Claire B. Ernhart

Alcohol teratogenicity in the human: A detailed assessment of specificity, critical period, and threshold

Though the occurrence of alcohol-related birth defects is well documented in the human and in animal models, definition of specificity, critical period, and dose-response threshold with a precision adequate for clinical risk assessment and management has been lacking. Data from a cohort of 359 neonates, from a large prospective observational study in which chronic alcohol problems and maternal drinking were assessed during pregnancy and standardized neonatal examinations were blinded for prenatal information, were analyzed with the use of multivariate techniques, with uniform control for confounding by eight factors. Craniofacial abnormalities were found to be definitively related to prenatal alcohol exposure in a dose-response manner (p less than 0.001); a significant, but less striking, relationship was observed for other anomalies (p less than 0.01). The critical period for alcohol teratogenicity was confirmed to be around the time of conception. Risk for anatomic abnormalities in the offspring was clearly defined among the 5.6% of infants whose mothers drank more than three ounces of absolute alcohol, that is, more than six drinks, per day. Because of a trend toward an increase in craniofacial abnormalities with increasing embryonic alcohol exposure at lower levels, a clear threshold could not be defined. These are not experimental data and the results should not be overinterpreted. Nevertheless, pending further studies with larger samples, the findings suggest that to completely avoid alcohol-related anatomic abnormalities, advice to discontinue drinking or at least to reduce it to a minimal level before conception is clinically appropriate.

Breastfeeding and cognitive development in the first 2 years of life

The relationship between breastfeeding and cognitive development in the first 2 years of life was examined in a cohort of children being followed in a study of risk factors in development. A significant difference between bottlefed children, children breastfed less than or equal to 4 months and those breastfed greater than 4 months was found on the Mental Development Index of the Bayley Scales at ages 1 and 2 years, favoring the breastfed children. At age 6 months, the direction of the relationship was the same but did not reach significance. Supplementary regression analyses examining the strength of the relationship between duration of breastfeeding and cognitive development similarly showed a small but significant relationship between duration of breastfeeding and scores on the Bayley at 1 and 2 years. Alternative explanations for the results are discussed.

Low level lead exposure in the prenatal and early preschool periods: Intelligence prior to school entry☆

The hypothesis that low level lead exposure in the fetal and early preschool years is related to neuropsychological deficit is being examined in a prospective study of development through the preschool years. Clear and consistent support for the hypothesis was not obtained in previously reported findings relating maternal and cord blood lead level (PbB) and PbB from venous samples through age three years to a series of developmental tests. It was concluded the relationship of lead level and cognitive development was primarily a function of the dependence of each on the quality of the caretaking environment. The cohort has now been examined at age four years, ten months, an age selected to assess development just prior to school age. The test used was the Wechsler Preschool and Primary Scale of Intelligence. We found statistically significant correlations between most of the PbB measures and the intelligence test scores, but these were attenuated, not statistically significant and not consistent in direction when relevant confounding variables were considered. Possible biases or threats to validity were reviewed. Effect sizes relating both prenatal and preschool lead exposure to intellectual development were small, not statistically significant, and not consistent in direction.

Significant Determinants of Susceptibility to Alcohol Teratogenicity

Typically, the rate of abusive drinking during pregnancy considerably exceeds the rates of fetal alcohol syndrome (FAS) and alcohol-related birth defects, suggesting that other factors may modify the impact of alcohol on the developing organism. Data in the literature supporting this susceptibility hypothesis are sparse. In this paper, two studies in different samples, using different analytic strategies to examine susceptibility to different adverse outcomes are presented. Among 176 pregnancies in which lowered birth weight for gestational age was detected as an effect attributable to frequent beer drinking, 27 infants weighted less than 2,700 grams and 149 weighed more. Using discriminant analysis to contrast these groups, lowered birth weight for gestational age was associated with black race and lower maternal weight and weight gain. The effects of these factors were additive with that of persistent alcohol exposure; no interactions were detected, but pregnancies with risks in addition to alcohol were more likely to yield growth-retarded infants. In a second study, pregnancies resulting in 25 FAS cases were contrasted with 50 controls. A four-factor model accounted for nearly two-thirds of the explainable variance in the occurrence of FAS. Adjusted for frequency of maternal drinking, chronic alcohol problems and parity, there was a sevenfold increase in risk for FAS among black infants. The findings from both studies are consistent with the susceptibility hypothesis and have potentially important implications for public health and clinical approaches to prevention, as well as for future research.

Congenital and postnatally acquired cytomegalovirus infections: Long-term follow-up

To determine long-term outcome of children with inapparent congenital cytomegalovirus infection, an assessment of congenitally infected children observed since birth was undertaken. Children with early postnatal acquisition of CMV infection were also evaluated. Cognitive, behavioral, neurologic, audiometric, and speech and language evaluations were performed in 48 patients, including 17 congenitally infected children, 10 children with postnatal infection, and 21 uninfected control subjects. Mean IQ of the three groups of children did not differ significantly. Behavioral, neurologic, speech and language examinations similarly failed to distinguish differences among the three groups. Audiologic abnormalities were present in four congenitally infected children, including one child with a severe unilateral sensorineural loss; in none of the children was hearing loss functionally significant. No hearing abnormalities were detected in postnatally infected children. Although inapparent CMV infection can result in audiologic sequelae, the continued lack of cognitive, behavioral, and neurologic sequelae in these school-age children reemphasizes the need to focus attention on prevention of primary maternal CMV infection to avoid the potentially devastating effects of intrauterine CMV infection.

Intrauterine Exposure to Low Levels of Lead: The Status of the Neonate

It has been suggested that lead (Pb) at low exposure levels is a behavioral teratogen. Blood lead (Pb-B) was measured in 185 samples of maternal blood and in 162 samples of cord blood drawn from members of a cohort of mother-infant pairs. Routine newborn assessments, an examination for minor anomalies, the Brazelton Neonatal Behavioral Assessment (NBAS), and part of the Graham/Rosenblith Behavioral Examination (G/R) were administered. Maternal and cord Pb-B correlated 0.80. In regression analyses, Apgar scores, birthweight, length, head circumference, neonatal anomalies, and seven behavioral scales were unrelated to either maternal or cord Pb-B. Three scales--the NBAS Abnormal Reflexes, the G/R Neurological Soft Sign, and the G/R Muscle Tonus Scales--were related minimally to either cord or maternal Pb-B. Because of the contrast in maternal and cord results, despite the high correlation of maternal and cord Pb-B, the data were reanalyzed for 132 cases with paired data. Only the Soft Sign Scale remained significant and that only for cord, but not maternal Pb-B. Regression analysis revealed a suppression with the Soft Sign Scale related to the variance of the cord Pb-B that was not common with maternal Pb-B. The possibility that the fetus under stress tends to accumulate Pb was considered.

Low level lead exposure in the prenatal and early preschool periods: Early preschool development☆

The hypothesis that low level lead exposure in the fetal and early preschool years is related to neuropsychological deficit was examined in a prospective study of child development. We also tested the hypothesis of reverse causality, i.e., that lead level is a function of prior developmental status. Fetal lead exposure was measured in maternal and cord blood while preschool lead level was measured in venous blood samples at ages six months, two years and three years. These blood lead measures (PbB) were related to concurrent and ensuing scores on developmental measures at six months, one year, two years, and three years. With statistical control of covariate measures (age, sex, race, birth weight, birth order, gestational exposure to other toxic substances, maternal intelligence, and several indicators of the quality of the caretaking environment) as well as potentially confounding risk factors (gestational exposure to alcohol and other toxic substances), most statistically significant associations of PbB with concurrent and later development were completely attenuated. Effects of lead exposure, significant or not, were not consistent in direction. In reverse-causality analyses, PbB was not related significantly to prior measures of developmental retardation or acceleration. It was concluded that the relationship of lead level and measures of development in these early years was primarily a function of the dependence of each on the quality of the caretaking environment.

PRENATAL ALCOHOL EXPOSURE AND SUSTAINED ATTENTION IN THE PRESCHOOL YEARS

The effects of fetal alcohol exposure on sustained attention in the preschool years were investigated in a sample of socioeconomically disadvantaged urban children. The 245 children examined were the 4-year 10-month-old follow-up cohort of a prospective study begun while their mothers were pregnant. Sustained attention was assessed by an automated, computer-administered vigilance task. Indices of fetal alcohol exposure included three measures of maternal drinking and two early indicators of fetal alcohol damage. The data were analyzed using univariate and multivariate regression techniques, both with and without adjustment for covariates. None of the analyses supported the hypothesis that fetal alcohol exposure adversely affects sustained attention performance in preschool children. We infer that, if there is an effect of prenatal alcohol exposure on vigilance task performance, that effect is small.

Prenatal alcohol exposure and cognitive development in the preschool years

The teratogenic effects of drinking during pregnancy on cognitive development were investigated in a cohort of disadvantaged mother-infant pairs. Three indices associated with maternal alcohol use were related to comprehensive cognitive measures obtained during five in-home assessments through age four years, ten months. Birth weight and a tally of craniofacial anomalies were also evaluated as early indicators of fetal alcohol damage. Multiple analyses relating the alcohol and cognitive measures provided no evidence of an adverse alcohol effect on cognitive development in the absence of Fetal Alcohol Syndrome. Assessments based on confidence intervals indicated that an average decrement of more than two points (IQ equivalent) on the five cognitive measures is unlikely for relatively heavy levels of maternal alcohol use or alcohol-related problems. Marginal relationships were observed between cognitive measures and the neonatal indicators, but depended heavily on the influence of a child exhibiting signs consistent with Fetal Alcohol Syndrome. This case therefore received special attention. The implications of the relationships involving the neonatal indicators were investigated using arguments based on path analysis.

Fetal lead exposure: Antenatal factors

It was hypothesized that maternal blood lead level at delivery and cord blood lead level of the neonate would be affected by maternal use of alcohol, history of alcohol abuse, and smoking. The possibility that iron status, as reflected in maternal serum ferritin, would be related to lead level was also explored. The maternal history of alcohol abuse was unrelated to lead level in 208 samples of maternal blood and 178 samples of cord blood. However, alcohol use during pregnancy was related in a dose-response fashion to maternal and to cord blood lead level. This effect was significant with and without control of maternal smoking. The effect of maternal smoking and serum thiocyanate on maternal and cord blood lead level were also highly significant with and without control of the maternal drinking variable. Serum ferritin was marginally related to lead level for white women and for black infants, but tests of the dichotomized maternal ferritin variable did not yield a significant linkage with maternal or cord blood lead level. The results further support recommendations that women abstain from alcohol consumption and cigarette smoking in pregnancy.