Fetal alcohol spectrum disorders (FASD) comprise the most common preventable cause of neurodevelopmental disabilities worldwide, with prevalence estimates of 1.1% to 5.0% in the United States and Western Europe1 and 13.6% to 20.9% in South Africa.2 Although children with FASD often phenotypically resemble children with other disorders, the neural underpinnings of FASD neurobehavioral deficits are unique and, therefore, require specifically targeted interventions. Accordingly, a growing number of evidence-based programs tailored to FASD are being studied, targeting arithmetic, maladaptive social behaviors, attention and self-regulation, language development, and literacy.3,4 However, the promise of these interventions is limited by diagnostic challenges in clinical practice because children with FASD are most difficult to identify in infancy and early childhood, when neurocognitive interventions may be most effective. In “Early-Life Predictors of Fetal Alcohol Spectrum Disorders” in this issue of Pediatrics, Kalberg et al5 address the issue of early diagnosis in a prospective longitudinal cohort in the Western Cape province of South Africa by examining the degree to which growth, dysmorphology, and neurobehavioral measures are predictive of FASD diagnosis at age 5 years.
Although the authors warn that dysmorphology examinations may not be reliable until age 9 months in clinical practice, it is impressive that the total score from a structured dysmorphology examination by expert dysmorphologists, which included weight and head circumference, differentiated between children with and without an FASD diagnosis as early as 6 weeks of age. In receiver operating characteristic curve analyses, total dysmorphology score sensitivity and specificity for FASD detection appeared to improve as the infants aged, with an area under the curve of 0.839 at age 18 months. The score also began to differentiate between FASD diagnostic groups as the infants aged. These findings indicate that a detailed dysmorphology examination conducted by an expert geneticist may identify children who are in need of interventions aimed at FASD neurobehavioral deficits. However, the cutoff used for the receiver operating characteristic curve analyses was not given, so these findings cannot currently be used for screening in clinical practice. The predictive validity of growth restriction presented is consistent with our previous findings that children who exhibit both pre- and postnatal growth restriction (length or height <10th percentile) have the most severe FASD neurobehavioral deficits, with long-term postnatal growth restriction being reliably present by age 12 months in >85% of the children.6
The Brazelton Neonatal Behavioral Assessment Scale and Bayley Scales of Infant Development did not reliably predict later FASD diagnoses at any age. The failure of these assessments to discriminate between groups is likely due to the nature of the tests, which provide global measures of overall current behavioral development but have poor predictive validity. By contrast, narrow-band measures targeting neural underpinnings of FASD neurobehavioral deficits that are unique to the teratogenic effects of alcohol (eg, recognition memory) have demonstrated predictive validity in later-childhood cognitive function.7–9 Of note, the only assessment of childhood cognition in the Kalberg et al5 study that was found to discriminate between the diagnostic groups, the Kaufman Assessment Battery for Children, was not assessed until the time of diagnosis (age 5 years).
The prospective longitudinal design of this study is a significant strength because serial measurements of the same children enabled the authors to examine the potential for growth, dysmorphology, and neurobehavioral assessments to predict subsequent FASD diagnoses from different time points. Of note, given the inclusion of both exposed and unexposed children from the same community, the findings presented are likely specific to prenatal alcohol exposure, rather than a reflection of socioenvironmental confounders, and are thus likely generalizable to other populations. Similarly designed studies in the Cape Town region of South Africa have yielded results that have been replicated in other human populations and animal models.10–12
The results of the Kalberg et al5 study underscore the critical need for biomarkers of FASD neurobehavioral deficits that can be readily obtained early in life. Although the discriminative validity of the dysmorphology score assessment is promising, it requires detailed examinations by expert dysmorphologists, of whom there are few worldwide and are even fewer in resource-poor areas where FASD is endemic. Narrow-band neurobehavioral assessments that perform better than the more global infant behavioral measures used in this study also require specialized expertise to administer. New approaches using three-dimensional photographs to detect facial dysmorphism may prove useful in identifying children of all ages who are at risk for FASD.13 Advances in large-scale –omics (eg, genome-wide methylation or transcriptome) data analysis methods and an ever-growing body of scientific literature elucidating the underlying mechanisms in the teratogenic effects of alcohol may provide opportunities to identify molecular biomarkers of FASD deficits. Epigenetic biomarkers are particularly promising because they may be less specific to time and tissue than direct markers of neuronal damage are. We recently demonstrated alcohol-related disruptions in placental imprinted gene expression that statistically mediated alcohol-related growth restriction later in infancy.14 In addition, as noted above, postnatal growth restriction remains an easily measurable biomarker of FASD deficits that may identify children who warrant further, more specialized assessments.6
Opinions expressed in these commentaries are those of the authors and not necessarily those of the American Academy of Pediatrics or its Committees.
FUNDING: No external funding.
COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2018-2141.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.