~ Pergamon PU: S0028-3932(96)00043-7 Neuropsycholoyia, Vol. 34, No. 12, pp. 1187-1196, 1996 Copyright ~ 1996 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0028 3932/96 $15.00+0.00 Impairments of number processing induced by prenatal alcohol exposure KAREN KOPERA-FRYE,* STANISLAS DEHAENEt and ANN P. STREISSGUTH* *Fetal Alcohol and Drug Unit, Department of Psychiatry and Behavioral Sciences, University of Washington Medical School, Seattle, WA 98195, U.S.A.; and tLaboratoire de Sciences Cognitives et Psycholinguistique, CNRS and EHESS, Paris, France (Received 4 July 1995; accepted 20 March 1996) Abstract--Prenatal alcohol exposure causes a variety of cognitive deficits, notably in mathematics and higher order processes such as abstraction. An exploratory battery was developed to examine specific types of number processing impairments in 29 adolescent and adult patients with Fetal Alcohol Syndrome (FAS) and Fetal Alcohol Effects (FAE) relative to controls matched for age, gender, and educational level. The battery included 11 tests: number reading and writing, exact calculation (addition, multiplication, subtraction), approximate calculation (selecting a plausible result for an operation), number comparison, proximity judgment, and cognitive estimation. The results indicated particular difficulties in calculation and estimation tests, with intact number reading and writing ability. The greatest impairment was found in the cognitive estimation test, which is sensitive to frontal lobe lesions. The patterns of deficit described may reflect either the diffuseness of brain damage incurred from prenatal alcohol exposure, or a cumulative deficit in comprehension which may be important for the acquisition of higher-order mathematical" abilities. Copyright © 1996 Elsevier Science Ltd Key Words: cognitive deficits; Fetal Alcohol Syndrome; cognitive estimation; number processing; prenatal alcohol exposure. Introduction In the last two decades, it has become clear that alcohol is a teratogen. A teratogen is defined as any agent capable of inducing abnormal or atypical fetal development. The first evidence of the teratogenic effects of alcohol came from the identification of the Fetal Alcohol Syndrome (FAS), the diagnostic term for a birth defect caused by prenatal alcohol exposure. Patients with FAS were first identified in France in 1968 [17] and in the United States in 1973 [16]. Since that time, thousands of patients with FAS have been identified in many countries [6, 7, 18, 20, 31, 32]. FAS is defined by three criteria: (1) growth deficiency of prenatal origin for height or weight; (2) a pattern of specific minor anomalies that includes characteristic facial features, e.g. short palpebral fissures, thin upper lip, smooth and/or long philtrum; and (3) central nervous system dysfunction, including microcephaly, delayed development, hyperactivity, attention deficits, learning disabilities, intellectual deficits, and so forth [6, 29]. Patients are often described as having possible Fetal Alcohol Effects (FAE) when they are heavily exposed to alcohol in utero and have some, but not all, of the hallmark features of FAS [6]. Both clinical case studies and large epidemiologic stud- ies have found alcohol-affected individuals to be impaired in standardized arithmetic achievement tests and IQ tests [7, 26, 32, 33]. While deficits with visual sequencing, visual spatial reasoning, attention, and memory have also been described [5, 11, 19, 24, 35], some of the most striking deficits are evident on number processing tasks. The exact nature of these deficits, however, has not been adequately studied. The purpose of the present study is to determine which components of number processing are impaired in patients with FAS and FAE by utilizing more specific tests and theories derived from cognitive neuro- psychology. Recently, there have been several careful studies of brain-lesioned patients with selective impairments in the numerical domain [10, 22, 25, 37]. This has led to the development of detailed models of the cognitive archi- tecture of number processing [3, 8, 21]. It is beyond the scope of this paper to describe these models in detail. However, two general principles of number processing were particularly relevant in guiding the present study. The first principle is that number-related processes appear to be largely modular. Single-case studies indicate that a focal brain lesion can have a highly selective effect on number processing. For example, knowledge of arith- 1187