Evidence increasingly suggests that mental health problems are associated with insults during critical times of brain development, and also by exposures over an individual's life course. These include in-utero exposures, which may lead to subtle neurobehavioral difficulties. Further, there is a body of evidence relating birth weight to cognitive function in childhood, though sibling-based analyses suggest the association is more likely to be due to fixed familial factors such as background socioeconomic position and behaviours/exposures that are similar for all siblings within a family. (1)

However, there is relatively little work on the role of early alcohol and tobacco exposure on the occurrence of a range of cognitive, behavioural and emotional difficulties. In-utero to these substances may cause a variety of adverse developmental outcomes. (2) Finally there is emerging evidence pointing to an association between alcohol and tobacco use in pregnancy and the development of addictive behaviours, such as addiction to alcohol and tobacco. (3-6)

The mechanisms underlying these associations are unclear. It is assumed that the effects are due to specific intrauterine exposure, but genetic factors and shared familial factors (including socioeconomic position, shared/learnt familiar behaviours) may well explain any link between maternal alcohol and tobacco consumption in pregnancy and later effects on offspring cognitive function and behaviour. One way to further explore this is to compare associations of maternal alcohol and tobacco use during pregnancy with offspring outcomes to the same associations with paternal alcohol and tobacco consumption. Analyses using these analytical techniques are needed to ascertain whether in-utero exposure to alcohol and tobacco consumption truly contributes to later developmental problems. If this is so it is also important to understand the mechanisms underlying these associations. To date only one study has done this and both George Davey Smith and I have been involved in preparing that paper (7).

In the previous paper published in 2008, we carried out comparisons of both alcohol and tobacco use with child's IQ at age 8 and found no important adverse effect of both moderate maternal alcohol and tobacco consumption on offspring IQ at age 8, though IQ scores were lower in children of mothers who reported greater quantity of 'binge'- like style of drinking. (7) Further to that study, a Welcome Trust project grant led by Dr Ron Grey, has allowed us to progress our work in this area and carry out maternal-offspring and paternal-offspring alcohol comparisons with a range of child's neuro-developmental outcomes, including cognitive abilities at age 11 (paper awaiting ALSPAC Exec approval).

The role of intra-uterine exposure to tobacco has not been investigated in this grant. Hence, we propose to examine the association between parental smoking and offspring cognitive abilities at age 11. Several studies have found an inverse association between parental (usually maternal) smoking during pregnancy and offspring IQ. Biological mechanisms involving in utero exposure to tobacco combustion products and their metabolites have been proposed however other evidence suggests that confounding by factors related to social position may be more important. [2,3] Unlike alcohol use, smoking at even relatively moderate levels may be a marker of adverse social environment in contemporary populations. This is likely to be particularly true of maternal smoking during pregnancy.

The aim of this proposal is to use ALSPAC data in order to determine whether intra-uterine exposure to tobacco consumption (assessed by maternal self-reports during pregnancy) is related to poorer academic abilities at age 11 (KS2 linked data). The association of maternal smoking during pregnancy with academic abilities will be compared to similar associations of paternal tobacco use (based on self-report by the partner of the mother, or where this is not available the mother's report of the fathers tobacco use, with restriction to those who define themselves as the biological father). This comparison will strengthen our ability to determine whether there is a true intrauterine effect, since if there is one, it should be only apparent with maternal tobacco use. Similar association between maternal and paternal tobacco use and child's cognitive skills would suggest that genetic factors and/or shared family environmental factors explain the association rather than specific intrauterine effects. We propose to examine the effect of maternal and paternal smoking during pregnancy and offspring KS2 scores at age 11, both before and after adjustment for covariates. I will conduct the analysis using the current Wellcome Trust dataset, and will need additional alcohol and tobacco variables in the postnatal period to include prenatal and postnatal parental effects in the analysis (see analysis used in the alcohol paper awaiting approval).