BACKGROUND: Maternal smoking in pregnancy lowers birthweight. It is unclear, however, whether smoking during pregnancy lowers offspring IQ, and, if it does, whether it is through the smoking effect on fetal growth. METHOD: Representative samples of low birthweight (<2500 g) and normal birthweight children born in 1983-85 from inner-city and suburban communities in southeast Michigan, USA were assessed at ages 6, 11, and 17, using Wechsler intelligence tests. Smoking during pregnancy was ascertained from mothers at the first assessment; and smoking at any time was ascertained at the first and second assessment. Generalized estimating equation models were used, with children's IQ at all three assessments as outcomes (n = 798). RESULTS: Without adjustment, offspring of mothers who smoked during pregnancy scored 6.8 IQ points lower than offspring of mothers who never smoked, on average. Low birthweight children scored 5.4 IQ points lower than normal birthweight children, on average. The statistical association of maternal smoking with offspring IQ was confounded by maternal characteristics, chiefly, maternal cognitive ability as measured by IQ and education; adjustment for these factors eliminated the association. By contrast, adjustment for maternal IQ and education as well as smoking during pregnancy had a negligible effect on the low birthweight-related IQ deficit. Low birthweight did not mediate the association of smoking and lowered IQ in offspring. CONCLUSION: Maternal smoking during pregnancy is a proxy for a matrix of vulnerabilities for adverse child cognitive development and has no direct causal effect on child's IQ. The relationship of low birthweight and IQ is independent of maternal smoking and maternal cognitive abilities.
BACKGROUND: Maternal smoking in pregnancy lowers birthweight. It is unclear, however, whether smoking during pregnancy lowers offspring IQ, and, if it does, whether it is through the smoking effect on fetal growth. METHOD: Representative samples of low birthweight (<2500 g) and normal birthweight children born in 1983-85 from inner-city and suburban communities in southeast Michigan, USA were assessed at ages 6, 11, and 17, using Wechsler intelligence tests. Smoking during pregnancy was ascertained from mothers at the first assessment; and smoking at any time was ascertained at the first and second assessment. Generalized estimating equation models were used, with children's IQ at all three assessments as outcomes (n = 798). RESULTS: Without adjustment, offspring of mothers who smoked during pregnancy scored 6.8 IQ points lower than offspring of mothers who never smoked, on average. Low birthweight children scored 5.4 IQ points lower than normal birthweight children, on average. The statistical association of maternal smoking with offspring IQ was confounded by maternal characteristics, chiefly, maternal cognitive ability as measured by IQ and education; adjustment for these factors eliminated the association. By contrast, adjustment for maternal IQ and education as well as smoking during pregnancy had a negligible effect on the low birthweight-related IQ deficit. Low birthweight did not mediate the association of smoking and lowered IQ in offspring. CONCLUSION: Maternal smoking during pregnancy is a proxy for a matrix of vulnerabilities for adverse child cognitive development and has no direct causal effect on child's IQ. The relationship of low birthweight and IQ is independent of maternal smoking and maternal cognitive abilities.