Weight growth in infants born to mothers who smoked during pregnancy

Introduction

A causal relation between smoking in pregnancy and birth weight is generally accepted. Mothers who smoke during pregnancy generally deliver infants weighing from 100 g to 300 g less than children born to non-smoking mothers.12 A dose-effect relation between number of cigarettes smoked by a pregnant woman and her infant's weight has been shown, whereas there are no differences in weight between babies born to women who never smoked during their lives and those who quit smoking only during pregnancy.1 3 4 5 Limited data on the follow up of children born to mothers who smoke suggest that the deficits at birth may remain in childhood and possibly even into adulthood.3 6 7 Almost all these data, however, were collected on small samples in the early 1970s in the United States and the United Kingdom, where smoking habit was often associated with poor social class and lower educational level. Furthermore this long term effect has not been found by other authors.8 9 The issue of maternal smoking during pregnancy and child development thus remains controversial.

We evaluated weight gain in the first six months of life of babies born to mothers who smoked during pregnancy.

Subjects and methods

Results

Table I shows the relation between socioeconomic index and mother's smoking habit during pregnancy. Smoking habit was more common in the upper classes, whereas the percentage of heavy smokers was independent of the socioeconomic class. Most subjects belonged to intermediate classes (groups 3 and 4). Table II shows the comparison at birth and at 3 and 6 months of age of the mean weight in each group of children born to non-smoking mothers and children born to mothers smoking up to nine or >/=10 cigarettes a day. No differences were seen between mean and median weights, and thus mean weights only are reported. The mean birth weight of children of non-smoking mothers was 3220 g for girls and 3373 g for boys. In children born to light smokers these weights were reduced by 88 g and 107 g, respectively; the reduction was higher in children born to heavy smokers (168 g for girls and 247 g for boys). As shown in table II, however, differences in weight observed at birth had decreased by 3 or 6 months of age. As a result of the weight gain the mean weight difference in the group of children born to light smokers, which at birth was -2.7% for girls and -3.2% for boys, became +0.1% and -0.8% at six months for girls and boys, respectively. Likewise in the group of children born to heavy smokers the difference was reduced from -5% in girls and -7.3% in boys at birth to -0.4% and -0.3% at 6 months.

To investigate confounding effects on weight growth in the first six months of life we considered for each sex the following variables: age of the mother, parity, and gestational age. Sixteen subgroups were identified accordingly. In this data set social class was not associated with a difference in weight between 0 and 6 months in children, as already reported,11 and thus it was not considered as a confounding variable. Table III shows for each subgroup the number and mean weights at 0, 3, and 6 months of children born to non-smoking mothers and weight differences with confidence intervals in children with mothers smoking up to 9 or >/=10 cigarettes a day. In almost all the subgroups the deficits of weight at birth in children of smoking mothers were overcome. Only in the subgroups of boys with gestational age <37 weeks (very few subjects) was weight recovery not evident.

The influence of mothers' smoking habit on children's growth was investigated with a multivariate model of the variance. The final model considered weight as the dependent variable and mother's smoking habit, parity, and gestational age as independent variables. Analysis was done in three steps to give the following results. Firstly, mean weight at birth in children with smoking mothers or non-smoking mothers was significantly different (P<0.0001 by Wilk's (lambda)). Secondly, mean weights at 6 months between children with smoking mothers or non-smoking mothers were not significantly different. Thirdly, the difference in speed of growth from 0 to 6 months between children with smoking mothers and non-smoking mothers was significant (P<0.0001 by Wilk's (lambda))–that is, babies born to smoking mothers have a greater rate of growth in the first six months. These findings were the same in both sexes. These results were confirmed in each step by the contrasts between smoking and non-smoking mothers and between non-smoking and heavy smoking mothers. Table IV gives the estimated differences (95% confidence intervals) of mean weights of children with non-smoking mothers in respect of children with smoking mothers and of children with mothers smoking >/=10 cigarettes a day (subdivided by sex and age).

Discussion

Data reported by Sexton and Hebel2 and Fox et al6 suggest that deficits of weight at birth in children born to mothers who smoke during pregnancy may remain throughout life. The number of children entered in their studies, however, was relatively small and the comparison was actually between smoking mothers who quit (188 subjects) or did not quit smoking (526 subjects) during pregnancy. Children from smoking mothers had a weight reduction of 92 g at birh and of 450 g at age three years, suggesting that the deficits associated with maternal smoking may not be overcome by 3 years of age.6 Difference in weight at 3 years, however, was significant only when other variables such as birth weight, household variables, or length of gestation, had not been adjusted for, and thus weight at 3 years could be more related to these variables than to smoking in pregnancy.2 In other studies in Finland and England children of smokers were found to be shorter and with poorer academic ability.3 4 In these studies smoking habit was associated with poorer socioeconomic factors.

In contrast with these results Barr et al found that in a small sample of primarily white, married, and well educated mothers in the United States maternal smoking during pregnancy was not significantly related to infant size (weight, height, and head circumference) at 8 months, although nicotine use was highly related to the birth size.8

The social context of our study was quite different from those mentioned. In the 1970s in Italy the highest prevalence of smoking was in women with high school diploma or university degree.12 14 This association was observed in our sample too (table I). In keeping with published data1 2 in our sample birth weights of girls and boys born to light or heavy smokers were, respectively, 88 g, 107 g, 168 g, and 247 g lower than those of infants born to non-smoking mothers.

The deficit of weight at six months of age was reduced to + 9 g for girls and -64 g for boys in children born to light smokers and to -28 g and -24 g in children born to heavy smokers. These data indicate that the deficits of weight at birth in children born to smoking mothers are overcome by 6 months of age. This trend was consistent in almost all subgroups evaluated.

The same findings have been obtained by multiple analysis of variance model. The analysis with parity and gestation considered as covariates showed that birth weight is related to mother's smoking habit; rate of growth from 0 to 6 months is higher in babies born to smoking mothers; weight differences between children born to smoking or non-smoking mothers are no longer measurable at 6 months. Weight recovery was faster in girls. Mother's age and social class were not associated with the weight of the baby.

Our data suggest that the well known effects of smoking during pregnancy on birth weight are not permanent and that smoking in pregnancy does not affect the growth in childhood when smoking habit during pregnancy is not associated with other unfavourable variables. Future investigations on this issue should evaluate carefully the role of other factors, such as exposure to environmental smoke after birth15 and socioeconomic class.