OBJECTIVE: We hypothesized that a neonatal marker of alcohol effect would improve the prediction of children at risk for school problems.STUDY DESIGN: A historical, prospective design was used to examine predictors of problem school behavior. Maternal-child dyads (N=140) previously participating in a prospective Fetal Alcohol Research Center Study were contacted; 116 agreed to have their six year old singleton term child participate. Absolute alcohol consumption (AAD) and cocaine exposure (COKE) were prospectively recorded during pregnancy. Dysmorphic examination during the neonatal period was performed by an observer blinded to fetal status. The child's first grade teacher, who was blinded to both study design and exposure, rated the child's behavior with the Conners' Teaching Rating Scale(CTRS) and the investigator-developed 14 item 10-point, visual analog scale, the PROBS, measuring behaviors believed by teachers to be specific to cocaine exposure. Previous data suggested that philtrum length (philtrum) was a significant alcohol related effect. In regression models AAD, COKE, and philtrum by AAD interaction were examined as predictors of CTRS and PROBS residualized for gender and years of teaching experience. RESULTS: Teacher data were obtained for 102 and philtrum for 82 children. AAD varied form 0 to 3.2 oz/day ([horizontal bar over]x=.16±.46) while 24 of the 82 subjects admitted to COKE. The CTRS was not predicted by any of the variables. For the PROBS-R, only philtrum was a significant predictor (r=.206, p<0.05, one-tailed); philtrum retained its one-tailed significance even when alcohol was added to the model. CONCLUSION: Long philtrum has been related to FAE in prior studies. This pilot study suggests that long philtrum, and alcohol related effect in prior studies, may serve as a marker for long term behavioral abnormalities. The behavior effects of alcohol exposure must be considered in any study of other suspected teratogens such as cocaine.
