There is an abundance of observational evidence that smoking during pregnancy is harmful to the unborn child.
Previous studies have seen an association between prenatal smoking and a range of behavioural and psychiatric problems in childhood and beyond. A new cohort study used magnetic resonance imaging (MRI) to study the brains of children whose mothers smoked during pregnancy.
The study participants were 6 to 8 years old and drawn from an ongoing cohort study in the Netherlands. 113 children who were exposed to tobacco smoke during pregnancy were matched with 113 similar children who were not exposed.
The researchers ascertained tobacco exposure with reference to questionnaire data that had been obtained during pregnancy. Three categories were established: “no smoking during pregnancy”, “until pregnancy was known” and “continued during pregnancy”
MRI scans were used to ascertain thickness of the cerebral cortex and the volume of a range of different brain structures. Behavioural problems were assessed using a validated questionnaire completed by the mother. Secondary measures included maternal age, educational level, ethnicity and drinking; child age, IQ, gestational age at birth and birth weight.
Adjustments were made for educational level, ethnicity, alcohol use and birth weight.
Continued maternal smoking was associated with smaller brain volume and thinner cortex in several brain areas.
- It was also associated with behavioural problems. Interestingly, behavioural problems were not associated with reduced brain volume but were associated with thinner frontal cortex.
- Children of women who quit smoking early in pregnancy did not show significant differences in brain development.
- Some caution is warranted in interpreting this research. There were important differences between the control and exposure groups. The rates of alcohol use were significantly higher in the mothers who smoked throughout pregnancy (although the association was still significant after alcohol was taken into account).
- Little actual data is shown comparing brain outcomes with controls. Mostly, we only have the p values for the comparisons. What this actually means in clinical terms is not clear. We also don’t know the relationship between the brain structures studied and clinical outcomes.
- The findings may also be confounded by the effects of passive smoking or other environmental factors associated with smoking, such as poor diet, low socioeconomic status and education. The study did not provide enough information about the process of matching with controls to be able to assess whether these possible confounders had been accounted for in the matching procedure.
- The authors cite similar research that identified differences in other brain structures at later stages in development. It is interesting that these were not consistent with the structures identified in their studies. Although they postulate possible mechanisms for these differences, it does suggest the possibility that they may be artifacts of imaging techniques. Further studies are warranted.
- Data from 31 subjects were not excluded from the analysis due to poor quality. The researchers state that this did not affect the results.
- On the other hand, this was a prospective study in which a relatively large number of children underwent neuroimaging.
- The researchers who interpreted the MRI scans were blinded to whether or not the subject had been exposed to tobacco.
- There was consistency of an affect across the different brain measures, suggesting that there is a real effect.
El Marroun H, Schmidt MN, Franken IHA, Jaddoe VWV, Hofman A, van der Lugt A, Verhulst FC, Tiemeier H, White T. Prenatal tobacco exposure and brain morphology: a prospective study in young children. Neuropsychopharmacology 2013. [Journal abstract]