Video Abstract

Video Abstract

BACKGROUND AND OBJECTIVES:

Fetal alcohol spectrum disorders (FASD) comprise the continuum of disabilities associated with prenatal alcohol exposure. Although infancy remains the most effective time for initiation of intervention services, current diagnostic schemes demonstrate the greatest confidence, accuracy, and reliability in school-aged children. Our aims for the current study were to identify growth, dysmorphology, and neurodevelopmental features in infants that were most predictive of FASD at age 5, thereby improving the timeliness of diagnoses.

METHODS:

A cohort of pregnant South African women attending primary health care clinics or giving birth in provincial hospitals was enrolled in the project. Children were followed longitudinally from birth to 60 months to determine their physical and developmental trajectories (N = 155). Standardized protocols were used to assess growth, dysmorphology, and development at 6 weeks and at 9, 18, 42, and 60 months. A structured maternal interview, including estimation of prenatal alcohol intake, was administered at 42 or 60 months.

RESULTS:

Growth restriction and total dysmorphology scores differentiated among children with and without FASD as early as 9 months (area under the receiver operating characteristic curve = 0.777; P < .001; 95% confidence interval: 0.705–0.849), although children who were severely affected could be identified earlier. Assessment of developmental milestones revealed significant developmental differences emerging among children with and without FASD between 18 and 42 months. Mothers of children with FASD were significantly smaller, with lower BMIs and higher alcohol intake during pregnancy, than mothers of children without FASD.

CONCLUSIONS:

Assessment of a combination of growth, dysmorphology, and neurobehavioral characteristics allows for accurate identification of most children with FASD as early as 9 to 18 months.

What’s Known on This Subject:

Drinking during pregnancy results in a continuum of disabilities in exposed offspring, fetal alcohol spectrum disorders. Although early infant intervention positively influences long-term developmental outcome of at-risk children, current diagnostic criteria reveal highest accuracy and reliability in school-aged children.

What This Study Adds:

Growth, dysmorphic features, and neurobehavioral characteristics in infancy can predict which children are at greatest risk of being assigned fetal alcohol spectrum disorder diagnoses at age 5, thereby aiding in timely diagnosis and initiation of intervention services in early life.

Fetal alcohol spectrum disorders (FASD) encompass a range of deleterious effects of maternal alcohol consumption during pregnancy. Together, they comprise the most common environmentally induced category of intellectual disability in the world, potentially affecting 1% to 6.5% of school-aged children (10–65 per 1000) in the United States.14  Although the most severe end of the FASD continuum (fetal alcohol syndrome [FAS]) occurs in 0.2% to 0.9% of live births (2–9 per 1000) in the United States,3  it is much more common elsewhere.5  For example, in some communities in the Western Cape Province of South Africa, 9.3% to 12.8% of children (93–128 per 1000) have documented FAS, whereas the full gamut of FASD affects 18.2% to 25.9% of children (182–259 per 1000).6,7 

Although several diagnostic schemes are used to assign diagnoses in the FASD continuum, the parameters set forth by the Institute of Medicine have been employed most extensively in international population-based studies.8,9  These diagnostic guidelines recently were updated as the Collaboration on Fetal Alcohol Spectrum Disorder Prevalence (CoFASP) Consensus Clinical Diagnostic Guidelines for FASD10  (Table 1).

TABLE 1

Updated Diagnostic Guidelines for FASD

CoFASP Consensus Clinical Diagnostic Guidelines for FASD
Prenatal Alcohol ExposureFAS Facial FeaturesaGrowth RestrictionbDeficient Brain GrowthcNeurobehavioral Impairment (<3 y)dNeurobehavioral Impairment (>3 y)eStructural Birth Defectsf
FAS       
 With confirmed or unconfirmed alcohol exposure — 
PFAS       
 With confirmed alcohol exposure — — — 
 With unconfirmed alcohol exposure X (or deficient brain growth) X (or deficient height and/or weight) — 
ARNDg       
 With confirmed alcohol exposure — — — N/A — 
 Alcohol-related birth defects       
 With confirmed alcohol exposure — — — — — 
CoFASP Consensus Clinical Diagnostic Guidelines for FASD
Prenatal Alcohol ExposureFAS Facial FeaturesaGrowth RestrictionbDeficient Brain GrowthcNeurobehavioral Impairment (<3 y)dNeurobehavioral Impairment (>3 y)eStructural Birth Defectsf
FAS       
 With confirmed or unconfirmed alcohol exposure — 
PFAS       
 With confirmed alcohol exposure — — — 
 With unconfirmed alcohol exposure X (or deficient brain growth) X (or deficient height and/or weight) — 
ARNDg       
 With confirmed alcohol exposure — — — N/A — 
 Alcohol-related birth defects       
 With confirmed alcohol exposure — — — — — 

Adapted from Hoyme HE, Kalberg WO, Elliott AJ, et al. Updated clinical guidelines for diagnosing fetal alcohol spectrum disorders. Pediatrics. 2016;138(2):e20154256. N/A, not applicable; X, required; —, not required.

a

The characteristic pattern of facial anomalies is defined by the presence of ≥2 of the following: (1) short palpebral fissures (≤10th percentile), (2) thin vermilion border (rank of 4 or 5 on racially normed lip-philtrum guide), and (3) smooth philtrum (rank of 4 or 5 on racially normed lip-philtrum guide).

b

Prenatal and/or postnatal growth deficiency: height and/or weight ≤10th percentile on sex-specific population-normed growth curves.

c

Deficient brain growth, morphogenesis, and/or neurophysiology is characterized by ≥1 of the following: (1) OFC ≤10th percentile, (2) structural brain abnormalities, and (3) recurrent nonfebrile seizures.

d

Affected children must display evidence of developmental delay ≥1.5 SD below the mean.

e

Children should be assessed for global, cognitive, and behavioral deficits. Global impairment: general conceptual ability, performance IQ, visual IQ, or spatial IQ ≥1.5 SD below the mean; cognitive deficit: ≥1.5 SD below the mean in 1 domain (executive function, specific learning impairment, memory impairment, or visual spatial impairment); behavioral impairment without cognitive impairment: behavioral deficit in 1 domain ≥1.5 SD below the mean in areas of self-regulation (mood or behavioral regulation impairment, attention deficit, or impulse control).

f

One or more major malformations demonstrated in animal models and human studies to be related to prenatal alcohol exposure, including cardiac defects (eg, atrial septal defects, ventricular septal defects, aberrant great vessels, conotruncal heart defects), musculoskeletal defects (eg, radioulnar synostosis, vertebral segmentation defects, large joint contractures, scoliosis), renal anomalies (eg, aplastic, hypoplastic, or dysplastic kidneys; “horseshoe” kidneys; ureteral duplications), eye anomalies (eg, strabismus, ptosis, retinal vascular anomalies, optic nerve hypoplasia), and/or hearing impairment (eg, conductive or neurosensory hearing loss).

g

For ARND, 2 domains of impairment are required for either cognitive deficit without behavioral impairment or behavioral impairment without cognitive deficit.

Many consequences of FASD are lifelong, and behavioral and learning difficulties constitute a significant burden. Neuroscience has revealed that the neural plasticity of young brains is positively enhanced by intervention through early infant stimulation and augmented nutrition.11,12  Although identification and referral of at-risk children within the first few months of life may be crucial for initiating effective early intervention services, diagnosis of the continuum of FASD in infants is rarely attempted. Such diagnostic inattention may be explained by a lack of phenotypic specificity in the newborn13  and by the higher confidence, accuracy, and reliability of diagnoses in school-aged children.9,10,1417 

In the current study, we seek to identify discriminating features of FASD in early life, which will lead to improved timeliness of diagnosis. Our aims for the study include comparing and contrasting young children with FASD with typically developing children on (1) growth patterns, (2) dysmorphic features, and (3) measures of temperament and development, thereby identifying features in infants that can most accurately predict a diagnosis of FASD at age 5.

An international multidisciplinary team of experienced investigators led the study, which was conducted over an 8-year period (2008–2015). Dysmorphology and developmental assessments were completed on each study child at 6 weeks (time point 1 [T1]) and at 9 (time point 2 [T2]), 18 (time point 3 [T3]), 42 (time point 4 [T4]) and 60 months of age (time point 5 [T5]).

Two regional communities in the Western Cape Province of South Africa with a high prevalence of documented FASD (comprising 5 towns and their surrounding rural areas) housed the project.6,7,1821  The province’s population encompasses a diverse racial makeup (mixed race, 50%; black, 33%; white, 16%; and Indian or Asian, 1%).22  The proportion of women who drink heavily during pregnancy in the 2 rural agricultural study communities is 23.7%.23  By 20 weeks’ gestation, 52.5% of pregnant women receive prenatal care, and most women schedule their first antenatal visit in the second trimester.24 

The Faculty of Medicine and Health Sciences Research Ethics Committees at Stellenbosch University and the University of New Mexico approved all study procedures and data collection tools. As an incentive and reimbursement for their time, participating women and children received a grocery store voucher at each clinic visit.

Trained research staff recruited pregnant women and their index study children from primary health care clinics and hospitals, explained the study to prospective participants, and, after obtaining consent, conducted screening interviews that included the 10-item Alcohol Use Disorders Test (AUDIT).25,26  Women with a history of drinking within the last year were advised about the dangers of drinking during pregnancy and counseled to stop drinking. The screened women reported the complete gamut of drinking behavior, ranging from abstinence to heavy drinking. The modal pattern of drinking occurred in binges of ≥3 standard drinks per day (14 g of absolute alcohol) on weekends. Initially, any woman who visited a primary health care clinic for antenatal care and who agreed to participate was recruited. After delivery, the children of enrolled women were evaluated at T1, T2, and T3. Once several hundred children were recruited and had been followed through T3, a subset for the longitudinal study was selected. The criteria for inclusion in the longitudinal cohort included completion of testing at all 3 time periods and maternal AUDIT scores >8 or <7. Of the 199 children initially enrolled in the study, mothers of 105 reported AUDIT scores >8 (52.8%), and mothers of 94 (47.2%) reported AUDIT scores <7. Of the women whose AUDIT scores were >8, 44 (42%) scored between 8 and 14, and 61 (58%) scored >15. Of the women whose AUDIT scores were <7, 52 (55%) scored 0, and 42 (45%) scored between 1 and 7. Therefore, the cohort reflected a representative range of maternal drinking by design.

One mother-child dyad dropped out immediately (before T1), and an additional 44 were later excluded from analysis because of insufficient data. Of the 44 children who were excluded from the study, 23 were discussed at a case conference after T5 and were found to have insufficient data for inclusion: 18 because of lack of developmental testing and 5 because both testing and prenatal alcohol exposure data were missing. An additional 21 children were not available and were therefore not examined by the dysmorphologists at T5, 9 of whom were last examined at T4 and 12 of whom were last evaluated at T3. There were no statistically significant differences (on t tests, P < .05) between the 44 children with insufficient data and those with complete data sets when compared on height, weight, occipitofrontal (head) circumference (OFC), total dysmorphology score, or Bayley Scales of Infant Development, Third Edition (BSID-III) cognitive percentile at T3, T4, or T5. A total of 155 children completed the entire protocol with a complete data set and are included in this analysis. Because of the complexity of the project, the number of children seen at each time period was variable (Fig 1).

FIGURE 1

Consolidated Standards of Reporting Trials chart for the longitudinal study. aAfter 44 subjects were excluded because of incomplete data, 155 children completed the entire battery.

FIGURE 1

Consolidated Standards of Reporting Trials chart for the longitudinal study. aAfter 44 subjects were excluded because of incomplete data, 155 children completed the entire battery.

The children were assessed on 3 domains: (1) growth (height, weight, and OFC), (2) dysmorphology, and (3) development. A structured maternal interview was also performed on all participants. After completion of the study, data were analyzed to determine which morphometric measures, dysmorphic features, developmental skills, or combination thereof was most predictive of a definitive FASD diagnosis at age 5.

Structured maternal interviews were used to gather information about maternal age, ethnicity, physical traits (height, weight, OFC, BMI), gravidity, parity, time of pregnancy recognition, and drinking before and during pregnancy by trimester.

Expert US clinical geneticists and dysmorphologists assigned diagnoses using the CoFASP diagnostic guidelines for FASD.10  The guidelines set forth a structured, empirically based dysmorphology assessment scheme and weighted numerical scoring system for quantifying growth restriction and minor anomalies common among children exposed to alcohol. Dysmorphology examinations adhered to a standardized protocol at each age interval (T1–T5). US or South African dysmorphologists, pediatricians, or specially trained research staff conducted the examinations. All staff were trained in the CoFASP diagnostic methodology10  by the expert US dysmorphologists. Examiners were blinded to each child’s prenatal alcohol exposure history and to the results of any previous study assessments at each time interval.

Specially trained research staff administered the Brazelton Neonatal Behavioral Assessment Scale (BNBAS-3). Assessment of each infant’s ability to regulate his or her state was a focus of this study.27  Self-regulation (including general irritability, ability to maintain attention, and general state maintenance) was hypothesized to be more difficult for children who were prenatally exposed to alcohol. Specific items on the BNBAS-3 explored aspects of the child’s state, affect, and temperament.28 

Sanctioned trainers instructed staff members on administration of the BSID-III. The BSID-III,29  designed to track attainment of developmental milestones and abilities, was used to assess infant and toddler development from 1 to 42 months of age across 5 domains (cognitive, language, motor, social-emotional, and adaptive abilities). Cognitive, language, and motor skills were assessed directly with the child, whereas social-emotional skills were derived from responses of the primary caregiver to a questionnaire. Composite standard scores were derived for cognitive, language, motor, and social-emotional skills.29 

The Kaufman Assessment Battery for Children, Second Edition (KABC-II)30  was used to measure processing and cognitive abilities of children at T5. This tool provides a mental processing index (MPI) that de-emphasizes acquired knowledge, instead focusing on sequential and simultaneous processing, learning ability, and planning ability. The KABC-II was chosen because it has been validated for use in previous South African and other African research initiatives.31,32  Two Afrikaans-speaking clinical psychologists who were blinded to alcohol exposure history administered the assessment.

During the initial clinical assessment at T1, a dysmorphology examination, BNBAS-3 assessment, and BSID-III testing were completed. The BSID-III was administered at assessment points T1 to T4, and the KABC-II was completed at T5. US dysmorphologists provided 2 physical examinations, 1 at T4 and 1 at T5. Children who were born before 36 weeks’ gestation were considered pre-term, and all developmental testing was scored on the basis of the child’s adjusted age for prematurity until 24 months. Interexaminer reliability data are not available for the present investigation but have been sound in previous studies that were conducted by using this methodology.18,19 

All dysmorphology data were systematically collected and recorded on a standard dysmorphology checklist,9,10  and developmental testing followed standardized procedures. After the data were collected, they were entered into Excel databases, quality assured, and exported into a master database in SPSS (version 22; IBM SPSS Statistics, IBM Corporation, Armonk, NY]).33 

Data from the 155 children diagnosed produced a data set for multiple–repeated-measures analysis of variance (ANOVA) to compare changes in scores over the 5 time points on the basis of the final diagnoses assigned. Categorical diagnoses assigned at case conferences included FAS, partial fetal alcohol syndrome (PFAS), alcohol-related neurodevelopmental disorder (ARND), or not FASD. No children met diagnostic criteria for alcohol-related birth defects.9,10  No other genetic or malformation syndromes were identified. Final diagnoses were assigned by the US dysmorphologists in a multidisciplinary case conference on the basis of the data gathered at T5 (psychological and behavioral assessments, growth, and dysmorphology examinations) and on the basis of alcohol exposure information obtained from the questionnaire responses provided at both the initial screening and from the full maternal interview completed at either T4 or T5. At the time of assignment of the final diagnosis, diagnosticians were blinded to any other data previously gathered or to preliminary diagnoses assigned at earlier time points.

SPSS was used for all data analyses.33  A repeated-measure ANOVA was used to measure differences between related population means over time because the primary longitudinal study aim was to identify factors early in a child’s life that would predict a later FASD diagnosis. In addition, a receiver operating characteristic (ROC) analysis was employed to strictly define the time period when diagnostic prediction, by using the dysmorphology score, achieved significance. The null hypothesis was that the means of FASD groups and the children without FASD would be equal over time.

In Table 2, the following are compared between mothers who gave birth to children with FASD and those who did not: (1) age, (2) growth parameters, (3) educational level, (4) race, (5) gravidity, (6) parity, (7) and self-reported drinking behavior before and during pregnancy.34,35  Mothers of children diagnosed with FASD at age 5 demonstrated significantly higher gravidity (3.5 vs 2.6) than did mothers of children without FASD. Self-reported drinking before and during pregnancy revealed significant differences in drinking during pregnancy (especially in the second and third trimesters) between mothers who had children with FASD and those who did not. By FASD diagnosis, the percentages of women who reported drinking during pregnancy were 90% for FAS, 100% for PFAS, 100% for ARND, and 51.4% for non-FASD. A comparison of other traits between mothers who drank during pregnancy and those who did not revealed significant differences in height (P < .001), weight (P = .006), OFC (P = .010), BMI (P = .024), number of years of education (P < .001), gravidity (P = .002), and parity (P = .003). The mothers of children exposed to alcohol were significantly shorter and lighter, had smaller OFCs and BMIs, and demonstrated higher gravidity and parity.

TABLE 2

Maternal Characteristics

Mothers of Children With FASD (n = 79)Mothers of Children Without FASD (n = 76)P
Age at interview, y, mean (SD) 32.5 (7.9) 29.3 (7.1) .01 
Height, cm, mean (SD) 154.5 (5.7) 158.7 (7.6) <.001 
Weight, kg, mean (SD) 55.4 (15.8) 70.2 (19.2) <.001 
OFC, cm, mean (SD) 54.2 (2.4) 55.2 (2.3) .010 
BMI, mean (SD) 22.9 (5.8) 27.7 (7.6) <.001 
Number of years of school, mean (SD)a 8.2 (2.1) 9.8 (2.2) .001 
Race, %    
 Mixed race 97.5 90.8 — 
 Black 2.5 9.2 — 
 White 0.0 0.0 .07 
Gravidity, mean (SD) 3.5 (1.6) 2.6 (1.6) .002 
Parity, mean (SD) 3.2 (1.7) 2.4 (1.3) .003 
Spontaneous abortions, mean (SD) 0.3 (0.6) 0.2 (0.5) .25 
Induced abortions, mean (SD) 0.0 (0.1) 0.1 (0.2) .17 
Stillbirths, mean (SD) 0.1 (0.2) 0.0 (0.3) .55 
Week of pregnancy recognition, mean (SD) 12.6 (5.4) 12.8 (6.2) .81 
Drinking during pregnancy, % yes 94.9 51.4 <.001 
Usual number of drinks per drinking day, mean (SD)    
 First trimesterb 8.3 (9.3) 5.9 (4.9) .14 
 Second trimesterb 7.4 (10.0) 3.8 (4.4) .04 
 Third trimesterb 5.3 (10.4) 1.8 (3.9) .02 
Mothers of Children With FASD (n = 79)Mothers of Children Without FASD (n = 76)P
Age at interview, y, mean (SD) 32.5 (7.9) 29.3 (7.1) .01 
Height, cm, mean (SD) 154.5 (5.7) 158.7 (7.6) <.001 
Weight, kg, mean (SD) 55.4 (15.8) 70.2 (19.2) <.001 
OFC, cm, mean (SD) 54.2 (2.4) 55.2 (2.3) .010 
BMI, mean (SD) 22.9 (5.8) 27.7 (7.6) <.001 
Number of years of school, mean (SD)a 8.2 (2.1) 9.8 (2.2) .001 
Race, %    
 Mixed race 97.5 90.8 — 
 Black 2.5 9.2 — 
 White 0.0 0.0 .07 
Gravidity, mean (SD) 3.5 (1.6) 2.6 (1.6) .002 
Parity, mean (SD) 3.2 (1.7) 2.4 (1.3) .003 
Spontaneous abortions, mean (SD) 0.3 (0.6) 0.2 (0.5) .25 
Induced abortions, mean (SD) 0.0 (0.1) 0.1 (0.2) .17 
Stillbirths, mean (SD) 0.1 (0.2) 0.0 (0.3) .55 
Week of pregnancy recognition, mean (SD) 12.6 (5.4) 12.8 (6.2) .81 
Drinking during pregnancy, % yes 94.9 51.4 <.001 
Usual number of drinks per drinking day, mean (SD)    
 First trimesterb 8.3 (9.3) 5.9 (4.9) .14 
 Second trimesterb 7.4 (10.0) 3.8 (4.4) .04 
 Third trimesterb 5.3 (10.4) 1.8 (3.9) .02 

—, not applicable.

a

FASD: n = 50; not FASD: n = 42.

b

Among drinkers only in the specific time period.

Table 3 presents a comparison of study children at age 5 who were diagnosed with FASD with those children who were not. Children diagnosed with FAS, PFAS, or ARND demonstrated significantly higher total dysmorphology scores than those without FASD. No significant difference was observed between the FASD and non-FASD groups on length of gestation, sex, or age at the Bonferroni-adjusted value of 0.005 at age 5.

TABLE 3

Child Age, Sex, Growth, and Dysmorphology Comparisons

FASD (n = 79)Not FASD (n = 76)Pa
Weeks’ gestation at birth, mean (SD) 38.0 (2.7) 37.8 (3.4) .70 
Child’s age at T5, mo, mean (SD)b 58.8 (7.3) 56.4 (6.5) .03 
Sex, % male 40.5 43.4 .71 
Height percentile at T5, mean (SD)c 14.3 (17.0) 34.4 (26.6) <.001 
Height z score at T5, mean (SD) −0.4 (0.7) 0.4 (1.1) <.001 
Weight percentile at T5, mean (SD)c 10.4 (16.2) 33.5 (30.1) <.001 
Weight z score at T5, mean (SD) −0.4 (0.6) 0.5 (1.2) <.001 
BMI percentile at T5, mean (SD)c 23.4 (25.8) 41.4 (31.3) <.001 
BMI z score at T5, mean (SD) −0.3 (0.9) 0.3 (1.1) <.001 
OFC percentile at T5, mean (SD)c 8.8 (15.4) 30.2 (26.4) <.001 
OFC z score at T5, mean (SD) −0.5 (0.7) 0.5 (1.1) <.001 
PFL percentile at T5, mean (SD)c 11.4 (12.4) 26.5 (15.7) <.001 
PFL z score at T5, mean (SD) −35 (0.8) 0.4 (1.0) <.001 
Philtrum ranking at T5, mean (SD) 3.5 (0.9) 2.8 (0.7) <.001 
Vermilion ranking at T5, mean (SD) 3.6 (0.9) 2.9 (0.8) <.001 
Total dysmorphology score at T5, mean (SD) 14.2 (5.0) 7.0 (3.8) <.001 
FASD (n = 79)Not FASD (n = 76)Pa
Weeks’ gestation at birth, mean (SD) 38.0 (2.7) 37.8 (3.4) .70 
Child’s age at T5, mo, mean (SD)b 58.8 (7.3) 56.4 (6.5) .03 
Sex, % male 40.5 43.4 .71 
Height percentile at T5, mean (SD)c 14.3 (17.0) 34.4 (26.6) <.001 
Height z score at T5, mean (SD) −0.4 (0.7) 0.4 (1.1) <.001 
Weight percentile at T5, mean (SD)c 10.4 (16.2) 33.5 (30.1) <.001 
Weight z score at T5, mean (SD) −0.4 (0.6) 0.5 (1.2) <.001 
BMI percentile at T5, mean (SD)c 23.4 (25.8) 41.4 (31.3) <.001 
BMI z score at T5, mean (SD) −0.3 (0.9) 0.3 (1.1) <.001 
OFC percentile at T5, mean (SD)c 8.8 (15.4) 30.2 (26.4) <.001 
OFC z score at T5, mean (SD) −0.5 (0.7) 0.5 (1.1) <.001 
PFL percentile at T5, mean (SD)c 11.4 (12.4) 26.5 (15.7) <.001 
PFL z score at T5, mean (SD) −35 (0.8) 0.4 (1.0) <.001 
Philtrum ranking at T5, mean (SD) 3.5 (0.9) 2.8 (0.7) <.001 
Vermilion ranking at T5, mean (SD) 3.6 (0.9) 2.9 (0.8) <.001 
Total dysmorphology score at T5, mean (SD) 14.2 (5.0) 7.0 (3.8) <.001 

PFL, palpebral fissure length.

a

Bonferroni-adjusted significance level = 0.005.

b

Senior dysmorphologist examinations used for analysis.

c

Centers for Disease Control and Prevention percentiles.

As is the case with any medical condition, sound clinical judgment was exercised in assigning diagnoses in the FASD continuum. Among the differential diagnoses considered were genetic disorders or conditions arising from other teratogens. Additionally, because OFC, growth, and many cognitive and behavioral characteristics display moderate to high degrees of heritability, when such information was available about the family, these data were considered in final diagnostic decisions. In the children included in the final analysis, all other apparent diagnoses, genetic or otherwise, were ruled out on the basis of the clinical impression of the dysmorphologists.

The total dysmorphology score has proven to be a useful and discriminating tool in evaluating individuals for potential FASD (Table 4). Figure 2 reveals the trajectories of weight, OFC, and dysmorphology scores for children with FASD and those without FASD at age 5. Children with FASD weighed less and demonstrated smaller OFCs at all time points. The children with FASD at age 5 also displayed significantly higher dysmorphology scores throughout early life compared with those children without FASD. ROC analysis was performed to quantify the level of discrimination provided by the dysmorphology score by time period. The analysis revealed that the area under the curve (AUC) (discrimination value) was 0.772 at T1 (95% confidence interval [CI]: 0.695–0.848), 0.777 at T2 (see below), and 0.839 at T3 (95% CI: 0.705–0.849), each of which was statistically significant (P < .001). The ROC value at 9 months (T2) is illustrated in Fig 3, and the AUC of 0.777 (P < .001; 95% CI: 0.705–0.849) is both statistically significant and robust in light of other findings at T2 and later time periods. Data used to classify each of the specific FASD diagnoses independently are illustrated in Fig 4. This classification reveals that the dysmorphology score clearly discriminates between children with FASD and those without FASD at 9 months of age because each of the specific FASD diagnostic group lines become separated by total dysmorphology score alone. Moreover, the total dysmorphology score was also highly discriminating between children diagnosed with FAS and those without FASD from T1 to T5 (Fig 4, Supplemental Table 6). These data indicate that making any diagnosis of FASD versus non-FASD at 9 months of age through dysmorphology score alone is more strongly supported than doing so at an earlier age. Additionally, data from the post hoc analysis in Supplemental Table 6 also support this finding. At T3 (18 months), children with FAS are significantly discriminated from both those without FASD and those with ARND on the dysmorphology score. Furthermore, at T4 (42 months), the dysmorphology score also discriminated children with PFAS from those without FASD. Finally, the post hoc analysis (Supplemental Table 6) reveals that the differences between each of the specific FASD diagnoses are significant at T5 (60 months).

TABLE 4

Dysmorphology Scoring System (a Weighted Score Based on Analysis of the Frequency of Growth Restriction and Minor Anomalies in 370 Children With FAS)

FeatureScore
OFC ≤10% 
Growth deficiency  
 Height ≤10% 
 Weight ≤10% 
Short PFL (≤10%) 
Smooth philtrum 
Thin vermilion 
Hypoplastic midface 
Epicanthal folds 
Decreased IPD or ICD (≤25%) 
Flat nasal bridge 
Altered palmar crease 
Fifth-finger clinodactyly 
Long philtrum (≥90%) 
Anteverted nares 
Camptodactyly 
Ptosis 
“Railroad track” ears 
Heart murmur or confirmed CHD 
Strabismus 
Limited elbow supination 
Hypoplastic nails 
Prognathism 
Hypertrichosis 
Total possible score 41 
FeatureScore
OFC ≤10% 
Growth deficiency  
 Height ≤10% 
 Weight ≤10% 
Short PFL (≤10%) 
Smooth philtrum 
Thin vermilion 
Hypoplastic midface 
Epicanthal folds 
Decreased IPD or ICD (≤25%) 
Flat nasal bridge 
Altered palmar crease 
Fifth-finger clinodactyly 
Long philtrum (≥90%) 
Anteverted nares 
Camptodactyly 
Ptosis 
“Railroad track” ears 
Heart murmur or confirmed CHD 
Strabismus 
Limited elbow supination 
Hypoplastic nails 
Prognathism 
Hypertrichosis 
Total possible score 41 

Adapted from Hoyme HE, Kalberg WO, Elliott AJ, et al. Updated clinical guidelines for diagnosing fetal alcohol spectrum disorders. Pediatrics. 2016;138(2):e20154256. CHD, congenital heart disease; ICD, intercanthal distance; IPD, interpupillary distance; PFL, palpebral fissure length.

FIGURE 2

Weight, OFC, and total dysmorphology score over time.

FIGURE 2

Weight, OFC, and total dysmorphology score over time.

FIGURE 3

ROC analysis: AUC for accuracy of the total dysmorphology score in discriminating children with FASD from children without FASD at T2 (9 months of age) evaluation.

FIGURE 3

ROC analysis: AUC for accuracy of the total dysmorphology score in discriminating children with FASD from children without FASD at T2 (9 months of age) evaluation.

FIGURE 4

Total dysmorphology score over time by diagnosis at 60 months (5 years); error bars: ± 1 SE; N = 94 (total number of children seen at all 5 time points); repeated measures analysis, within subjects effect, time: F = 24.263, P < .001; repeated measures analysis, within subjects effect, time × group: F = 2.370, P < .002; repeated measures analysis, between subjects effect, group: F = 21.338, P < .001; Mauchly’s test of Sphericity has been violated: χ2(9) = 17.118, P = .047.

FIGURE 4

Total dysmorphology score over time by diagnosis at 60 months (5 years); error bars: ± 1 SE; N = 94 (total number of children seen at all 5 time points); repeated measures analysis, within subjects effect, time: F = 24.263, P < .001; repeated measures analysis, within subjects effect, time × group: F = 2.370, P < .002; repeated measures analysis, between subjects effect, group: F = 21.338, P < .001; Mauchly’s test of Sphericity has been violated: χ2(9) = 17.118, P = .047.

BNBAS-3 data at 6 weeks were explored to see if self-regulation would be lower for infants exposed to alcohol during pregnancy. Infants whose mothers were abstinent during pregnancy were compared with infants whose mothers drank during the first trimester of pregnancy. This comparison revealed significant differences on the examiner’s emotional response variable between children who were exposed to alcohol during the first trimester of pregnancy and those who were not exposed. There was also a significant difference between the children who were exposed and the children who were unexposed on the cost of attention and examiner facilitation variables only (Supplemental Table 7).

Cognitive scores from the BSID-III were analyzed on the basis of the assigned FASD diagnostic categories at T5; they began to differentiate among the groups between 9 and 18 months of age (Fig 5), especially those with FAS. The BSID-III did not discriminate among specific FASD diagnostic groups in a clinically interpretable way at any age (although the ROC analysis revealed significant discrimination [P = .003] of FASD versus non-FASD at 18 months [AUC = 0.628; 95% CI: 0.544–0.711]). By T5, when assessed by using the KABC-II, the various FASD groups continued to show increasing decline in overall cognitive abilities, as would be expected. The children with FAS were the most different from subjects without FASD (Table 5). Additional BSID-III testing results are presented in Supplemental Table 8 for all testing domains at each time period by diagnostic category. Significant differences among diagnostic categories began to emerge over T3 (18 months) and T4 (42 months).

FIGURE 5

Cognitive percentile scores over time; error bars: ± 1 SE; the BSID-III was used in T1 to T4; the KABC-II was used in T5; N = 57 (total number of children seen at all 4 time points); repeated measures analysis, within subjects effect, time: F = 8.687, P < .001; repeated measures analysis, within subjects effect, time × group: F = 1.340, P = .198; repeated measures analysis, between subjects effect, group: F = 0.174, P = .971; Mauchly’s test of Sphericity has been violated: χ2(5) = 22.736, P < .001.

FIGURE 5

Cognitive percentile scores over time; error bars: ± 1 SE; the BSID-III was used in T1 to T4; the KABC-II was used in T5; N = 57 (total number of children seen at all 4 time points); repeated measures analysis, within subjects effect, time: F = 8.687, P < .001; repeated measures analysis, within subjects effect, time × group: F = 1.340, P = .198; repeated measures analysis, between subjects effect, group: F = 0.174, P = .971; Mauchly’s test of Sphericity has been violated: χ2(5) = 22.736, P < .001.

TABLE 5

KABC-II Global Percentile Rank, KABC-II Simultaneous or GV Percentile Rank, and KABC Sequential Number Recall or GSM Perentile Rank Compared byDiagnostic Groups

FAS (n = 34)PFAS (n = 13)ARND (n = 18)Not FASD (n = 64)P
Mean (SD)Mean (SD)Mean (SD)Mean (SD)
KABC-II global percentile rank 4.6 (4.8) 6.2 (10.2) 6.0 (6.6) 19.1 (18.5) <.001a,b,c 
KABC-II simultaneous or GV percentile rank 2.5 (3.8) 5.1 (9.7) 2.7 (3.0) 16.0 (20.7) <.001a,b,c 
KABC-II sequential or GSM percentile rank 15.9 (17.6) 20.5 (18.7) 18.4 (16.0) 25.6 (22.0) .12 
KABC-II learning or GLR (Atlantis) percentile rank 29.2 (26.1) 23.1 (19.7) 30.7 (22.4) 50.4 (26.3) <.001a,b,c 
FAS (n = 34)PFAS (n = 13)ARND (n = 18)Not FASD (n = 64)P
Mean (SD)Mean (SD)Mean (SD)Mean (SD)
KABC-II global percentile rank 4.6 (4.8) 6.2 (10.2) 6.0 (6.6) 19.1 (18.5) <.001a,b,c 
KABC-II simultaneous or GV percentile rank 2.5 (3.8) 5.1 (9.7) 2.7 (3.0) 16.0 (20.7) <.001a,b,c 
KABC-II sequential or GSM percentile rank 15.9 (17.6) 20.5 (18.7) 18.4 (16.0) 25.6 (22.0) .12 
KABC-II learning or GLR (Atlantis) percentile rank 29.2 (26.1) 23.1 (19.7) 30.7 (22.4) 50.4 (26.3) <.001a,b,c 

GLR, long-term storage and retrieval; GSM, short term memory; GV, visual processing.

a

Post hoc comparisons are significantly different between FAS and not FASD.

b

Post hoc comparisons are significantly different between PFAS and not FASD.

c

Post hoc comparisons are significantly different between ARND and not FASD.

The KABC-II overall MPI global scores, the simultaneous visual processing scores, and the learning scores were significantly lower for children with FAS, PFAS, and ARND when compared by ANOVA with those for the children who did not have FASD (Table 5). Pairwise differences (individual diagnostic groups), were also significantly lower for each of the 3 FASD diagnostic groups than for those without FASD for global, simultaneous visual processing, and learning percentile ranks. Any percentile score equal to or less than the eighth percentile would place the child’s skills ≥1.5 SDs from the mean. All global and visual processing scores fell >1.5 SDs below the mean. The percentile-rank mean scores on the MPI and visual processing subtests range from 2.5% to 6.2%, which fall in the borderline range of ability for children diagnosed with FAS, PFAS, and ARND. The learning domain scores were also significantly lower for children with FASD, although the percentile scores fell in the average range. Interestingly, however, the short-term memory scores were not significantly different between children who were affected and those who were not.

This study reveals the accuracy of using dysmorphology assessments for FASD diagnostic purposes in children under the age of 3, especially in conjunction with appropriate behavioral evaluations. Such assessments progressively discriminate among specific FASD diagnostic groups, from 18 to 60 months (when the KABC-II can discriminate each of the FASD diagnostic groups from one another). These early signs from behavioral testing reveal significant potential, especially when coupled with expert dysmorphology evaluations, for guiding surveillance of children who may be at risk for a specific FASD diagnosis because of prenatal alcohol exposure.

Our data indicate that dysmorphic features differentiate between children with FASD and unaffected children as early as 9 months of age, although children with FAS who are severely affected can be identified earlier with a significant degree of probability.36,37  Data from the dysmorphology examinations reveal that the specificity of facial features observed in FAS and PFAS begins to differentiate among diagnostic categories within the FASD continuum at 9 months but do so most clearly from 18 months onward. Growth restriction, a clear criterion needed for a diagnosis of FAS, is evident at 9 months as well, if not before.

Assessment of early behavioral and developmental milestones and ROC analysis revealed significant developmental differences between children with FASD and those without FASD between 18 and 42 months of age. The results of the BNBAS-3 at T1 revealed that infants prenatally exposed to alcohol have significantly more difficulty maintaining a state of attention (cost of attention). Infants who are frailer or those who have a known developmental issue may have greater difficulty maintaining a state of attention.29  When an infant has such difficulty, the clinician may observe shallow or irregular breathing; motor disorganization and exhaustion displayed by flailing, hyper- or hypotonicity, jerkiness, or complete shut down into a sleeping state; or state overload displayed by crying, hiccups, yawns, regurgitating, or gagging. When an infant exhibits these traits, the examination becomes more difficult to complete or facilitate (examiner facilitation), and the examiner may experience more stress when conducting the assessment (examiner’s emotional response).

Children who were prenatally exposed to alcohol did not score as highly on the KABC-II (MPI global score) at 60 months and had the most difficulty in the visual processing subtests. Conceptual thinking and overall cognitive deficits were reported in the earliest studies of children with FASD.38,39  Although children who were affected scored significantly lower on the learning cluster, their scores were still in the average range.

Attrition of 44 children because of incomplete data (of 199 entering the evaluation protocol) did not skew the results. There were no statistical differences (P < .05) in height, weight, OFC, total dysmorphology score, or BSID-III cognitive percentile at T3, T4, or T5 among those with a diagnosis of FASD or non-FASD who were discussed at case conference and those excluded because of insufficient data or not being assessed at T5. Of the 44 children who were excluded from analysis, 23 were discussed in detail at a multidisciplinary case conference at 60 months of age and determined to have insufficient data for final analysis in the study. Eighteen were determined to have insufficient data because of the lack of testing (KABC-II), and none were found to have insufficient data because of a lack of alcohol exposure information. Five had insufficient data because both testing and alcohol exposure information were missing. Twenty-one individuals were not evaluated by the dysmorphologists at T5. Of the 21 not evaluated at T5, 9 were last assessed at T4, and 12 were last assessed at T3.

Other investigators previously have attempted to determine features in early infancy that are most predictive of FASD as the child with alcohol exposure grows older. Coles et al40  used neonatal microcephaly (OFC less than the fifth percentile), pre- or postnatal growth deficiency, and level of self-reported maternal alcohol use during pregnancy to predict which infants at 6 or 12 months would be at highest risk for developmental delay. Stoler and Holmes41  employed a standardized facial assessment scale to attempt to differentiate newborns exposed to alcohol from newborns not exposed to alcohol in a blinded fashion. The facial assessment differentiated the groups to some extent; however, individual diagnoses were not assigned and developmental follow-up was not attempted. Thus, the predictive value of their facial assessment scale in terms of developmental outcome in individual children with alcohol exposure was not evaluated. Carter et al42  demonstrated that growth trajectory in infancy and childhood is a marker of which heavily alcohol-exposed infants are at greatest risk of cognitive developmental deficits, with those displaying prenatal growth restriction, persistent postnatally, at greatest risk. Mesa et al43  recently published results of a study in which cardiac-orienting responses were used as an early scalable biomarker of alcohol-related neurodevelopmental impairment. The study revealed that the cardiac-orienting response at 6 months was more predictive of developmental delay on the 12-month BSID-III testing than the 6-month BSID-III score.43  Although further work is needed to determine the long-term predictive accuracy of this technique, cardiac-orienting responses may be an early and straightforward way to determine individual risk for later developmental delay.

Although diagnoses at the more severe end of the FASD continuum may be suspected in early infancy, no accepted diagnostic criteria for infants and preschool-aged children with FASD currently exist. Thus, such diagnoses are, by definition, preliminary until more comprehensive neurobehavioral testing can be accomplished when children reach school age. However, lack of defined and accepted diagnostic criteria for FASD in infancy should not deter referral of infants exposed to alcohol for appropriate diagnostic services. Children with prenatal alcohol exposure (especially those demonstrating developmental delay or behavioral concerns or those who are in foster care) should be evaluated at any age. Even without a definitive FASD diagnosis, such children may benefit from specific interventions and therapies.

Although any child with prenatal alcohol exposure falls into an at-risk category for developmental disabilities, in the current study, we have established that determination of a diagnosis within the FASD continuum is possible earlier in childhood than has previously been appreciated. Assessment of a combination of growth, dysmorphic features, and neurobehavioral characteristics allows for accurate identification of most children with FASD as early as 9 to 18 months.

Ms Kalberg is a co–principal investigator for the Oxnard Foundation–funded portion of this project and conceptualized and designed the study, helped design the neurodevelopmental assessment protocol for enrolled infants and children, drafted the initial manuscript, and serves as the first author; Dr May is the principal investigator of the National Institute on Alcohol Abuse and Alcoholism–funded studies in South Africa on which this article is based and substantially contributed to the conception and design of the study as well as to the analysis and interpretation of data gathered and revised the manuscript critically for important intellectual content, including all statistical analyses and epidemiological designs and analyses; Mr Buckley and Ms Hasken make up the data analysis group for the fetal alcohol spectrum disorders team collaboration and supervised the acquisition, storing, and analysis of sensitive subject data, produced the tables and figures for the manuscript, and revised the manuscript critically for important intellectual content; Ms Marais and Ms De Vries are the local South African project managers of the present longitudinal cohort study and substantially contributed to the conception and design of the study as well as to the acquisition, analysis, and interpretation of data and revised the article critically for important intellectual content; Drs Bezuidenhout, Manning, Robinson, Adam, and Derek B. Hoyme substantially aided in the conception and design of the study, extensively contributed to data acquisition by providing dysmorphology examinations of all children in the current study, extensively contributed to the analysis and interpretation of data by assigning final diagnoses to all subjects in multidisciplinary case conferences throughout the project, and revised the article critically for important intellectual content; Drs Parry and Seedat are the South African coinvestigators of the National Institute on Alcohol Abuse and Alcoholism–funded studies on which this article is based and actively participated in the conception and design of the study (including providing critical input about and liaison with the local South African communities in which subjects were recruited), participated in the vital analysis and interpretation of data, and revised the article critically for important intellectual content; Dr Elliott assisted with the design of the current investigation (specifically with substantial planning of the neurodevelopmental assessment protocol for infants and children in the study), trained the South African staff in the administration of the neurodevelopmental tests performed, participated in the analysis and interpretation of the neurodevelopmental data gathered, and critically revised the manuscript for important intellectual content; Dr H. Eugene Hoyme is a coprincipal investigator for the Oxnard Foundation–funded portion of the current study and conceptualized and designed the study, analyzed and interpreted the growth and dysmorphology data acquired, reviewed the manuscript critically for important intellectual content, and performed the final edits of the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

FUNDING: Funded by National Institute on Alcohol Abuse and Alcoholism grants R01 AA11685 and R01/UO1 AA01115134. Funding was also provided by the Oxnard Foundation (Newport Beach, CA). Funded by the National Institutes of Health (NIH).

     
  • ANOVA

    analysis of variance

  •  
  • ARND

    alcohol-related neurodevelopmental disorder

  •  
  • AUC

    area under thecurve

  •  
  • AUDIT

    Alcohol Use Disorders Test

  •  
  • BNBAS-3

    Brazelton Neonatal Behavioral Assessment Scale

  •  
  • BSID-III

    Bayley Scales of Infant Development, Third Edition

  •  
  • CI

    confidence interval

  •  
  • CoFASP

    Collaboration on Fetal Alcohol Spectrum Disorder Prevalence

  •  
  • FAS

    fetal alcohol syndrome

  •  
  • FASD

    fetal alcohol spectrum disorder

  •  
  • KABC-II

    Kaufman Assessment Battery for Children, Second Edition

  •  
  • MPI

    mental processing index

  •  
  • OFC

    occipitofrontal (head) circumference

  •  
  • PFAS

    partial fetal alcohol syndrome

  •  
  • ROC

    receiver operating characteristic

  •  
  • T1

    time point 1

  •  
  • T2

    time point 2

  •  
  • T3

    time point 3

  •  
  • T4

    time point 4

  •  
  • T5

    time point 5

1
Sampson
PD
,
Streissguth
AP
,
Bookstein
FL
, et al
.
Incidence of fetal alcohol syndrome and prevalence of alcohol-related neurodevelopmental disorder
.
Teratology
.
1997
;
56
(
5
):
317
326
2
May
PA
,
Gossage
JP
,
Kalberg
WO
, et al
.
Prevalence and epidemiologic characteristics of FASD from various research methods with an emphasis on recent in-school studies
.
Dev Disabil Res Rev
.
2009
;
15
(
3
):
176
192
3
May
PA
,
Baete
A
,
Russo
J
, et al
.
Prevalence and characteristics of fetal alcohol spectrum disorders
.
Pediatrics
.
2014
;
134
(
5
):
855
866
4
May
PA
,
Chambers
CD
,
Kalberg
WO
, et al
.
Prevalence of fetal alcohol spectrum disorders in 4 US communities
.
JAMA
.
2018
;
319
(
5
):
474
482
5
Lange
S
,
Probst
C
,
Gmel
G
, et al
.
Global prevalence of fetal alcohol spectrum disorder among children and youth: a systematic review and meta-analysis
.
JAMA Pediatr
.
2017
;
171
(
10
):
948
956
6
May
PA
,
de Vries
MM
,
Marais
AS
, et al
.
The continuum of fetal alcohol spectrum disorders in four rural communities in South Africa: prevalence and characteristics
.
Drug Alcohol Depend
.
2016
;
159
:
207
218
7
May
PA
,
Marais
AS
,
de Vries
MM
, et al
.
The continuum of fetal alcohol spectrum disorders in a community in South Africa: prevalence and characteristics in a fifth sample
.
Drug Alcohol Depend
.
2016
;
168
:
274
286
8
Institute of Medicine
;
Committee to Study Fetal Alcohol Syndrome
. In:
Stratton
K
,
Howe
C
,
Battaglia
F
, eds.
Fetal Alcohol Syndrome: Diagnosis, Epidemiology, Prevention, and Treatment
.
Washington, DC
:
The National Academy Press
;
1996
9
Hoyme
HE
,
May
PA
,
Kalberg
WO
, et al
.
A practical clinical approach to diagnosis of fetal alcohol spectrum disorders: clarification of the 1996 Institute of Medicine criteria
.
Pediatrics
.
2005
;
115
(
1
):
39
47
10
Hoyme
HE
,
Kalberg
WO
,
Elliott
AJ
, et al
.
Updated clinical guidelines for diagnosing fetal alcohol spectrum disorders
.
Pediatrics
.
2016
;
138
(
2
):
e20154256
11
Knudsen
EI
.
Sensitive periods in the development of the brain and behavior
.
J Cogn Neurosci
.
2004
;
16
(
8
):
1412
1425
12
Engle
PL
,
Black
MM
,
Behrman
JR
, et al;
International Child Development Steering Group
.
Strategies to avoid the loss of developmental potential in more than 200 million children in the developing world
.
Lancet
.
2007
;
369
(
9557
):
229
242
13
Little
BB
,
Snell
LM
,
Rosenfeld
CR
,
Gilstrap
LC
 III
,
Gant
NF
.
Failure to recognize fetal alcohol syndrome in newborn infants
.
Am J Dis Child
.
1990
;
144
(
10
):
1142
1146
14
Astley
SJ
,
Clarren
SK
.
Diagnosing the full spectrum of fetal alcohol-exposed individuals: introducing the 4-digit diagnostic code
.
Alcohol Alcohol
.
2000
;
35
(
4
):
400
410
15
Astley
S
.
FASD 4-Digit Diagnostic Code (2004). Available at: https://depts.washington.edu/fasdpn/htmls/4-digit-code.htm. Accessed June 9, 2019
16
Chudley
AE
,
Conry
J
,
Cook
JL
, et al;
Public Health Agency of Canada’s National Advisory Committee on Fetal Alcohol Spectrum Disorder
.
Fetal alcohol spectrum disorder: Canadian guidelines for diagnosis
.
CMAJ
.
2005
;
172
(
suppl 5
):
S1
S21
17
Cook
JL
,
Green
CR
,
Lilley
CM
, et al;
Canada Fetal Alcohol Spectrum Disorder Research Network
.
Fetal alcohol spectrum disorder: a guideline for diagnosis across the lifespan
.
CMAJ
.
2016
;
188
(
3
):
191
197
18
Viljoen
DL
,
Gossage
JP
,
Brooke
L
, et al
.
Fetal alcohol syndrome epidemiology in a South African community: a second study of a very high prevalence area
.
J Stud Alcohol
.
2005
;
66
(
5
):
593
604
19
May
PA
,
Brooke
L
,
Gossage
JP
, et al
.
Epidemiology of fetal alcohol syndrome in a South African community in the Western Cape Province
.
Am J Public Health
.
2000
;
90
(
12
):
1905
1912
20
May
PA
,
Gossage
JP
,
Marais
AS
, et al
.
The epidemiology of fetal alcohol syndrome and partial FAS in a South African community
.
Drug Alcohol Depend
.
2007
;
88
(
2–3
):
259
271
21
May
PA
,
Blankenship
J
,
Marais
AS
, et al
.
Approaching the prevalence of the full spectrum of fetal alcohol spectrum disorders in a South African population-based study
.
Alcohol Clin Exp Res
.
2013
;
37
(
5
):
818
830
22
Census
2011
. Provincial profile: Western Cape. Available at: http://www.statssa.gov.za/publications/Report-03-01-70/Report-03-01-702011.pdf. Accessed October 1, 2019.
23
Brinkmann
S
;
Department of Health of the Western Cape
.
Birth and Antenatal Data, Information Management
.
Cape Town, South Africa
:
Department of Health, Western Cape Government
;
2008
24
de Vries
MM
,
Joubert
B
,
Cloete
M
, et al
.
Indicated prevention of fetal alcohol spectrum disorders in South Africa: effectiveness of case management
.
Int J Environ Res Public Health
.
2015
;
13
(
1
):
76
25
Babor
TF
,
Higgins-Biddle
JC
.
Alcohol screening and brief intervention: dissemination strategies for medical practice and public health
.
Addiction
.
2000
;
95
(
5
):
677
686
26
Dawson
DA
.
Methodological issues in measuring alcohol use
.
Alcohol Res Health
.
2003
;
27
(
1
):
18
29
27
Porges
SW
.
Physiological regulation in high-risk infants: a model for assessment and potential intervention
.
Dev Psychopathol
.
1996
;
8
(
1
):
43
58
28
Brazelton
TB
,
Nugent
JK
.
Neonatal Behavioral Assessment Scale
, 3rd ed.
London, United Kingdom
:
Mac Keith Press
;
1995
29
Bayley
N
.
Bayley Scales of Infant and Toddler Development
, 3rd ed.
San Antonio, TX
:
Pearson Publishers
;
2006
30
Kaufman
AS
,
Kaufman
NL
.
Kaufman Assessment Battery for Children
, 2nd ed.
Bloomington, MN
:
Pearson Executive Office
;
2004
31
Skuy
M
,
Taylor
M
,
O’Carroll
S
,
Fridjhon
P
,
Rosenthal
L
.
Performance of black and white South African children on the Wechsler Intelligence Scale for Children–revised and the Kaufman Assessment Battery
.
Psychol Rep
.
2000
;
86
(
3, pt 1
):
727
737
32
Bangirana
P
,
Seggane-Musisi
,
Allebeck
P
, et al
.
A preliminary examination of the construct validity of the KABC-II in Ugandan children with a history of cerebral malaria
.
Afr Health Sci
.
2009
;
9
(
3
):
186
192
33
IBM SPSS Statistics for Windows [Computer Program]. Version 22.0
.
Armonk, NY
:
IBM Corporation
;
2013
34
May
PA
,
Hasken
JM
,
De Vries
MM
, et al
.
A utilitarian comparison of two alcohol use biomarkers with self-reported drinking history collected in antenatal clinics
.
Reprod Toxicol
.
2018
;
77
:
25
32
35
Babor
TF
,
de la Fuente
JR
,
Saunders
J
,
Grant
M
.
The Alcohol Use Disorders Identification Test. Guidelines for Use in Primary Health Care
, 1st ed.
Geneva, Switzerland
:
World Health Organization
;
1992
36
Jones
KL
,
Smith
DW
.
Recognition of the fetal alcohol syndrome in early infancy
.
Lancet
.
1973
;
302
(
7836
):
999
1001
37
Tangsermkijsakul
A
.
Fetal alcohol syndrome in sudden unexpected death in infancy: a case report in medicolegal autopsy
.
Am J Forensic Med Pathol
.
2016
;
37
(
1
):
9
13
38
Streissguth
AP
,
LaDue
RA
.
Fetal alcohol. Teratogenic causes of developmental disabilities
.
Monogr Am Assoc Ment Defic
.
1987
;(
8
):
1
32
39
Streissguth
AP
,
Barr
HM
,
Kogan
J
,
Bookstein
FL
.
Understanding the Occurrence of Secondary Disabilities in Clients With Fetal Alcohol Syndrome (FAS) and Fetal Alcohol Effects (FAE). Final Report
.
Seattle, WA
:
University of Washington School of Medicine, Department of Psychiatry and Behavioral Sciences
;
1996
40
Coles
CD
,
Kable
JA
,
Drews-Botsch
C
,
Falek
A
.
Early identification of risk for effects of prenatal alcohol exposure
.
J Stud Alcohol
.
2000
;
61
(
4
):
607
616
41
Stoler
JM
,
Holmes
LB
.
Recognition of facial features of fetal alcohol syndrome in the newborn
.
Am J Med Genet C Semin Med Genet
.
2004
;
127C
(
1
):
21
27
42
Carter
RC
,
Jacobson
JL
,
Molteno
CD
, et al
.
Fetal alcohol growth restriction and cognitive impairment
.
Pediatrics
.
2016
;
138
(
2
):
e20160775
43
Mesa
DA
,
Kable
JA
,
Coles
CD
, et al;
CIFASD
.
The use of cardiac orienting responses as an early and scalable biomarker of alcohol-related neurodevelopmental impairment
.
Alcohol Clin Exp Res
.
2017
;
41
(
1
):
128
138

Competing Interests

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.

Supplementary data