0145-6008/96/2004-0763$03.00/0 ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH Vol. 20, No. zy 4 zy June 1996 zy Prenatal Alcohol Exposure and Academic Achievement at Age Six: A Nonlinear Fit Lidush Goldschmidt, Gale A. Richardson, David S. Stoffer, Diklah Geva, and Nancy L. Day This is a report on the effects of prenatal alcohol exposure on the academic achievement of children at 6 years of age. In this longitu- dinal study, women were interviewedat the end of each trimester of pregnancy, at delivery, and at 8,18,36, and 72 months postpartum. The women were of lower socioeconomic status, high school-edu- cated, and moderate users of alcohol. The offspring received age- appropriate physical and developmental assessments at each fol- low-up. Linear regression and nonlinear curve fitting were used to investigate the nature and shape of the relationship between prena- tal alcohol exposure and achievement. In addition, the role of child IQ in this relationship was explored. Alcohol exposure during the sec- ond trimester predicted deficits in each of the three subtests of the Wide RangeAchievement Test-Revised(WRAT-R): reading, spelling, and arithmetic. The relationship was partially reduced by the addition of IQ to the model, but prenatal alcohol exposure still predicted sig- nificant deficits in achievement, even after controlling for IQ. Tests for the shape of the relationship demonstrated that the effect of prenatal exposure on the arithmetic subtest of the WRAT-R was a linear or doswesponse relationship. By contrast, the relationships between prenatal alcohol exposure and performanceon the spelling and readingsubtests of the WRAT-R were better modeledas thresh- old effects. The thresholds for both were -1 drinwday in the second trimester. Key Words: Alcohol, Pregnancy, Wide Range Achievement Test- Revised, IQ, Threshold Effects. 0 DATE, there have been few studies of the refation- T ship between prenatal alcohol exposure and the subse- quent academic achievement of the exposed offspring. This is an important issue, however, because school perfor- mance is a marker of behavioral and cognitive problems, as well as a predictor of future achievement. Two studies have evaluated the relationship between prenatal alcohol exposure and school achievement. Streiss- guth et a1.l reported that drinking five or more drinks per occasion in the month before pregnancy recognition pre- zyxwvu From the Western Psychiatric Institute and Clinic (L.G.), University zyxwvutsr of Pittsburgh Medical Center; Departments of Psychiatry ( G A R . , N.L.D.), Epidemiology (G.A.R.,N.L.D.), and Pediatrics (N.L.D.),University of Pitts- burgh School of Medicine; and Department of Mathematics and Statistics (D.S.S. ), University of Pittsburgh, Pittsburgh, Pennsylvania; and the Edith Wolfson Medical Center Holon (D.G.) and Tel Aviv University Medical School (D.G.), Tel Aviv, Israel. Received for publication May 9, 1995; accepted January 29, 1996 This study was supported by the National Institute on Alcohol Abuse and Alcoholism (Grants AA06390 and zyxwvutsrqpo AA06666) and by the National Institute on Drug Abuse (Grants DA03209 and DA03874). This study was conducted at the University of Pittsburgh School of Medicine, Western Psychiatric Insti- tute and Clinic, and the Magee- Womens Hospital. Reprint requests: Nancy L. Day, Ph.D., Western Psychiatric Institute and Clinic, 3811 O'Hara Street, Pittsburgh, PA 15213-2593. Copyright zyxwvutsrqp 0 1996 by The Research Society on Akoholism. Alcohol Clin Exp Res, Vol20, No 4, 1996: pp 763-770 dicted a deficit of 3 points on both the arithmetic and reading scores of the Wide Range Achievement Test-Re- vised (WRAT-R) at 7.5 years. Coles et a1.' compared 5- to 8-year-old offspring of three groups of women: (1) women who drank throughout pregnancy, (2) women who stopped drinking midpregnancy, and (3) women who did not drink during pregnancy. They found significant deficits in Kauf- man Assessment Battery for Children (K-ABC) premath and reading skills among the alcohol-exposed children (groups 1 and 2), compared with the children who were not exposed to alcohol prenatally (group 3). Cognitive deficits are part of fetal alcohol syndrome (FAS). Decreased cognitive functioning has been noted as a fetal alcohol effect, although reports on the relationship between prenatal alcohol exposure and cognitive develop- ment have been mixed. Streissguth et aL3 reported a sig- nificant negative relationship between prenatal alcohol ex- posure and the Wechsler Preschool and Primary Scales of Intelligence (WPPSI) IQ of 4-year-olds in a predominantly Caucasian, middle-class sample. At age 7.5 years, in this same sample, Streissguth et al.,' using the Wechsler Intel- ligence Scale for Children-Revised found that a 7-point decrease in IQ was associated with exposure at the rate of 2 or more drinkslday in midpregnancy. Coles et aL2 found significant deficits in K-ABC sequential processing and mental composite scores among children in group 1 (ex- posed throughout pregnancy), compared with the children in the other two groups. Russell et aL4 found lower WPPSI Verbal IQ and Token Test scores among 6-year-olds of mothers who had two or more indications of problem drinking during pregnancy. Other studies have not found an effect of prenatal alco- hol exposure on cognitive outcome. Greene et al.,5 in a lower socioeconomic status sample, did not find an adverse effect of prenatal alcohol exposure on the WPPSI Full- Scale IQ scores of children assessed at age 4 years 10 months. Fried et a1.,6 in a Caucasian, middle-class sample, did not find significant relationships between low to mod- erate levels of alcohol exposure and cognitive development or receptive language at either 5 or 6 years of age using the McCarthy Scales of Children's Abilities and the Peabody Picture Vocabulary Test. The reports of the effects of prenatal alcohol exposure on decreased academic performance were from the same studies that reported a relationship between alcohol use during pregnancy and decreased IQ. The relationship be- z 763