Mothers who drink alcohol during pregnancy are more likely to have badly behaved children, newspapers reported today.
The Daily Telegraph said that a study had found that “conduct problems increased with every day that expectant mothers had a drink”.
The Daily Mail said that "U.S. academics claim the research shows alcohol's effect on the unborn baby has consequences for the child's behaviour several years later, even after genetic and parenting factors are taken into account."
Analysis of this study reveals that it did not take into account all of the potential factors that could increase the risk of behavioural problems, and the researchers themselves say that their study ‘cannot prove causality’.
This study cannot prove that alcohol exposure during pregnancy causes conduct problems in children. However, women are advised to avoid alcohol in pregnancy anyway; this advice is based on strong evidence that heavy exposure can lead to foetal alcohol syndrome (growth and nervous system abnormalities that can occur in the baby).
Where did the story come from?
Dr D’Onofrio and colleagues from various academic institutions across the USA, including Indiana University and the University of Chicago, conducted this research. The study was published in the (peer-reviewed) medical journal: Archives of General Psychiatry.
What kind of scientific study was this?
This study used information from two previous studies; the National Longitudinal Survey of Youth (NLSY), and its follow up, the Children of the National Longitudinal Survey of Youth (CNLSY).
The NLSY study was carried out between 1979 and 2004 and investigated adolescents and young adults who were aged 14 to 22 years when they were first surveyed in 1979. It looked at several areas, including health status, education, alcohol and drug use during pregnancy, and delinquent behaviour.
The CNLSY was a separate study of all the children who were born to women enrolled in the NLSY. The CNLSY assessed children every two years between the ages of 4 and 11 years and collected information on health, background, education and behaviour.
Using these two studies together meant that the researchers had information from 4912 women who had a child who was at least 4 years old by the end of the study, and detailed assessments of their 8621 children. In particular, they used the mothers’ ratings of their child’s behaviour over this period. This was measured on a questionnaire called the ‘behaviour problem index’ which was used to determine whether the children were suffering from attention/impulsivity problems (which would include ADHD) or conduct problems (repetitive behaviour which can include aggression, destruction of property, deceitfulness).
Using the information from the two studies, the researchers determined whether there were any behavioural differences between the children of women who drank alcohol during pregnancy, and those who did not.
The researchers adjusted for certain factors which they thought might influence the relationship between alcohol use during pregnancy and the child’s behaviour. This included the mothers’ delinquent behaviour when they were aged 15 to 22 years, maternal age at the birth of her first child, maternal income, level of education (number of years in school), and mothers’ intellectual ability.
For a smaller sample of the children (3,977), there was information available on whether their mothers were also exposed to drugs such as marijuana and cocaine. The researchers analysed this sample to see how exposure to drugs might be affecting the behavioural outcomes in the children.
In the final part of the study, the researchers looked only at women who had more than one pregnancy. They compared the behaviour of the offspring where the mother had said she took alcohol during that pregnancy with the behaviour of the offspring from another pregnancy where reported alcohol intake was different.
What were the results of the study?
The results from the ‘behaviour problem index’ questionnaire returned different results for 'conduct problems' and 'attention/impulsivity problems' with alcohol during pregnancy.
The researchers found that when they took into account the factors mentioned above (i.e. mothers’ delinquent behaviour, age at birth of first child, income, level of education, and intellectual ability), prenatal alcohol exposure increased the risk that the child would have conduct problems.
However, they found no evidence that alcohol during pregnancy increased the risk of attention/impulsivity problems.
When they compared different pregnancies from the same mother, they still found that alcohol exposure increased the risk of conduct problems.
What interpretations did the researchers draw from these results?
The researchers conclude that there is “an environmentally mediated causal effect” between prenatal alcohol exposure and childhood conduct problems, but that the relationship between alcohol exposure and attention/impulsivity problems is more likely to be caused by other factors which are related to maternal drinking during pregnancy.
What does the NHS Knowledge Service make of this study?
This is a large and well-designed study; however, several important features must be considered when we interpret the results:
- As time went on, and the general public became more aware of the negative effects of alcohol use during pregnancy, women may have been less likely to report their true alcohol consumption in subsequent pregnancies. This may have biased the results.
- Only 3,977 offspring had information available on their mothers’ exposure to drugs such as marijuana and cocaine; this information was only collected in the NLSY from 1986 onwards. When the researchers analysed this sample, they found that there was no statistically significant association between alcohol intake during pregnancy and conduct problems. The researchers acknowledge this, but are not concerned by the lack of statistical significance, saying that the ‘magnitude of the estimate was [similar in size and direction] with the results of the full sample’. We would argue that this non-significant result, which essentially suggests that when considering drug use, alcohol itself may not have any effect on behavioural outcomes, could be important and should be explored further using larger sample sizes.
- The researchers are careful to use language that does not imply that their study can show that alcohol exposure causes behavioural problems. In fact, they say that their study ‘cannot prove causality’.
- The researchers also highlight that their study did not account for all of the potential factors that could be increasing the risk of behavioural problems. This weakness is another reason why the study cannot prove causality. Importantly, the study did not take into account parenting skills, or characteristics of the father which may be linked to behaviour in children.
- Information about alcohol exposure during pregnancy was collected through a questionnaire that was given to the mother up to two years after the birth. Some women will have been unable to remember exactly how much they were drinking up to two years ago. Although the researchers say that there ‘is evidence that retrospective reports of prenatal substance use are reliable’, it’s unclear how much this will have influenced the results.
- The study also relies on the mothers’ reports of their children’s behaviour. Further information from teachers, or children’s self report may have added to the quality of these data.
This is an interesting study and the findings will no doubt add to the discussion about alcohol use during pregnancy. The Department of Health advises women to avoid alcohol during pregnancy. If women choose to drink, they should not drink more than one or two units of alcohol once or twice a week and should not get drunk. This advice is based on evidence that excessive alcohol intake can damage the foetus and can lead to Foetal Alcohol Syndrome.
Sir Muir Gray adds...
There is already good enough evidence to advise pregnant women to avoid alcohol. This study would strengthen that evidence base if the findings were that the relationship is one of cause and effect and not one of coincidence. We need a systematic review of all the research on this topic.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Daily Telegraph, 6 November 2007
Daily Mail, 6 November 2007
Links to the science
Arch Gen Psychiatry 2007; 64: 1296-1304.