Association of Fatty Acid Ethyl Esters in Meconium and Cognitive Development during Childhood and Adolescence

Abstract

Objective

To examine associations between amounts of fatty acid ethyl esters (FAEEs) in meconium and cognitive development in school aged children exposed to alcohol and drugs in utero.

Study design

A secondary analysis of a prospective cohort of children, primarily African-American and of low socioeconomic status, that were recruited at birth. FAEEs were quantified with gas chromatography using a flame ionization detector. Meconium was analyzed for FAEEs in 216 newborns. 191 of these infants were assessed for IQ at ages 9, 11, and 15 years using the Wechsler Intelligence Scales for Children-Fourth Edition (WISC-IV).

Results

Longitudinal mixed model analyses indicated that, controlling for maternal and child covariates, higher concentrations of FAEEs (ethyl myristate, ethyl oleate, ethyl linoleate, ethyl linolenate) were associated with lower WISC-IV Verbal Comprehension Index (VCI), Working Memory Index (WMI), and Full Scale IQ (FSIQ). Associations of FAEEs with VCI, WMI, and FSIQ did not vary over time. No association of FAEEs with Perceptual Reasoning and Processing Speed Indices were found.

Conclusion

Elevated levels of FAEEs in meconium are potential markers for identifying newborns at risk for poor cognitive development related to prenatal alcohol exposure.

Fetal alcohol spectrum disorders (FASDs) represent a wide range of developmental disabilities resulting from alcohol exposure in utero, including fetal alcohol syndrome (FAS) and alcohol-related neurodevelopmental disorder (ARND). FAS, characterized by growth restriction, a distinct pattern of facial features, and evidence of central nervous system dysfunction,¹ is the leading known preventable cause of intellectual disability.² ARND refers to various neurologic abnormalities associated with prenatal alcohol exposure such as problems with memory, learning ability, behavioral problems, and lower intelligence³ without the facial dysmorphology. Even low-to-moderate levels of maternal drinking during pregnancy (< 7 drinks per week), although less consistently, have been related to decreased cognitive ability in African American children,⁴ poor memory,⁵ and behavioral regulation problems.⁶ It is estimated that as many as 2–5% of younger school children in the US and Western Europe are affected by FASDs, with FAS affecting 2–7 per 1000 live births.⁷ A recent meta-analysis⁸ reported the overall prevalence of FAS and other FASDs among children in child welfare systems to be 6% and 17% respectively.

Although early identification of infants at risk for alcohol-related problems is critical to reduce secondary disabilities,⁹ identifying such infants is quite challenging when the distinctive facial features of FAS are not present. Women often under-report drinking due to the stigma associated with drinking during pregnancy. Furthermore, although structured in-depth interviews given by trained professionals in research studies can elicit reliable information,¹⁰ no reliable clinical tools for assessing levels of drinking in pregnant women and identifying newborns who were exposed to alcohol have been established.

Fatty acid ethyl esters (FAEEs), the non-oxidative metabolites of ethanol analyzed in meconium, have been investigated as biomarkers for identifying alcohol exposed neonates.^11,12 Increased concentrations of FAEEs in meconium correlate with fetal exposure to alcohol.^13–15 We previously reported the associations of FAEEs with mental and psychomotor development during the first 2 years of life.¹⁶ To our knowledge, no studies have examined an association between FAEEs in meconium and cognitive outcomes during childhood and adolescence. The purpose of the present study is to extend our previous findings to examine relationships between FAEEs and cognitive development at ages 9, 11, and 15 years. We hypothesize that the relationship of the higher concentration of FAEEs in meconium with poorer cognitive outcomes will persist into older ages.

Methods

This study included 191 children (84 boys, 107 girls) recruited at birth (September 1994 to June 1996) from a large, urban, teaching hospital for a longitudinal study on the neurobehavioral effects of prenatal cocaine exposure.¹⁷ Women at high risk for drug use due to lack of prenatal care, behavior suggesting intoxication, a history of involvement with the Department of Human Services, or self-admitted substance use had drug toxicology screenings at delivery. Women with a psychiatric history (major depression, bipolar disorder, or schizophrenia), low intellectual functioning (diagnosis of mental retardation indicated in medical chart review), HIV-positive status, or chronic medical illness were excluded, as were infants with Down Syndrome, FAS, or congenital heart defects. Random samples of meconium were collected from 248 newborns¹⁴ following informed consent, and 216 had adequate analysis of meconium.¹⁵ Of the 216 children, 14 had missing interview data, 2 children died, and 9 dropped out or were lost to contact, yielding the current sample of 191 (89% retention rate for living children with adequate analysis of meconium). Of the 191 children, 187 (98%) completed the 9 year assessment; 186 (97%) completed the 11 year assessment; and 176 (92%) completed the 15 year assessment. Eighty-nine percent (n=170) had all three IQ assessments, with 98.5% having at least two assessments.

Children and their caregivers were seen by separate examiners at the developmental research laboratory for approximately 5 hours at each follow-up visit at ages 9, 11, and 15 years. Children were assessed by a clinical psychologist or master’s level research assistant; caregivers were assessed by a social worker or trained research assistant. All research assistants were trained and supervised by a licensed clinical psychologist. Examiners were blinded to mother and infant alcohol and drug exposure status. All participants were given a monetary stipend, lunch and transportation costs. This study was approved by the Institutional Review Board of the participating hospital. Parental written informed consent and child assent were obtained. A Certificate of Confidentiality (DA-98-91) was obtained from the Department of Health and Human Services.

Meconium was collected shortly after birth and stored at −70°C until analysis. The FAEEs were extracted with acetone/hexane and isolated using silica gel chromatography. Isolated FAEEs were identified and quantitated by gas chromatography using a flame ionization detector (GC/FID) ^14,15. Six FAEE analytes were examined: ethyl myristate, ethyl palmitate, ethyl oleate, ethyl linoleate, ethyl linolenate, and ethyl arachidonate.¹⁵ Ethyl stearate was not analyzed due to background noise on the chromatograms,¹⁴ and ethyl palmitoleate was excluded as it did not correlate with alcohol exposure in humans¹⁵ or sheep.¹⁸ Meconium analyses were performed by investigators blinded to infant alcohol exposure status.

At the newborn visit, birth mothers were asked to recall frequency and amount of alcohol and drug use for the month prior to and for each trimester of pregnancy. Women were asked the number of drinks consumed per drinking day and what size serving they had. The number of standard drinks (0.5 oz. of absolute alcohol) of beer, wine, or hard liquor per drinking day was computed. Frequency of drinking was recorded on a Likert-type scale ranging from 1 (less than once a month) to 7 (daily use) and converted to reflect the average number of drinking days per week. Number of drinking days per week was multiplied by the number of drinks per drinking day to compute an average number of alcohol drinks per week in the month prior to pregnancy and in each trimester, which were then averaged to obtain a total average drinks per week over the four periods of time. Birth mothers were also asked to recall more than the usual number of drinks (“On the days that you drank more than the usual number of drinks, how many drinks do you have?”). Risk drinking during pregnancy was assessed with TWEAK,¹⁹ with a total score of ≥ 2 indicating pregnancy risk drinking.²⁰ Other substance use during pregnancy, number of tobacco cigarettes, marijuana joints smoked, and crack cocaine “rocks” consumed and the amount of money spent per day, were also collected along with the frequency of use, computing a total average score for each substance (cigarettes, marijuana, and cocaine). The alcohol and drug assessment was updated with the child’s current caregiver at the 9, 11, and 15 year follow-up visits to measure recent (prior 30 day period) caregiver alcohol and drug use.

Demographic and medical characteristics, including maternal age at birth, gestational age, birth weight and length, head circumference, and Hobel Neonatal Risk score, were extracted from the medical records of the mothers and infants. Socioeconomic status was calculated using the Hollingshead Index.²¹ Maternal vocabulary was assessed at birth using the Peabody Picture Vocabulary Test-Revised (PPVT-R),²² and updated using its third edition (PPVT-III)²³ at later assessments. The Block Design and Picture Completion subtests of the Wechsler Adult Intelligence Scale-Revised (WAIS-R)²⁴ were used to estimate maternal non-verbal intelligence at infant birth. Maternal psychological distress was assessed using the Global Severity Index (α = 0.95), a summary scale of the Brief Symptom Inventory,²⁵ at birth and at each follow-up visit. At each visit, the child’s placement (with either biological mother/relative or adoptive/foster caregiver) was noted and data on the current caregiver were updated to provide concurrent assessment of caregiver intelligence and psychological distress.

At 9, 11, and 15 years, children’s intelligence was assessed using the Wechsler Intelligence Scales for Children-Fourth Edition (WISC-IV).²⁶ The WISC-IV yields four summary indices (Verbal Comprehension, Perceptual Reasoning, Processing Speed, and Working Memory) and a Full Scale IQ. The quality of the caregiving environment was assessed via interview at 9, 11 and 15 years using the Home Observation for Measurement of the Environment (HOME) with the middle childhood version used at age 9 and the early adolescent version used at ages 11 and 15.²⁷

Statistical analyses

Data that were positively skewed were normalized using a log transformation (alcohol and drug use, maternal psychological distress) prior to analyses. Means and standard deviations were reported by the variables’ original distribution, with transformations used in analyses. Concentrations for each FAEE were transformed by log₁₀ (FAEE+100) to correct skewed distribution.¹⁶ Spearman correlation (r_s) was conducted to examine the association between specific FAEE and maternal, child, and caregiver characteristics. The relationship of each FAEE with cognitive outcomes were evaluated using a mixed linear model approach with maximum likelihood estimation procedures. Unstructured covariance matrix, estimating each variance and covariance uniquely from the data by making no assumptions regarding the nature of the residual correlations between the repeated measures, was used to account for correlated responses within a subject. We tested the homogeneity of FAEE effects, as well as the effects of sex and other covariates on children’s cognitive outcomes over time by including an interaction term with time. If the interaction was not significant at p < .10, the interaction terms were removed from the model. Missing data were modeled using full-information maximum likelihood, which utilizes all available information from the observed data. Variables that were associated with FAEE analytes or with outcomes at p ≤ .20 for at least one time point were entered into the model using backward elimination approach. Each FAEE was considered separately.

Results

The 191 mothers and children were primarily African American and of low SES (Table I). Only 13% of mothers were married at the child’s birth; 37% (n=70) had not finished high school, with a mean (± SD) number of education years of 11.8 (± 1.5). Sixty percent of the 191 birth mothers reported alcohol use during pregnancy, with 63% engaged in risk drinking (TWEAK ≥ 2) with the mean (± SD) number of alcohol drinks per week of 6.5 (± 11.9). Twenty-seven (14% of the total sample) mothers reported ≥ 7 drinks per week and 15 mothers ≥ 14 drinks per week, indicating that most of the women in this study were low-to-moderate drinkers. More than half of the sample (n= 117, 64%) smoked cigarettes, 46 (24%) used marijuana, and 87 (46%) used crack cocaine during pregnancy. The mean (± SD) gestational age was 38 (± 3) weeks with a mean birth weight of 2994 (± 697) grams. Characteristics of caregivers and the home environment are presented in Table II. Between 8–10% of the children were cared for by non-relative foster or adoptive parents at any given assessment. The mean (± SD) Full Scale IQ (unadjusted) was 87 (± 13) at age 9, 86 (± 13) at 11, and 83 (± 13) at 15.

Table I.

Biological Maternal and Child Characteristics (N = 191)

	n (%)/Mean ± SD	Median (10% – 90%)
Biological maternal
African American	154 (80.6)
Low socioeconomic status	191 (100)
Married	24 (12.6)
Age at birth	27.30 ± 5.28	26 (21 – 35)
Years of education	11.80 ± 1.49	12 (10 – 14)
Number of prenatal visits	7.49 ± 4.94	8 (1 – 14)
Parity	2.83 ± 1.69	3 (1 – 5)
Global Severity Index	0.67 ± 0.69	0.45 (0.08 – 1.51)
PPVT-R Standard Score	76.76 ± 13.40	77 (60 – 93)
WAIS-R Block Design	7.05 ± 2.11	7 (5 – 9)
WAIS-R Picture Completion	6.92 ± 2.32	6 (5 – 10)
Measures of drinking during pregnancy (n=114)a
Number of drinks on drinking day	2.14 ± 2.79	1.5 (0.25 – 4.5)
Number of drinking days per week	1.37 ± 1.49	0.75 (0.06 – 3.5)
Number of drinks per week	6.51 ± 11.95	1.98 (0.06 – 18)
More than usual number of drinks	5.74 ± 8.83	3 (0 – 14)
Risk drinking (TWEAK ≥2)	72 (63)
Other substance use during pregnancya
Cigarettes/day (n = 117)	12.06 ± 10.83	10.5 (1.2 – 20)
Marijuana joints/week (n = 46)	2.91 ± 4.27	1.1 (0.1 – 7)
Cocaine units/week (n = 87)	24.71 ± 51.83	5 (0.4 – 70)
Child
Male	84 (44.0)
Gestational age (weeks)	38.3 ± 2.9	39 (35 – 41)
Prematurity (< 37 wks gestational age)	41 (21.5)
Hobel Neonatal Risk score	5.9 ± 15.3	0 (0 – 15)
Birth weight (grams)	2994 ± 679	3130 (2050 – 3710)
Birth length (cm)	48.8 ± 3.9	49.0 (44.4 – 53.0)
Head circumference (cm)	33.1 ± 2.4	33.0 (30.2 – 35.5)

^aBased on users (n) only

Table II.

Caregiver Characteristics

	9 year (n = 187)	11 year (n = 186)	15 year (n = 176)
Non-kinship foster/adoptive care, n (%)	17 (9.1%)	15 (8.1%)	17 (9.7%)
HOME environmenta	43.50 ± 6.53	46.21 ± 6.33	48.00 ± 6.74
Global Severity Indexa	0.37 ± 0.45	0.42 ± 0.51	0.37 ± 0.46
PPVT Standard Scorea	80.01 ± 14.34	80.05 ± 14.07	79.74 ± 13.53
WAIS-R Block Designa	7.14 ± 2.29	7.18 ± 2.19	7.19 ± 2.06
WAIS-R Picture Completiona	7.16 ± 2.43	7.31 ± 2.46	7.40 ± 2.44
Substance use (past 30 days), n, Median (10% – 90%)b
Alcohol drinks/week	67, 1.5 (0.5–9)	62, 3.3 (0.5–10)	104, 1 (0.3– 6)
Cigarettes/day	85, 10 (2–20)	80, 10 (3–20)	72, 10 (3–20)
Marijuana joints/week	6, 4.3 (0.3–28)	4, 19 (1–35)	8, 1.5 (0.3–70)

^aPresented with Mean ± SD

^bMedian (10% – 90%) is based on users (n) only. No caregivers reported cocaine use in the past 30 days, except one caregiver reporting 1.5 cocaine units per week at 9 year assessment.

Table III gives associations between four FAEE analytes (ethyl myristate, ethyl oleate, ethyl linoleate, ethyl linolenate) and key maternal and infant characteristics. Because preliminary analyses indicated no relationship of two FAEE analytes (ethyl palmitate and ethyl arachidonate) with outcomes, our analysis was focused on these four FAEEs. Maternal report of prenatal alcohol use (yes/no) was associated with ethyl oleate (r_s =.13, p= .09), which correlated weakly with self-reported amount of average alcohol use per week which was averaged over the three trimesters and the month prior to pregnancy (r_s =.11, p= .13). More than the usual number of drinks was marginally associated with ethyl oleate (r_s =.13, p=.086), although a measure of the usual number of drinks on drinking day was not associated with FAEEs. No other drug variables (cigarettes, marijuana, and cocaine) were associated with FAEEs (all p > .20). Maternal psychological distress at birth and parity were related to a higher concentration of ethyl myristate, ethyl oleate, and ethyl linoleate. For infant measures, birth weight and birth length were associated with higher FAEEs.

Table III.

Association (r_s) between FAEE and maternal and infant characteristics

	Ethyl myristate	Ethyl oleate	Ethyl linoleate	Ethyl linolenate
Prenatal drinking (yes/no)	.08	.12^	.10	.04
Number of drinks on drinking day	.08	.10	.08	< .01
Number of drinking days per week	.06	.09	.06	< .01
Number of drinks per week	.08	.11	.09	.02
More than usual number of drinks	.07	.13^	.11	.04
Risk drinking (TWEAK ≥2)	< .01	.05	.05	−.02
Cigarettes per day	.04	.02	.03	−.01
Marijuana joints per week	.10	.06	.08	.05
Cocaine units per week	.08	.05	.03	.04
Maternal Global Severity Index at birth	.14^	.15*	.14^	.09
Parity	.13^	.18*	.14^	.08
Gestational age	.11	.08	.13^	.12
Birth weight	.13^	.12	.18*	.18*
Birth length	.13^	.13^	.20**	.18*

^p < .10;

^*p < .05;

^**p < .01

Relationships of the four FAEE analytes with cognitive outcomes after adjustment for covariates are presented in Table IV. Covariates adjusted for each outcome are listed as footnotes, with significant covariates italicized. In the longitudinal analysis with mixed linear models, none of the interaction terms between the FAEEs and age/visits were significant, indicating that the association of FAEEs with cognitive outcomes did not vary over time. After controlling for covariates, higher concentrations of ethyl oleate, ethyl linoleate, and ethyl linolenate were associated with lower Verbal Comprehension Index. For example, one unit (ng) change in log₁₀ (ethyl oleate + 100) was related to a reduction of 2.05 points in Verbal Comprehension Index on average over time. Higher concentrations of ethyl myristate, ethyl oleate, and ethyl linolenate were associated with lower Working Memory Index. Also, increased concentrations of the four FAEE analytes were associated with lower Full Scale IQ. No association of FAEEs with the Perceptual Reasoning and Processing Speed Indices were found.

Table IV.

Adjusted Association of Fatty Acid Ethyl Esters

	Verbal Comprehensiona		Working Memoryb		Full Scale IQc
	b ± se	p	b ± se	p	b ± se	p
Ethyl myristate	−3.26 ± 1.67	.053	−4.17 ± 1.96	.035	−3.60 ± 1.79	.046
Ethyl oleate	−2.05 ± 0.91	.025	−2.31 ± 1.06	.032	−2.07 ± 0.97	.035
Ethyl linoleate	−1.81 ± 0.83	.031	−1.90 ± 0.98	.054	−1.77 ± 0.88	.047
Ethyl linolenate	−2.31 ± 1.02	.024	−2.63 ± 1.19	.029	−2.62 ± 1.09	.017

Note. b = regression coefficient, se= standard error; Significant (p < .05) covariates are listed in italics.

^aAdjusted for age/visit, birth length, birth mother’s education and psychological distress at birth, current caregiver’s PPVT scores, HOME score, and child race

^bAdjusted for age/visit, birth length, birth mother’s education and psychological distress at birth, current caregiver’s Block Design scores, and child sex

^cAdjusted for age/visit, birth length, birth mother’s education and psychological distress at birth, current caregiver’s Block Design scores, HOME score, and child race

Discussion

We demonstrated that the level of cognitive functioning during childhood and adolescence was inversely associated with increased concentrations of FAEEs in low SES, primarily African-American, urban children who were poly-drug exposed prenatally. Higher levels of the four FAEE analytes were associated with poorer Verbal Comprehension, Working Memory, and Full Scale IQ, supporting our hypotheses. The negative association between higher concentration of FAEEs in meconium and cognitive development noted during the first 2 years of life continued at ages 9, 11, and 15 years. These findings are consistent with our previous studies based on maternal report of alcohol use. Maternal self-report of average alcohol use during pregnancy was related to lower vocabulary scores and working memory at 9 years²⁸ and to lower Verbal Comprehension and Working Memory Indices at 15 years.²⁹ Despite weak correlations between maternal reported alcohol use and FAEEs, their relationships with cognitive development converged. However, neither study utilizing maternal self-report found associations with Full Scale IQ, suggesting higher sensitivity of FAEEs in identifying alcohol exposed children.

Some findings related to FAEEs are noteworthy. Greater birth length and birth weight were correlated with increased concentration of FAEEs. Although heavy drinking during pregnancy is related to restricted growth, effects of low-to-moderate drinking during pregnancy on birth outcomes are not consistent, with some studies reporting greater birth weight and less prematurity among infants with low-to-moderate prenatal alcohol exposure.³⁰ In this study, infants with FAS diagnosed at birth were excluded based on the selection criteria of the parent study.¹⁷ Additionally, higher maternal psychological distress was correlated with increased concentration of FAEEs, suggesting that mothers experiencing greater distress also drank more.

Limitations in our study should be noted. For FAEE determination, the use of the GC/FID is less sensitive and may falsely identify peaks as FAEE when gas chromatography tandem mass spectroscopy (GC/MS/MS) would correctly identify them as FAEE or not. Self-reported alcohol consumption during pregnancy was collected retrospectively, producing possible recall bias. Further, under-reporting or denial due to the stigma associated with pregnancy drinking may be responsible for the weak correlations between various measures of alcohol use during pregnancy and FAEEs observed in this study. Also, birth length and head circumference from medical records may not be reliable. The effects of behavioral, educational, cognitive, or social skills intervention that children may have received were not controlled due to lack of such data.

Despite these limitations, this study has significant strengths. A comprehensive set of covariates and confounders, including other prenatal substance exposures (cigarette, marijuana, cocaine) and caregiving environments, were evaluated and controlled statistically when necessary. Assessment at multiple time points, combined with a high rate of retention for follow-up, allowed repeated observations of an association between FAEEs and cognitive outcomes at different developmental stages, increasing confidence in the findings. Longitudinal analyses allowed us to test whether the relationship of FAEEs on cognitive outcomes changed over time. The present study extends previous studies of FAEEs by demonstrating that poor cognitive development noted in infancy persists into childhood and adolescence. It provides additional evidence that elevated FAEEs in meconium may be an informative biological marker for the early identification of children at risk for cognitive developmental delay, which will serve as a guidance for early intervention. It also establishes predictive validity of FAEEs for determining alcohol exposure in utero.

Acronyms

FAEEs

Fatty acid ethyl esters

FAS

Fetal alcohol syndrome

PPVT

Peabody Picture Vocabulary Test

HOME

Home Observation for Measurement of the Environment