A Fetal Alcohol Behavior Scale

zy zyxwvu zy zy zyxwvutsrqponm 0145-600819812202-0325$03.00/0 ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH Val. 22, No. 2 April 1998 A Fetal Alcohol Behavior Scale Ann P. Streissguth, Fred L. Bookstein, Helen M . Barr, Shoshanna Press, and Paul D. Sampson This research aimed to develop a Fetal Alcohol Behavior Scale (FABS) that describes the behavioral essence of fetal alcohol syndrome (FAS) and fetal alcohol effects (FAE), regardless of age, race, sex, and 10.Using a referencesample of 472 diagnosed patientswith FAS or FAE, ages 2 to 51, five studies are described.The FABS demonstrates high item-to-scale reliability (Cronbach's (Y = 0.91) and good test-retest reliability (r = 0.69) over an average interval of 5 years. It identifies many of the subjects with known or presumed prenatal alcohol exposure in detection studies using both prison and general samples. FABS scores also predict dependent living among adult patients with FASIFAE. The FABS is uncorrelatedwith 10, sex, age, race, and diagnosis (FAS versus FAE). We outline areas of further work to define the specificity and utility of this FABS. Key Words: Fetal alcohol syndrome (FAS), Fetal alcohol effects, Behavioralteratology,Alcohol-relatedneurodevelopmentaldisorder (ARND), Behavior scale. ETAL ALCOHOL syndrome (FAS) and fetal alcohol effects (FAE) are important causes of developmental disabilities in children and adults.'-7 However, because of the subtle and variable nature of the defining physical and central nervous system (CNS) characteristics and their changes with age,879many children and adults with this disability are never diagnosed and so are cut off from the services and interventions that might help them. The goal of this research is to construct a short, easy-to-administer scale that will capture the behavioral essence of FAS and FAE, regardless of age, race, sex, or IQ, and thus have utility across various populations and across the life span. The diagnosis of FAS has traditionally been based on three types of criteria: growth deficiency apparent at birth; a pattern of dysmorphic features primarily recognizable in the face; and some manifestations of CNS dysfunction.1,8,10-15 As growth deficiency and CNS dysfunction have many causes, the facial dysmorphology has historically been the distinguishing feature linking this birth defect with its prenatal alcohol etiology. The CNS dysfunction associ- F ated with FAS, while more prevalent as a prenatal alcohol effect, has not been considered as specific or unique as the facial dysmorphology. Since the mid-l970s, it has been clear that the range of intellectual disabilities associated with the FAS diagnosis was very broad16 and that no specific level of intellectual functioning could ever reasonably serve as the distinguishing CNS characteristic for defining FAS. Some children with FAS are mentally retarded, but many are not. On the other hand, confining the concept of alcohol-related birth defects to just those produced during the specific phase of prenatal exposure necessary for producing facial dysmorphology seems overly restrictive in light of recent The term FAE (as well as PFAS and PFAE, terms used for probable or possible FAS or FAE) has been used historically for patients who have a history of prenatal alcohol exposure and have some but not all of the characteristics of FAS.',22 Partial manifestations of a birth defects syndrome are not unusual and, in the case of alcohol, a large number of CNS characteristics have been linked to prenatal alcohol exposure, both in animal studies and human st~dies.'~-'~ The Institute of Medicine's report on FAS9 addressed this issue by suggesting a new term, alcohol-related neurodevelopmental disorder (ARND), to reflect the CNS component of fetal alcohol effects. The ARND criteria include structural brain anomalies, decreased cranial size at birth, neurological hard and soft signs, and/or evidence of a complex pattern of behavior or cognitive abnormalities that are inconsistent with developmental level and cannot be explained by familial background or environment alone. Although general categories of cognitive abnormalities (i.e., learning difficulties, problems with memory) and behavioral abnormality (i.e., poor impulse control, poor judgment) were listed, no specific criteria were suggested. Research from our laboratory has revealed a wide variety of cognitive, neuropsychological, and learning disabilities associated with prenatal alcohol exposure in a long-term prospective epidemiological ~ t u d y . ~ We ~ - *have ~ recently clustered these into a performance-based neurodevelopmental framework for estimating the prevalence of FAE, including FAS and ARND.27 The studies described in the present paper are an attempt to quantify the behavioral phenotype of fetal alcohol. Clinically, individuals identified as either FAS or FAE often share a similar behavioral profile.5328Despite the wide range of primary disabilities that people with FAS and FAE may manifest, their parents and caretakers often describe them in terms of some relatively characteristic be- zyxwvutsrq zyxwvutsrqp zyxwvutsrqpo z zyxwvutsrq From the Department of Psychiatiy and Behavioral Sciences (A.P.S., H.M.B., S.P.), University of Washington School of Medicine, Seattle, Washington;Institute of Gerontology (F.L.B.), University of Michigan, Ann Arbor, Michigan; and the Department of Statistics (P.D.S.), University of Washington School of Arts and Sciences, Seattle, Washington. Presented at the 1996 Borchard Foundation Symposium on the Behavioral Effects in Children following Prenatal Alcohol Exposure, Missillac, France, July 28-30, 1996. This study was funded by the Centers for Disease Control (Grant R041 CCROO8515-01-04) and by the National Institute on Alcohol Abuse and Alcoholism (Grant ROI-AA01455-01-22). Reprint requests: Ann P. Streissguth, Ph.D., Department of Psychiatry and Behavioral Sciences, University of Washington, School of Medicine, Box 359112, Seattle, WA 98195. Copyright 0 1998 by The Research Society on Alcoholism. Alcohol Clin Exp Res, Val 22, No 2, 1998: pp 325-333 325 zyxwvutsrqponmlkj zy zyxwvutsrqp 16 STREISSGUTH ET AL. Table 1. Demographic Characteristics of the Five Study Samples Demographic characteristic Age at evaluation Reference sample (n = 472) Detection study sample (n = 81) - Normative sample Test-retest sample Prediction study sample 8’ 8 14 7 0 0 7 63 (n = 186) (n = 37) (n = 70) 2-6 7-1 1 12-17 18-51 Sex 165 96 124 87 1 80 52 20 28 86 Male Female Ethnicity White Black Native American Other Alcohol-related diagnosis FAS 271 201 81 0 96 85 21 16 36 34 273 31 129 37 76 3 2 0 125 17 4 34 24 2 11 0 45 4 20 1 169 303 - - 27 10 43 27 FAE * Age - at initial administration of the PBC haviors. For example, “Talks a lot but says little; is chatty but with shallow content.” “Makes ‘off the wall’ comments; sometimes says things that seem completely out of context.” “Overreacts to situations; emotional responses are often stronger than you would expect.” “Often demands attention or monopolizes a conversation.” Such behavioral descriptors by parents, which predated popular writings about FAS, refer to characteristics that transcend scores on performance tests, and as a group, do not seem characteristic of other childhood disorders. Over the years, these behavioral descriptors were transcribed by Streissguth as they were used by parents and caretakers to describe their children with FAS and FAE. Those most frequently encountered were assembled into a list of 68 short descriptors, which we called a “Personal Behaviors Checklist” (PBC), that could be answered in a yesho format by someone familiar with the child’s behavior. For many years, parents and caretakers of patients with FASFAE associated with our Fetal Alcohol and Drug Unit routinely filled out PBCs. The purpose of this paper is to report a series of studies conducted over a period of several years that used these data more formally. First, we performed a Derivation Study to condense this checklist into a scale that we called the Fetal Alcohol Behavior Scale (FABS). Then, we conducted a Detection Study to see whether this scale could be used to detect people with FAS or FAE from among a deviant subgroup of the population (i.e., those in prison). Then we did a Normative Study to determine the sensitivity of this scale for identifying children of mothers with alcohol problems from within a nonclient sample of parents. Next, we performed a Test-Retest Study to evaluate the stability of these items in describing an individual’s behavior over time. Finally, we conducted a Prediction Study to see whether the scale could have any value in predicting to dependent living as an adult. Standard demographic descriptors for each of these five samples are collected in Table 1. The five studies are described herein. zyxwvutsrq METHODS A PBC was developed by Streissguth in the 1970s comprised of items used by parents and caretakers to describe their children with FASPAE. Items represented the following categories: communication and speech, personal manner, emotions, motor skills and activities, social skills and interactions, academic/work performance, and bodily and physiological functions. A total of 472 patients who had previously been diagnosed as FAS, PFAS, FAE, or PFAE by experienced dysmorphologists,were ultimately rated on the PBC by their parents or caretakers (see Table 1, column 1, for demographic characteristics). These comprised the FAS/ FAE reference sample. Subsets of this reference sample were used for the Derivation Study, the Test-Retest Study, and the Prediction Study. The Derivation Study zy zyx By 1994, PBCs were available on 134 patients under the age of 35 years. These data were used for principal components analyses based on covariance matrices for each of the four age groups defined in Table 1. Individuals scoring high on the first principal component are considered to reflect the behavioral essence of FAS and/or FAE. To define a scale less than half the length of the PBC, we selected those PBC items having high item-to-scale correlation for each of the four age groups. We defined a 26-item FABS by requiring minimum item-to-scale correlation of 0.32 or better across the four age groups. By 1995, a larger, more representative sample of PBCs for patients under the age of 35 years (n = 322) was available, and a second principal components analysis was conducted, resulting in a second scale with high item-to-scale validity (also 0.32). These two scales correlated 0.92. The 36 items representing the union of the items selected by these two analyses is referred to as the FABS. The Detection Study The Detection Study was conducted in late 1994, after the first principal components analysis of the Derivation Study had been performed. The Detection Study was conducted in a special unit for developmentally disabled, emotionally disturbed (nonpsychotic) male inmates within the Washington State Prisons System. Eighty-one of these inmates met the study criterion of having been known to the prison staff for at least 3 months and consented to participate. They are described in Table 1 (column 2). The respondents, two corrections officers, and a prison counselor filled out a single FABS on each of the 81 men by group consensus. zyxwvuts zyxwvu FETAL ALCOHOL BEHAVIOR SCALE impulsive stubborn unaware conseq poor attention cutel ixie ish can’t Eke hint incompl tasks tantrums too easily led overreacts mood sw)ngs Sensitive interrupts poor judgement center of attn fearless looses things people oriented overstimulated very active likeA%ZX demands attentn tries hard,but... opinionated out of context overly friend1 sleeping prx item attracted superf.friends hygiene prb loves to climb messy touches freq indistinct speech chats,no content zyxwvutsr 321 A A V A A V A V v’ A A A A V V zyxwvu A A V A A A V V V zyxwvutsrqponmlk A Fig. 1. The 68 items on the PBC ranked according to percent “yes“ responses by parenWinformants on 472 patients with FAS/FAE. For full text of the items, contact the senior author. poor manners talks fast not capable canfhp;y?$? dif’ty performing seems brighter dif‘ty learning unusual topic low self-esteem toilet training inappro lhome inappropkutside repeats often noise sensitive enjoys fixing freq phrases feeding prb ioud,unus voice vision prb light sensitive prb sex funct flirts hearing prb) stomach aches dry skin unusual smell bangs head rocks poor sch attnd V V A A v 0% zyxwvutsrqpon ; zyxwvutsrqp zyxwvu zyxwvutsrqp zyxwvutsr V I V A V V V V V V V V 25% i Patients with FAS/FAE (FAS/FAk reference sample, n=472) ! 36 items chosen for FABS A 32iitems not chosen for FABS v 50% percent ‘yes’ 75% Information on whether or not each inmate had a biological mother with alcohol problems was obtained independently from questionnaires filled out by the inmates themselves. The Normative Study The Normative Study of the FABS was conducted among parents in a general practice waiting room at the University of Washington Medical Center in 1995 after the second principal components analysis was completed.” While waiting to be called for their appointments, all consenting adults who had children of any age (n = 186) were asked to fill out a questionnaire regarding one of their children selected randomly by Shoshonna Press (see Table 1, column 3). Average time to complete the questionnaire (the FABS and a few demographic questions) was approximately 5 min. In this normative sample, Cronbach’s coefficient a was equal to 0.89, indicating satisfactorilyhigh item-to-scalereliabilit~.~’ If one were to divide the FABS scale into two subscales in all possible ways, this value of 0.89 is a weighted average of all the correlations between the scores on the two halves. 100% The Test-Retest Reliability Study There were 41 patients in the FAS/FAE reference sample of 472 who had had two PBCs filled out at two different ages by the same respondent. The 37 patients with at least a 1-year interval between the two FABS scores comprised the sample for the Test-Retest Study. Mean age at the first PBC was 13.4 years; at the second, 18.7 years (see Table 1,column 4). The Prediction Study The Prediction Study conducted in 1997 included a subset of 70 adults with FASFAE from the FASPAE reference sample who had previously had a PBC and later had a Life History Interview (LHI) administered to a caretaker, spouse, or informant in 1996 at least 1 year after the PBC?l Of particular interest to this FABS study was a summaly score on the LHI called Dependent Living, indicating that the adult with FASPAE was unable to live independentl~.~~ Mean age of patients in the Prediction Study at the time of evaluation for Dependent Living on the LHI was 28.1 years. 328 zyxwvutsrqponml STREISSGUTH ET AL. overreacts chats.no content unusual topic demands attentn unaware conseq incomplete tasks inapproploutside likes to talk interrupts center of attn touches freq can’t ia team canfta{e hint sleeping prb poor manners overstimulated out of context klutz hygiene prg phys loving repeats often messy talks fast inapgrophome supe icial friend fidgety poor jud ement dif’ty per8orming loud,unus voice overly friendly loses things noise sensitv mood swings poor attention prb sex funct tries hard but... ... freq phases fearless item attracted flirts too easily led indistinct speech tantrums opinionated stomach aches light sensitive impulsive low self-esteem toilet training not capable ve active gubborn people oriented seems &; :b loves to climb unusual smell dif‘ty learning rocks hearing prb dry skin cutelpixyjsh sensitive bangs head poor sch attnd vision prb enjoys fixin feeding prg A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A zyxwvutsrq zyxw zyx zyxwvutsrq zyxwvutsrq A A A A Fig. 2. The same 68 items on the PBC as in Fig. 1, ranked according to their item-to-scale correlations. The correlation plotted is the maximum (across the two derivation analyses, n = 134 and 322, respectively) of the minimum item-to-scale correlation over the four age groups of Table 1. v V V n V V V V V V V zyxwv zyxwvutsrqpo V V V V Patients with FASIFAE (two derivation samples, n=l34/322) 36 items chosen for FABS A 32 items not chosen for FABS v V V V V V V V V V zyxwvutsrq zyxwvutsrq 0 0.2 0.4 0.6 item-to-scale correlation RESULTS ing to their item-to-scale correlations. These range from “Overreacts . . .,” with an item-to-scale correlation of 0.58, Deriving the FABS from the PBC Items down to “Tries hard, but . . .,” with an item-to-scale correFigure 1 ranks the frequency with which each of the 68 lation of 0.32. Items in Figs. 1 and 2 with open triangles did items in the PBC was endorsed by the 472 parentdcaretaknot meet this criterion. Cronbach’s coefficient ly30 was ers in describing an individual with F A S F M . “Impulsive” equal to 0.91 in the Derivation Study sample of 322, indiand “stubborn” at the top are the two most frequently cating high item-to-scale reliability in the FABS. endorsed items at over 85%. “Poor school attendance” and “Scree plots” (not shown) indicate that there is only one “rocks rhythmically” are the least frequently endorsed main principal component of the items contributing to the items at under 20%. But, we did not specifically select items FABS in any of the four age groups. The item frequencies for the FABS because of having a high frequency. In fact, are sufficiently close to 50% that we can replace the score those selected have frequencies between 29% and 79%. on this first principal component by a simple sum of “yes’ The black triangles in Figs. 1 and 2 represent those items responses to the FABS items, which yields the FABS sumselected for the FABS from the PBC according to their item-to-scale correlations in the principal components for mary score. Figure 3 shows that the FABS score computed in this way each of the four age blocks examined. Figure 2 presents the same 68 items as Fig. 1, but here they are ordered accord- is independent of not only the patients’ ages, but also IQ, FETAL ALCOHOL BEHAVIOR SCALE zyxw zyxwv zyxwvutsrqp zyxw 329 zyxwvu . ............ . ....... .... . . .. zyx . ....... -.-. 8 ........ ..... .-. .... . . r= -0.03 36 h -r Relationship of FABS to other common determinants of outcome II 5 24 v) m a U -g (data from FASlFAE reference sample) .- 12 cb m 0 ......... .._.."_ .... ....... - . . .. .. .... -..... ...... .............. ................ . ..-..-. ........ .. .-...-... . . ............ ...-..... .. . . ......... . .......... ........... . . . . . . ......... .. .. . .. .. . . . . .. . ....... ....... . . ....... . . 40 60 80 100 120 140 Full Scale IQ age at FABS evaluation (years) zyxwvutsrq 36 3 v) 24 c E .- $12 0 male n= 273 129 31 39 Wh NAm BI 0th female Sex of Client Total Race 472 FAS Total FAE Alc-related Diagnosis Total Fig. 3. Relationship of the FABS to age, IQ, sex, race, and FAS versus FAE. The box in a boxplot spans the middle half of the FABS scores from the 25th to 75th percentiles. The dark horizontal line within a box indicates the median FABS score. The whiskers extending vertically from each box reach to the lowest and highest FABS. The open boxes on the lower and upper whiskers indicate the 10th and 90th percentiles respectively. Wh, White; NAm, Native American; El, Black oth, other: Alc, alcohol. sex, race, and diagnosis (FAS or FAE). Thus, no adjust- were on inmates who themselves had independently rements of the FABS summary score are needed for any of ported having mothers with alcohol problems. these potential confounds. In the FAS/FAE reference sample, individual patients' FABS scores ranged from 0 to 36 Using the FABS to Detect Children with Alcoholic Mothers (Fig. 4). The mean FABS score in this sample of 472 fiom Among a Group of Parents patients with FAS/FAE is 20.3; the median 21.0. Seven The boxplots in Fig. 6 show fairly impressive separation percent of the sample had a score of 6 or less. In the Detection Study and the Normative Study (see of the FABS scores of the FASFAE reference sample from herein), the FABS scores of this reference sample of 472 the FABS scores of the Normative Study sample. The mean patients with FASFAE are compared with other populations FABS score in the Normative Study sample is 6.6 (on the to see whether individuals with heavy prenatal alcohol expo- 1995 35-item FABS). About 80% of the FABS scores of the sure can be detected among them. Because of the historical reference sample fall below a score of 11 or 12, whereas development of the FABS, there are different numbers of about 80% of the FASFAE patients in the FASFAE items in the FABS for the 1994 Detection Study (Fig. 5 ) and reference sample have a FABS score above 11 or 12. Only about one-third as many items are endorsed in the Normathe 1995 Normative Study (Fig. 6). tive Study, compared with those in the Derivation Study. As Fig. 6 shows, having a father with alcohol problems did not Using the FABS to Detect Men with Alcoholic Mothers affect FABS scores in the Normative Study. However, havfrom Among Prison Inmates ing a mother with alcohol problems raised the FABS scores The boxplots in Fig. 5 show fairly impressive separation to a range substantially overlapping that of the Derivation of the FABS scores on the reference sample of 472 patients Study patients with FAS/FAE. with FASFAE from the FABS scores of the inmates in the Detection Study. We find that 85% of the inmates have FABS scores below 6 or 7, whereas 85% of the FASFAE Demonstrating the Reliability of the FABS Score Over Time patients have FABS scores above 6 or 7 (on the 1994 For the 37 patients in the Test-Retest Study, FABS 26-item FABS). Three of the four highest FABS scores as scores were correlated at 0.69 over an average duration of filled out by the corrections officers in the Detection Study 5.0 years (range: 1.5 to 9.4). The mean discrepancy between zyxwvut 330 zyxwvutsrqpon zyxwvutsrqp STREISSGUTH ET AL. 100% .* .* . . . . 75% 4- zyxwvutsrqponmlkjih . S a, 2 a, a .-$ 3 zyxwvut zyxwvuts . 50% i 3 0 25% .* .*= .. . .. . 0% 12 0 Fig. 4. Cumulative frequency distribution of FABS scores for the FAS/FAE reference sample (n = 472). Determination of the median score'' is indicated by the dashed lines from this plot. . . .. zyxwvuts 21 24 36 (median) The 1996 36-item FABS (data from FAS/FAE reference sample, N=472) zyxw than 9 points higher (on the 1996 36-item FABS). The correlation of 0.69 is a more appropriate estimate of reliability than the Cronbach a, because it incorporates instability over time in the implicit criteria being used by the caretaker and in the patient's own behavioral lability. Examination of a scatterplot showing the change in FABS score by duration of testhetest interval revealed no systematic trends over time. Using the FABS Scores to Predict Dependent Living in Adults with FASIFAE FAS/FAE reference Detection Study Sample Patients with high FABS scores are much more likely to sample still be living dependently as adults (Fig. 8). The median Fig. 5. Boxplots comparing the FAS/FAE FABS scores from the FAS/FAE score on the 1996 36-item FABS for the 60 adults with reference sample (n = 472) with the FABS scores of prlson inmates in the Detection Study (n = 81). Additional boxplots break out the inmate sample FAS/FAE living dependently is 21.5, compared with 8.5 for according to inmate's report of whether or not the inmate thought his biological those not living dependently (n = 10). The mean age of the mother had an alcohol problem. Individualscores are represented by dots to the Prediction Study sample at the administration of the FABS right of the boxplot, for offspring whose biological mothers had alcohol problems. Note that the boxplot data for the FAS/FAE reference sample are based on the was 24.8 years (range: 12.6 to 43.0 years). The mean age at same 26 items used with the inmates for the Detection Study. the LHI administration was 28.1 years (range: 21 to 50 years). The mean duration between FABS and LHI was 3.3 FABS 1 and FABS 2 was only about 1.5 points. As Fig. 7 years (range: 1.1 year to 12.3 years). The mean duration shows, only three patients had scores more than 9 points between the FABS and the LHI was not different for lower on retest, whereas only two patients had FABS more patients who were and were not classified as living depenn= 472 81 Total 13 44 24 Yes no unkn Biologic Mother Had Alcohol Problems zyxwvu zyxwvutsrqpo zyxwvutsrq zyxwvutsrq zyxw FETAL ALCOHOL BEHAVIOR SCALE 331 zyxwvutsrq 35 (I) m 2 E 30 25 Fig. 6. Boxplots comparing the FABS scores of the FAS/FAE reference sample (n = 472) with the FABS scores on children in the Normative Study (fl = 186). Additional boxplots show the Normative Study sample broken out according to respondent’s report of alcohol problems for the target child’s biological parents. The individual scores for the two smallest groups (“mother” and “no info” are shown with dots. Note that the boxplot data for the FAS/FAE reference sample are based on the same 35 items used with the parents for the Normative Study. The 35-item FABS differs from the 36-item FABS because of a clerical omission. .-2 20 d 0 m 15 In m ;10 c I- 5 0 n= 472 186 Total FASIFAE reference sample 13 mother 27 138 8 father neither no only parent info Biologic Parents Had Alcohol Problems Normative Study Sample zyxwvutsrqpon mean FABS at older age = 20.0 k 8.6 mean FABS at younger age = 21.5 f 9.3 correlation r = 0.69 Fig. 7. FABS scores at two different ages for the 37 patients who had PBCs filled out by the same respondent after at least a 1-year interval. These data are from the 1996 36-item FABS. unkn, Unknown. /* 9 18 36 27 The 1996 36-item FABS at younger age (Test 1) dently as adults. Dependent living as an adult is one of the severe long-term consequences of FAS.31,32 DISCUSSION The FABS emerging from this series of studies is a 36-item scale in a yes/no format. The FABS score is a simple count of yes responses. It is not self-administered, but rather is filled out by a person or caretaker who knows the patient well or by the consensus of a group who collectively know the patient’s behavior well, each from different standpoints. The FABS score has adequate test-retest reliability and is uncorrelated with age, sex, race, IQ, and alcohol-related diagnosis, so has maximum usefulness across various groups from age 2 up through age 35. Because FAS was only identified as such in 1973, there are not presently enough geriatric patients available for study. The behavioral phenotype reflected in the FABS score is 332 zy zyxwvutsrqponml STREISSGUTH ET AL. FABS in a research context. Further study is needed to clarify its utility in a clinical or screening context. Instruments like the FABS should not be used clinically (i.e., for diagnostic purposes) without additional evidence of prenatal alcohol exposure. Additional studies are needed to determine the specificity of this behavioral phenotype to alcohol teratogenesis, to evaluate contemporaneous interrater reliability, and to ascertain the conditions under which it would be a useful tool. Studies comparing FABS scores of individuals with FAS/FAE compared with patient groups with other developmental disabilities (e.g., Down’s syndrome), to those with high behavioral problems scores on the CBCL, and to more children of alcoholic fathers versus alcoholic mothers will further elucidate the extent to which the FABS succeeds in capturing the specific behavioral essence of FAE and alcohol-related neurodevelopmental disabilities. zyxwvutsrq zyxwvutsrqpon n= 10 no 60 yes Dependent Living Boxplots comparing FABS scores for adults with FAS/FAE who are living dependently (n = 60) or not living dependently (n = 10).These data are from the 1996 36-item FABS. Fig. 8. not characteristic of the normative sample as rated by their parents nor of the group of prison inmates as rated by the corrections staff. In both groups, those individuals thought to have had a biological mother with alcohol problems had scores more often in the range of FAS/FAE patients. FABS scores appear to be correlated with maternal alcohol problems, but not with paternal alcohol problems (Fig. 6). Thus, the FABS appears to reflect the behavioral phenotype of fetal alcohol fairly specifically rather than the behavioral consequences of being raised in an alcoholic family. This does not appear to be a “children of alcoholics” finding. In the FASiFAE reference sample (Table S), additional psychometric data were available for approximately 400 subjects. The FABS is almost perfectly uncorrelated with IQ (Verbal IQ X FABS, r = -0.03; Performance IQ X FABS, r = -0.05), as well as uncorrelated with age, sex, and race (Fig. 3 ) . The FABS correlates negligibly with achievement scores (Wide Range Achievement Test-Revised; Reading: X FABS, r = -0.05; Spelling: r = -0.15; and Arithmetic: r = -0.16). The FABS correlates moderately (r = -0.36) with the Vineland Adaptive Behavior Composite (VABS) and at magnitudes near 0.6 with certain subscales of the Child Behavior Checklist (CBCL), namely, Social (r = -0.61), Attention (Y = -0.63), and Total Problems (r = -0.63). These latter findings are not surprising, because previous studies have shown that patients with FAS/FAE have elevated scores on the CBCL and depressed scores on the VABS:,32,33 scores reflecting the general level of behavioral problems and adaptive living problems characteristic of individuals with FAS/FAE. One advantage of the FABS is its brief administration time. All of the respondents in these five studies filled out the ACKNOWLEDGMENTS The authors thank Sterling K. Clarren, M.D., John M. Graham, M.D., James H. Hanson, M.D., David F. Smith, M.D., Kenneth Lyons Jones, M.D., and the late David W. Smith, M.D., for diagnostic assessments of patients; Robert Jones, Ph.D., for facilitating the Detection Study; and Alan Ellsworth, Pharm.D, for facilitating the Normative Study. The technical assistance of Karen Kopera-Frye, Julia Kogan, and Kaylin Anderson, and the assistance of John Anzinger and Kristi Cove11 in manuscript preparation are gratefully acknowledged. zyxw zyxwv z zyxwvutsr zyxwvutsr REFERENCES 1. Clarren SK, Smith DW: The fetal alcohol syndrome. N Engl J Med 298:1063-1067, 1978 2. Steinhausen HC, Nestler V, Spohr H L Development and psychopathology of children with the fetal alcohol syndrome. J Dev Behav Pediatr 3:49-54, 1982 3. May PA, Hymbaugh KJ, Aase JM, Samet JM: Epidemiology of fetal alcohol syndrome among American Indians of the Southwest. SOC Biol 30:374-387, 1983 4. Aronson M, Oleglrd R: Children of alcoholic mothers. Pediatrician 1457-61, 1987 5. Streissguth AP, Aase JM, Clarren SK, Randels SP, LaDue RA, Smith DF: Fetal alcohol syndrome in adolescents and adults. JAMA 26511961-1967, 1991 6. Lemoine P, Lemoine PH: Avenir des enfants de mbres alcooliques (Etude de 105 cas retrouvts ii I’iige adulte) et quelques constatations d’inttrCt prophylactique. Ann Pediatr (Paris) 39:226-235, 1992 7. Spohr HL, Willms J, Steinhausen H-C: Prenatal alcohol exposure and long-term developmental consequences. Lancet 341:907-910, 1993 8. Streissguth AP, Clarren SK, Jones KL: Natural history of the fetal alcohol syndrome: A ten-year follow-up of eleven patients. Lancet 2:8591, 1985 9. Institute of Medicine (U.S.) Division of Biobehavioral Sciences and Mental Disorders Committee to Study Fetal Alcohol Syndrome: In Stratton K, Howe C, Battaglia F, (eds): Fetal Alcohol Syndrome: Diagnosis, Epidemiology, Prevention, and Treatment. Washington, D.C., National Academy Press, 1996 10. Jones KL, Smith DW, Ulleland CN, Streissguth Ap: Pattern of malformation in offspring of chronic alcoholic mothers. Lancet 815:12671271, 1973 11. Jones KL, Smith D W Recognition of the fetal alcohol syndrome in early infancy. Lancet 836:999-1001, 1973 12. Rosett H L A clinical perspective of the fetal alcohol syndrome. zyxw zyxwvutsrq zyxwvuts zyxwv zyxw zyxwvu z FETAL ALCOHOL BEHAVIOR SCALE Alcohol Clin Exp Res 4:119-122, 1980 13. Jones KL: Fetal alcohol syndrome. Pediatr Rev 8:122-126, 1986 14. Sokol RJ, Clarren S K Guidelines for use of terminology describing the impact of prenatal alcohol on the offspring. Alcohol Clin Exp Res 4:597-598, 1989 15. Aase JM: Clinical recognition of FAS: Difficulties of detection and diagnosis. Alcohol Health Res World 18:5-9, 1994 16. Streissguth AP,Herman CS, Smith DW: Intelligence, behavior, and dysmorphogenesis in the Fetal Alcohol Syndrome: A report on 20 patients. J Pediatr 92:363-367, 1978 17. West JR (ed): Alcohol and Brain Development. New York, Oxford University Press, 1986 18. Riley EP, Vorhees CV: Handbook of Behavioral Teratology. New York Plenum Press, 1986 19. Vorhees CV, Mollnow E: Behavioral teratogenesis: Long-term influences on behavior from early exposure to environmental agents, in Osofsky JD (ed): Handbook of Infant Development, ed 2. New York, Wiley, 1987, pp 913-971 20. Day N L Effects of prenatal alcohol exposure, in Zagon IS, Slotkin TA (eds): Maternal Substance Abuse and the Developing Nervous System. San Diego, Academic Press, 1992, pp 26-43 21. Goodlett CR, West JR: Fetal alcohol effects: Rat model of alcohol exposure during the brain growth spurt, in Zagon IS, Slotkin TA (eds): Maternal Substance Abuse and the Developing Nervous System. San Diego, Academic Press, Inc., 1992, pp 45-75 22. Streissguth AP:Fetal Alcohol Syndrome: A Guide for Families and Communities. Baltimore, Paul H. Brookes Publishing Co., 1997 23. Streissguth AP,Barr HM, Bookstein FL, Sampson PD: The Enduring Effects of Prenatal Alcohol Exposure on Child Development: Birth Through 7 Years: A Partial Least Squares Solution. Ann Arbor, University of Michigan, 1993 24. Carmichael Olson H, Sampson PD, Barr HM, Streissguth AP, Bookstein FL: Prenatal exposure to alcohol and school problems in late childhood: A longitudinal prospective study. Dev Psychopath01 4:341-359, 333 1992 25. Streissguth AP,Sampson PD, Carmichael Olson H, Bookstein FL, Barr HM, Scott M, Feldman J, Mirsky AF: Maternal drinking during pregnancy and attentiodmemory performance in 14-year old children: A longitudinal prospective study. Alcohol Clin Exp Res 18:202-218, 1994 26. Streissguth AP, Barr HM, Sampson PD, Bookstein FL Prenatal alcohol and offspring development: The first fourteen years. Drug Alcohol Depend 36:89-99, 1994 27. Sampson PD, Streissguth AP,Bookstein FL, Little RE, Clarren SK, Dehaene P, Hanson JW, Graham JM: The incidence of fetal alcohol syndrome and prevalence of alcohol-related neurodevelopmental disorder. Teratology 56:317-326,1997 28. Mattson S, Riley E: Neurobehavioral and neuroanatomical effects of heavy prenatal exposure to alcohol, in Streissguth AP,Kanter J (eds): The Challenge of Fetal Alcohol Syndrome: Overcoming Secondary Disabilities. Seattle, University of Washington Press, 1997, pp 3-14 29. Press, S: Parental Report of “Child Behavior and Relationship to Maternal Alcohol Problems. Seattle, University of Washington, 1997 30. Cronbach LJ:The Dependability of Behavioral Measurement. New York, Wiley, 1972 31. Streissguth AP, Barr HM, Kogan J, Bookstein F L Primary and secondary disabilities in fetal alcohol syndrome, in Streissguth AP,Kanter J (eds): The Challenge of Fetal Alcohol Syndrome: Overcoming Secondary Disabilities. Seattle, University of Washington Press, 1997, pp 25-39 32. Streissguth AP,Barr HM, Kogan J, Bookstein F L Understanding the Occurrence of Secondary Disabilities in Clients with Fetal Alcohol Syndrome (FAS) and Fetal Alcohol Effects (FAE). Final Report to the Centers for Disease Control and Prevention (CDC). Seattle, University of Washington, Fetal Alcohol & Drug Unit, Technical Report No, 96-06, August, 1996 33. Steinhausen HC, Willms J, Spohr H L Long-term psychopathological and cognitive outcome of children with fetal alcohol syndrome. J A m Acad Child Adolesc Psychiatry 35:5, 990-994, 1993