 2003), the weaker functional connectivity seen in this region may reflect poorer verbal processing and visual–spatial working memory during number processing in alcohol-exposed children. Notably, mental arithmetic assessed with the WISC-IV requires significant attention and working memory—both of which are known to be impacted by PAE (Coles et al., 2011; Mattson et al., 2011).

It is of interest that both the inferior frontal gyrus and superior parietal lobule have been implicated in structural MRI studies of PAE, demonstrating thicker cortex (Yang et al., 2012) and smaller volumes (Archibald et al., 2001; Chen et al., 2012), respectively, in those with FASD compared to controls. These abnormalities may play a role in the number processing (Santhanam et al., 2009) and spatial and verbal working memory (Roussotte et al., 2012; Spadoni et al., 2007) impairments observed in FASD, and contribute to the lower RSFC between these regions.

Lower RSFC between the R horizontal intraparietal sulcus and the R precentral gyrus was also seen in both syndromal and nonsyndromal heavily exposed children compared to controls. The R horizontal intraparietal sulcus is involved in the nonverbal representation of quantity. This area is hypothesized to support number processing regardless of notation, i.e., whether representation is comprised of symbolic (e.g., Arabic numbers) or nonsymbolic (e.g., numbers of dots) sequences (Dehaene et al., 2003). The precentral gyri are bilaterally activated during magnitude comparison (Kaufmann et al., 2005) and in sensorimotor processes during the visual identification of Arabic digits (Pinel et al., 1999).

We found an adverse effect of PAE on activation of the R horizontal intraparietal sulcus during number processing in our previous studies focusing on the same parietal regions used as seeds in this study: increasing PAE was related to weaker activation in the R horizontal intraparietal sulcus during proximity judgment and single digit addition (Woods et al., 2015) and, during nonsymbolic number comparison, controls activated this region more strongly than the FAS/PFAS group (Woods et al., 2018). In a whole-brain fMRI analysis, we found a PAE-dependent response in the R horizontal intraparietal sulcus, with controls showing more activity during a proximity judgment task (Meintjes et al., 2010). During a subtraction task, adults with alcohol-related dysmorphology showed weaker activation in the R inferior parietal region, including the horizontal intraparietal sulcus (Santhanam et al., 2009). Involvement of the precentral gyrus during number processing has been attributed to the role of finger representation in the development of numerical cognition (Andres et al., 2008). Finger counting is an early numerical ability and establishes an external aid to represent numbers (Lafay et al., 2013). The lower RSFC between the R horizontal intraparietal sulcus and the R precentral gyrus in children with PAE may therefore indicate impaired finger-based numerical representation, which has been linked to poorer arithmetic skills in children (Moeller et al., 2011).

In contrast to the two previous connections, lower RSFC between the R horizontal intraparietal sulcus and L cerebellar lobule VI was only seen in the FAS/PFAS group. During a single digit addition task, Meintjes et al. (2010) reported that children with FAS or PFAS showed higher functional activation than controls in a region of the L cerebellar cortex (TT coordinates: −24, −44, −18) that is in close proximity to the cluster within L lobule VI found in the current study (TT coordinates: −22.5, −67.5, −24.5). Notably, previous studies also only found cerebellar deficits in the FAS/PFAS group (Fan et al., 2015, 2017; Meintjes et al., 2010; Spottiswoode et al., 2011).

The cerebellum has been found to be involved in more challenging number processing tasks (Kaufmann et al., 2006), as well as in tasks that target language, spatial processing, working memory, and executive functioning (Stoodley & Schmahmann, 2009). The cerebellum was found to be activated during a task of non-symbolic numerosity estimation in only high-ability children (Kovas et al., 2009). It is possible that the cerebellum is used more by control children with better math ability, resulting in a stronger connection between the core math R horizontal intraparietal sulcus region and the cerebellum.

Other studies have also found more extensive behavioral and neural impairment in the syndromal FAS and PFAS children than in the nonsyndromal HE group. For example, the cerebellum plays a central role in delay and trace eyeblink conditioning, which we found to be most impaired in children with FAS compared to nonsyndromal heavily alcohol exposed and control children (Jacobson et al., 2008; Jacobson, Stanton, et al., 2011). Fan et al. (2016, 2017) reported that in nonsyndromal HE children connectivity differences, both structural and functional, were smaller and evident in only a subset of regions affected in children with FAS or PFAS. Similarly, during nonsymbolic number comparison, of the regions affected in children with FAS and PFAS, some were functionally affected in the nonsyndromal heavily exposed children, while in other regions the functioning in these children was apparently spared (Woods et al., 2018).

Because detailed maternal drinking histories during pregnancy were well documented and ascertained prospectively in this cohort, a novel aspect of our study was the ability to quantitatively relate the extent of PAE to RSFC alterations during number processing. The continuous measure of PAE, oz AA/day, was associated with RSFC between the R posterior superior parietal lobule and the L inferior frontal gyrus and between the R horizontal intraparietal sulcus and both the R precentral gyrus and L lobule VI, suggesting dose-dependent effects, which remained significant after controlling for potential confounders. Moreover, the RSFCs between the parietal seeds and the L inferior frontal gyrus and the R precentral gyrus were associated with WISC-IV Arithmetic scores, and multiple regression indicated that these RSFC deficits partially mediated adverse effects of PAE on arithmetic.

All neuroimaging assessments, especially in pediatric studies, are limited by subject motion and related imaging artifacts. Due to the low resolution of rs-fMRI acquisitions, voxelwise analysis is limited by potential co-registration errors. In our study, all rs-fMRI data were visually checked during processing and care was taken to ensure good co-registration. To control for Type 1 error, we determined the minimum size of clusters for significance in each RSN using the new “mixed ACF” methodology, which accounts for non-Gaussianity in the spatial noise distribution (Cox et al., 2017). This approach limits our ability to identify small regions with GM abnormalities. Although we considered potential confounding by eight important variables—three child characteristics, three maternal characteristics, and total gray matter volume and maximum displacement during the scan—it is not possible to control for all extraneous variables that might provide alternative explanations for observed effects. The parietal seeds used in this study are based on those cited in Dehaene et al. (2003), and further studies might consider other regions related to number processing. It is uncertain whether the weaker RSFC seen here is a result of gray matter damage in the seed or target regions, or impairment of the white matter tracts linking the regions. Previously, we found gray matter connectivity deficits that appeared to be related to deficits in white matter tracts providing connections within resting state networks (Fan et al., 2017). Future studies should consider using tractography to determine whether the RSFC differences observed here are due to gray or white matter changes.

CONCLUSIONS

This study examined the effects of PAE on whole-brain functional connectivity in parietal seeds known to play a crucial role during number processing. Three connections demonstrated dose-dependent PAE-related reductions in RSFC—two fronto-parietal connections and one between the R horizontal intraparietal sulcus and L cerebellar lobule VI. Lower RSFC in both the fronto-parietal connections partially mediated the effect of PAE on arithmetic performance. The RSFC analyses presented here advance our understanding by revealing PAE-related impairment in neural circuits involved in number processing. The weaker connectivity between the R posterior superior parietal lobule and the L inferior frontal gyrus in the children with PAE suggests less proficiency in using verbal processing and visual–spatial working memory in conjunction with number processing, and the weaker connectivity between the R horizontal intraparietal sulcus and the R precentral gyrus suggests poorer finger-based numerical representation, which has been linked to computational skills in children. These findings therefore suggest fetal alcohol-related deficits in two distinct aspects of mathematical processing.

RSFC was more impaired in the more severely affected children with FAS/PFAS than in the HE group. Even though WISC-IV Arithmetic performance appears to have been spared in the HE group, the observed RSFC deficits would likely be associated with impairment in more challenging math and/or other aspects of cognitive function. Future WM tractography studies focusing on the structural connectivity between the clusters identified in this study and the parietal seeds may provide additional insight into the mechanisms underlying PAE-related number processing deficits in individuals with FASD. Future studies could employ graph theory to examine resting state networks and aim to identify behavioral scores or sub-tests that correlate with network measures. Such an analysis would offer valuable insights into the functional implications of altered RSFC patterns in individuals affected by FAS/PFAS. Moreover, it is crucial for future studies to explore potential treatment interventions aimed at improving outcomes for affected individuals. This may entail investigating strategies to modulate components of the mediation model identified in our study, thereby enhancing cognitive functioning and behavioral outcomes.

Supplementary Material

Appendices

ACKNOWLEDGMENTS

The authors thank the Cape Universities Brain Imaging Centre (CUBIC) radiographers Marie-Louise de Villiers and Nailah Maroof and our University of Cape Town and Wayne State University research staff Nicolette Hamman, Mariska Pienaar, Maggie September, Emma Makin, Neil Dodge, and Renee Sun. We thank the three FASD dysmorphologists, H. Eugene Hoyme (HEH), Luther K. Robinson (LKR), and Nathaniel Khaole, who conducted the examinations of the children during our 2005 FASD dysmorphology clinic; HEH and LKR, who conducted the examinations in our 2009 diagnostic clinic; and HEH who led the dysmorphology team in 2013 and 2016. We also greatly appreciate the contributions of the mothers and children who have participated in our Cape Town Longitudinal Cohort research.

FUNDING INFORMATION

This work was funded by grants from the NIH/National Institute on Alcohol Abuse and Alcoholism (R01 AA016781, R21 AA017410, R03 TW007030 and U01 AA014790); DST/ National Research Foundation (NRF); (NRF) South African Research Chairs Initiative; NRF Focus Area grant FA2005040800024; and grants from the Lycaki-Young Fund from the State of Michigan; and the NRF Free-standing Postdoctoral, NRF Thuthuka, and Harry Crossley Clinical Research Fellowships.

Funding information

National Institutes of Health, Grant/Award Number: R01 AA016781, R03 TW007030, R21 AA017410 and U01 AA014790; National Research Foundation; (NRF) South African Research Chairs Initiative; Grant/Award Number: FA2005040800024; Lycaki-Young Fund from the State of Michigan; Grant/Award Number: 99625; Grant/Award Number: 129832; Harry Crossley Clinical Research Fellowships

Footnotes

CONFLICT OF INTEREST STATEMENT

The authors declare no competing financial interests.

SUPPORTING INFORMATION

Additional supporting information can be found online in the Supporting Information section at the end of this article.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

REFERENCES

  1. Abramov DM, Cunha CQ, Galhanone PR, Alvim RJ, de Oliveira AM & Lazarev VV (2019) Neurophysiological and behavioral correlates of alertness impairment and compensatory processes in ADHD evidenced by the Attention Network Test. PLoS One, 14(7), e0219472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andres M, Olivier E. & Badets A. (2008) Actions words and numbers: a motor contribution to semantic processing? Current Directions in Psychological Science, 17(4), 313–317. [Google Scholar]
  3. Archibald SL, Fennema-Notestine C, Gamst A, Riley EP, Mattson SN & Jernigan TL (2001) Brain dysmorphology in individuals with severe prenatal alcohol exposure. Developmental Medicine and Child Neurology, 43(3), 148–154. [PubMed] [Google Scholar]
  4. Burden MJ, Jacobson SW & Jacobson JL (2005) Relation of prenatal alcohol exposure to cognitive processing speed and efficiency in childhood. Alcoholism: Clinical and Experimental Research, 29(8), 1473–1483. [DOI] [PubMed] [Google Scholar]
  5. Chen X, Coles CD, Lynch ME & Hu X. (2012) Understanding specific effects of prenatal alcohol exposure on brain structure in young adults. Human Brain Mapping, 33(7), 1663–1676. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chiodo LM, Jacobson SW & Jacobson JL (2004) Neurodevelopmental effects of postnatal lead exposure at very low levels. Neurotoxicology and Teratology, 26(3), 359–371. [DOI] [PubMed] [Google Scholar]
  7. Clogg CC, Petkova E. & Shihadeh ES (1992) Statistical methods for analyzing collapsibility in regression models. Journal of Educational Statistics, 17(1), 51–74. [Google Scholar]
  8. Coles CD, Goldstein FC, Lynch ME, Chen X, Kable JA, Johnson KC et al. (2011) Memory and brain volume in adults prenatally exposed to alcohol. Brain and Cognition, 75, 67–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cox RW, Chen G, Glen DR, Reynolds RC & Taylor PA (2017) fMRI clustering in AFNI: false-positive rates redux. Brain Connectivity, 7(3), 152–171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dehaene S, Piazza M, Pinel P. & Cohen L. (2003) Three parietal circuits for number processing. Cognitive Neuropsychology, 20(3–6), 487–506. [DOI] [PubMed] [Google Scholar]
  11. Dodge NC, Thomas KGF, Meintjes EM, Molteno CD, Jacobson JL & Jacobson SW (2019) Spatial navigation in children and young adults with fetal alcohol spectrum disorders. Alcoholism: Clinical and Experimental Research, 43(12), 2536–2546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fan J, Jacobson SW, Taylor PA, Molteno CD, Dodge NC, Stanton ME et al. (2016) White matter deficits mediate effects of prenatal alcohol exposure on cognitive development in childhood. Human Brain Mapping, 37(8), 2943–2958. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fan J, Meintjes EM, Molteno CD, Spottiswoode BS, Dodge NC, Alhamud AA et al. (2015) White matter integrity of the cerebellar peduncles as a mediator of effects of prenatal alcohol exposure on eyeblink conditioning. Human Brain Mapping, 36(7), 2470–2482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fan J, Taylor PA, Jacobson SW, Molteno CD, Gohel S, Biswal BB et al. (2017) Localized reductions in resting-state functional connectivity in children with prenatal alcohol exposure. Human Brain Mapping, 38(10), 5217–5233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fulbright RK, Manson SC, Skudlarski P, Lacadie CM & Gore JC (2003) Quantity determination and the distance effect with letters, numbers and shapes: a functional MR imaging study of number processing. American Journal of Neuroradiology, 24(2), 193–200. [PMC free article] [PubMed] [Google Scholar]
  16. Gruber O, Indefrey P, Steinmetz H. & Kleinschmidt A. (2001) Dissociating neural correlates of cognitive components in mental calculation. Cerebral Cortex, 11(4), 350–359. [DOI] [PubMed] [Google Scholar]
  17. Gullick MM, Sprute LA & Temple E. (2011) Individual differences in working memory, nonverbal IQ, and mathematics achievement and brain mechanisms associated with symbolic and nonsymbolic number processing. Learning and Individual Differences, 21(6), 644–654. [Google Scholar]
  18. Howell KK, Lynch ME, Platzman KA, Smith GH & Coles CD (2006) Prenatal alcohol exposure and ability, academic achievement, and school functioning in adolescence: a longitudinal follow-up. Journal of Pediatric Psychology, 31(2), 116–126. [DOI] [PubMed] [Google Scholar]
  19. Hoyme HE, May PA, Kalberg WO, Kodituwakku P, Gossage JP, Trujillo PM et al. (2005) A practical clinical approach to diagnosis of fetal alcohol spectrum disorders: clarification of the 1996 Institute of Medicine criteria. Pediatrics, 115(1), 39–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jacobson JL, Dodge NC, Burden MJ, Klorman R. & Jacobson SW (2011) Number processing in adolescents with prenatal alcohol exposure and ADHD: differences in the neurobehavioral phenotype. Alcoholism: Clinical and Experimental Research, 35(3), 431–442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jacobson SW, Chiodo LM, Sokol RJ & Jacobson JL (2002) Validity of maternal report of alcohol, cocaine, and smoking during pregnancy in relation to infant neurobehavioral outcome. Pediatrics, 109(5), 815–825. [DOI] [PubMed] [Google Scholar]
  22. Jacobson SW, Hoyme HE, Carter RC, Dodge NC, Molteno CD, Meintjes EM et al. (2021) Evolution of the physical phenotype of fetal alcohol spectrum disorders from childhood through adolescence. Alcoholism: Clinical and Experimental Research, 45(2), 395–408. [DOI] [PubMed] [Google Scholar]
  23. Jacobson SW, Jacobson JL, Sokol RJ, Chiodo LM & Corobana R. (2004) Maternal age, alcohol abuse history, and quality of parenting as moderators of the effects of prenatal alcohol exposure on 7.5-year intellectual function. Alcoholism: Clinical and Experimental Research, 28(11), 1732–1745. [DOI] [PubMed] [Google Scholar]
  24. Jacobson SW, Stanton ME, Dodge NC, Pienaar M, Fuller DS, Molteno CD et al. (2011) Impaired delay and trace eyeblink conditioning in school-age children with fetal alcohol syndrome. Alcoholism: Clinical and Experimental Research, 35(2), 250–264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jacobson SW, Stanton ME, Molteno CD, Burden MJ, Fuller DS, Hoyme HE et al. (2008) Impaired eyeblink conditioning in children with fetal alcohol syndrome. Alcoholism: Clinical and Experimental Research, 32(2), 365–372. [DOI] [PubMed] [Google Scholar]
  26. Kaemingk KL, Mulvaney S. & Halverson PT (2003) Learning following prenatal alcohol exposure: performance on verbal and visual multi-trial tasks. Archives of Clinical Neuropsychology, 18(1), 33–47. [PubMed] [Google Scholar]
  27. Kalberg WO., Provost B., Tollison SJ., Tabachnick BG., Robinson LK., Hoyme HE. et al. (2006) Comparison of motor delays in young children with fetal alcohol syndrome to those with prenatal alcohol exposure and with no prenatal alcohol exposure. Alcoholism: Clinical and Experimental Research, 30(12), 2037–2045. [DOI] [PubMed] [Google Scholar]
  28. Kaufmann L, Koppelstaetter F, Delazer M, Siedentopf C, Rhomberg P, Golaszewski S. et al. (2005) Neural correlates of distance and congruity effects in a numerical Stroop task: an event-related fMRI study. NeuroImage, 25(3), 888–898. [DOI] [PubMed] [Google Scholar]
  29. Kaufmann L, Koppelstaetter F, Siedentopf C, Haala I, Haberlandt E, Zimmerhackl L-B et al. (2006) Neural correlates of the number–size interference task in children. Neuroreport, 17(5), 587–591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kingdon D, Cardoso C. & McGrath JJ (2016) Executive function deficits in fetal alcohol spectrum disorders and attention-deficit/hyperactivity disorder – a meta-analysis. Journal of Child Psychology and Psychiatry, 57(2), 116–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kovas Y, Giampietro V, Viding E, Ng V, Brammer M, Barker GJ et al. (2009) Brain correlates of non-symbolic numerosity estimation in low and high mathematical ability children. PLoS One, 4(1), e4587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kucian K, Loenneker T, Dietrich T, Dosch M, Martin E. & von Aster M. (2006) Impaired neural networks for approximate calculation in dyscalculic children: a functional MRI study. Behavioral and Brain Functions, 2, 31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kucian K, Loenneker T, Martin E. & von Aster M. (2011) Non-symbolic numerical distance effect in children with and without developmental dyscalculia: a parametric fMRI study. Developmental Neuropsychology, 36(6), 741–762. [DOI] [PubMed] [Google Scholar]
  34. Lafay A, Thevenot C, Castel C. & Fayol M. (2013) The role of fingers in number processing in young children. Frontiers in Psychology, 4, 488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Lewis CE, Thomas KG, Dodge NC, Molteno CD, Meintjes EM, Jacobson JL et al. (2015) Verbal learning and memory impairment in children with fetal alcohol spectrum disorders. Alcoholism: Clinical and Experimental Research, 39(4), 724–732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. MacKinnon DP, Lockwood CM, Hoffman JM, West SG & Sheets V. (2002) A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7(1), 83–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mattson SN, Crocker N. & Nguyen TT (2011) Fetal alcohol spectrum disorders: neuropsychological and behavioral features. Neuropsychology Review, 21, 81–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. May PA, Chambers CD, Kalberg WO, Zellner J, Feldman H, Buckley D. et al. (2018) Prevalence of fetal alcohol spectrum disorders in 4 US communities. JAMA, 319(5), 474–482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. May PA, Gossage JP, Marais AS, Adnams CM, Hoyme HE, Jones KL et al. (2007) The epidemiology of fetal alcohol syndrome and partial FAS in a South African community. Drug and Alcohol Dependence, 88(2–3), 259–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Meintjes EM, Jacobson JL, Molteno CD, Gatenby JC, Warton C, Cannistraci CJ et al. (2010) An fMRI study of number processing in children with fetal alcohol syndrome. Alcoholism: Clinical and Experimental Research, 34(8), 1450–1464. [DOI] [PubMed] [Google Scholar]
  41. Moeller K, Martignon L, Wessolowski S, Engel J. & Nuerk HC (2011) Effects of finger counting on numerical development – the opposing views of neurocognition and mathematics education. Frontiers in Psychology, 2, 328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Molko N, Cachia A, Riviere D, Mangin JF, Bruandet M, Le Bihan D. et al. (2003) Functional and structural alterations of the intraparietal sulcus in a developmental dyscalculia of genetic origin. Neuron, 40(4), 847–858. [DOI] [PubMed] [Google Scholar]
  43. Oldfield RC (1971) The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia, 9(1), 97–113. [DOI] [PubMed] [Google Scholar]
  44. Pei J, Job J, Kully-Martens K. & Rasmussen C. (2011) Executive function and memory in children with fetal alcohol spectrum disorder. Child Neuropsychology, 17(3), 290–309. [DOI] [PubMed] [Google Scholar]
  45. Piazza M, Mechelli A, Butterworth B. & Price CJ (2002) Are subitizing and counting implemented as separate or functionally overlapping processes? NeuroImage, 15(2), 435–446. [DOI] [PubMed] [Google Scholar]
  46. Pinel P, Le Clec’H G, van de Moortele P-F, Naccache L, Le Bihan D. & Dehaene S. (1999) Event-related fMRI analysis of the cerebral circuit for number comparison. Neuroreport, 14(4), 1005–1014. [DOI] [PubMed] [Google Scholar]
  47. Power JD, Barnes KA, Snyder AZ, Schlaggar BL & Petersen SE (2012) Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. NeuroImage, 59(3), 2142–2154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Renwick JH & Asker RL (1983) Ethanol-sensitive times for the human conceptus. Early Human Development, 8(2), 99–111. [DOI] [PubMed] [Google Scholar]
  49. Rosenberg-Lee M, Barth M. & Menon V. (2011) What difference does a year of schooling make? Maturation of brain response and connectivity between 2nd and 3rd grades during arithmetic problem solving. NeuroImage, 57(3), 796–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Roussotte FF, Sulik KK, Mattson SN, Riley EP, Jones KL, Adnams CM et al. (2012) Regional brain volume reductions relate to facial dysmorphology and neurocognitive function in fetal alcohol spectrum disorders. Human Brain Mapping, 33(4), 920–937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Santhanam P, Li Z, Hu X, Lynch ME & Coles CD (2009) Effects of prenatal alcohol exposure on brain activation during an arithmetic task: an fMRI study. Alcoholism: Clinical and Experimental Research, 33(11), 1901–1908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Simon O, Mangin JF, Cohen L, Le Bihan D. & Dehaene S. (2002) Topographical layout of hand, eye, calculation, and language-related areas in the human parietal lobe. Neuron, 33(3), 475–487. [DOI] [PubMed] [Google Scholar]
  53. Smith SM, Fox PT, Miller KL, Glahn DC, Fox PM, Mackay CE et al. (2009) Correspondence of the brain’s functional architecture during activation and rest. Proceedings of the National Academy of Sciences of the United States of America, 106(31), 13040–13045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H. et al. (2004) Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage, 23, S208–S219. [DOI] [PubMed] [Google Scholar]
  55. Spadoni AD, McGee CL, Fryer SL & Riley EP (2007) Neuroimaging and fetal alcohol spectrum disorders. Neuroscience & Biobehavioral Reviews, 31(2), 239–245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Spottiswoode BS, Meintjes EM, Anderson AW, Molteno CD, Stanton ME, Dodge NC et al. (2011) Diffusion tensor imaging of the cerebellum and eyeblink conditioning in fetal alcohol spectrum disorder. Alcoholism: Clinical and Experimental Research, 35(12), 2174–2183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Stoodley CJ & Schmahmann JD (2009) Functional topography in the human cerebellum: a meta-analysis of neuroimaging studies. NeuroImage, 44(2), 489–501. [DOI] [PubMed] [Google Scholar]
  58. Tisdall MD, Hess AT, Reuter M, Meintjes EM, Fischl B. & van der Kouwe AJ (2012) Volumetric navigators for prospective motion correction and selective reacquisition in neuroanatomical MRI. Magnetic Resonance in Medicine, 68(2), 389–399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. van der Kouwe AJ, Benner T, Salat DH & Fischl B. (2008) Brain morphometry with multiecho MPRAGE. NeuroImage, 40(2), 559–569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Winer BJ (1971) Statistical principles in experimental design, 2nd edition, New York, NY: McGraw-Hill. [Google Scholar]
  61. Woods K, Meintjes E, Molteno C, Jacobson S. & Jacobson J. (2015) Parietal dysfunction during number processing in children with fetal alcohol spectrum disorders. NeuroImage: Clinical, 8, 594–605. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Woods KJ, Jacobson SW, Molteno CD, Jacobson JL & Meintjes EM (2018) Altered parietal activation during non-symbolic number comparison in children with prenatal alcohol exposure. Frontiers in Human Neuroscience, 11, 627. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Yang Y, Roussotte F, Kan E, Sulik KK, Mattson SN, Riley EP et al. (2012) Abnormal cortical thickness alterations in fetal alcohol spectrum disorders and their relationships with facial dysmorphology. Cerebral Cortex, 22(5), 1170–1179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Zago L, Pesenti M, Mellet E, Crivello F, Mazoyer B. & Tzourio-Mazoyer N. (2001) Neural correlates of simple and complex mental calculation. NeuroImage, 13, 314–327. [DOI] [PubMed] [Google Scholar]
  65. Zago L, Petit L, Turbelin MR, Andersson F, Vigneau M. & Tzourio-Mazoyer N. (2008) How verbal and spatial manipulation networks contribute to calculation: an fMRI study. Neuropsychologia, 46, 2403–2414. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Appendices
NIHMS2053142-supplement-Appendices.docx (70.5KB, docx)

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

RESOURCES