Pregnant drinking and child behaviour

Tuesday March 22 2011

“Mothers who drink in early pregnancy are ‘more likely to have unruly children’,” reported the Daily Mail. The newspaper says that a study has found a three-fold risk of antisocial behaviour among 16 year olds whose mothers drank as little as one alcoholic drink per day during early pregnancy.

The US study assessed the possible association between drinking during the first trimester of pregnancy and the risk of a psychiatric condition known as ‘conduct disorder’ in adolescents up to 16 years of age. The disorder can lead to a persistent, marked pattern of repetitive antisocial behaviour that is beyond simply being unruly.

Although the study found an association between conduct disorder and maternal pregnancy, it should be remembered that it is a relatively uncommon condition, and that only 67 adolescents (about 12% of the study population) had experienced it. Therefore further research is needed to reliably assess the influence of prenatal alcohol exposure on the risk of developing the condition.

Current advice is that women trying to conceive and pregnant women, particularly those in the first three months of pregnancy, should refrain from consuming alcohol altogether.

Where did the story come from?

The study was carried out by researchers from The University of Pittsburgh. It was funded by grants from the US National Institute of Alcohol and Alcoholism and the US National Institute of Drug Abuse.

The study was published in the peer-reviewed Journal of the American Academy of Child Adolescent Psychiatry.

This research was covered by the Daily Mail, which reported that alcohol consumption during pregnancy was associated with “unruly behaviour”. It should be emphasised that Conduct Disorder is a specific psychiatric condition diagnosed by a persistent, marked pattern of repetitive antisocial behaviour. It is not clear from this study how alcohol consumption during pregnancy affects minor or short-term symptoms of unruly behaviour.

What kind of research was this?

This was a prospective cohort study, which investigated whether a mother’s alcohol consumption during pregnancy was associated with an increased risk of their child having conduct disorder.

What did the research involve?

The researchers used data from two longitudinal studies, which had looked at the effects of substance exposure during pregnancy. One had focused on alcohol consumption and one had focused on marijuana use. But as their study designs were identical, the researchers combined the data. In total, these studies provided data on 829 women who had been recruited from antenatal clinics. The study began in 1982.

The researchers recorded data on the amount and frequency of alcohol consumption over the three pregnancy trimesters. The researchers also collected data on drug and tobacco use.

From the original cohort, the researchers were able to gather data on 763 live singleton births (some of the mothers moved from the area or did not participate in follow-up). The children were followed from birth for 22 years. At the age of 16 years, 572 of them completed a psychiatric interview to assess psychiatric disorders, both current and during their lifetime. The mothers and adolescents were interviewed separately about their own symptoms. The researchers focused on whether the adolescents had conduct disorder, a psychiatric condition that may cause people to be repeatedly aggressive or destructive and to behave outside of social norms.

Additionally, the children and mothers were assessed at birth, and at the ages of 8 months, 18 months, and 3, 6, 10 and 14 years. During these visits, various aspects of  the children’s home life was assessed, such as whether their biological father or another male adult was involved in their life; how strict they thought their parenting was, whether they regularly ate meals with their family, participated in family activities and performed chores. They were also asked about their participation in sports, their interests and their hobbies.

The researchers also recorded whether the children had experienced a number of specific positive and negative life events, as well as data on the families’ socioeconomic status, the mother’s marital status, the child’s IQ and education.

The researchers restricted their analysis to the reported volume of alcohol drunk during the first three and last three months of pregnancy.

What were the basic results?

The study started in 1982. Of the women enrolled, 73% had completed high school and 23% were working or going to school. At birth, the average age of mothers was 23 years; 55% were African American and 68% were single.

The average alcohol use was 0.6 drinks a day (this ranged from 0 to 20). The average marijuana use was 0.4 joints a day (range 0 to 9), and the average number of cigarettes smoked were 8 per day (range 0 to 50). Eight per cent of the women reported illicit drug use other than marijuana, and 3% reported cocaine use.

When the offspring were 16, the average age of the mothers was 41 years. Fifty per cent were married or living with a male partner, and 72.5% worked or attended school. On average, the women had 12.2 years of education.

The researchers found that 11.7% of the adolescents had a lifetime prevalence of conduct disorder, and 5% met the criteria for a current diagnosis of conduct disorder (at the age of 16). Sixty per cent of those with conduct disorder were male.

Youths who rated their parents as being more strict or more involved were less likely to have conduct disorder (CD). When quality of home environment was ranked on an ascending 10-point scale, the average rating was 5.34 from adolescents with CD compared to 6.07 from those without (p=0.005). Adolescents who had CD had, on average, experienced more key life events in the last year, as reported by their mothers (3.7 versus 2.8, p=0.005).

In total, 35% of children with CD were exposed to at least one drink per day in the womb compared with 16% among adolescents who did not have CD (p=0.003). There were no differences between the number of CD and non-CD adolescents whose mothers had consumed less than this amount during pregnancy.

Of the 67 adolescents who had CD, 24 (36%) had mothers who had consumed at least one drink per day during the first trimester of pregnancy, whereas 22 adolescents (33%) had mothers who had not consumed alcohol at all during this period. Of the 505 adolescents who did not have a diagnosis of CD, 80 had mothers who consumed at least one drink a day during pregnancy (16%), whereas 185 (37%) had mothers who had not consumed any alcohol during this period.

Risk of conduct disorder was not associated with alcohol consumption in the third trimester or with drug use during pregnancy as a whole (where marijuana, cocaine and other drugs were assessed separately). There was a borderline association between CD and cigarette smoking in the first trimester.

The researchers then produced a model in which they took into account the influence of demographic variables, drug and cigarette use, measures of parenting practices, home environment and past-year events. After these adjustments, they found that drinking more than one alcoholic drink per day was associated with an approximately three-fold increase in the odds of having CD as an adolescent (odds ratio [OR] =2.74; 95% confidence interval [CI] = 1.50 to 5.01). They also found that strict parenting reduced the odds of conduct disorder by 10% (OR =0.90; 95% CI, 0.83 to 0.96) and experiencing one of the notable life events in the past year increased the odds by 20% (OR =1.20; 95% CI; 1.07 to 1.34).

How did the researchers interpret the results?

The researchers said that “prenatal alcohol exposure above the level of one drink per day predicts a three-fold increase in the rate of conduct disorder in exposed offspring at 16 years of age”. They say that prenatal alcohol exposure should be considered as another risk factor for conduct disorder.


While this study has demonstrated an increased risk of conduct disorder with drinking one or more alcoholic drinks per day in the first trimester, there are several limitations to this study that should be taken into account when interpreting these results.

  • The sample of US women were from a prenatal clinic. Sixty-eight per cent were single, and 55% were African American and generally of a lower socioeconomic status. It is not clear whether these women would be representative of a general British population or of British mothers.
  • The authors reported that they did not have information about the psychiatric status of the biological fathers, so they were unable to control for this variable.
  • Conduct disorder is relatively uncommon, and in this study only 67 adolescents had a diagnosis of conduct disorder. Performing multiple analyses of these small numbers in sub-groups increases the possibility that some associations were found due to chance rather than any genuine relationship between factors.
  • The focus of this research was on alcohol use during the first and third trimester. However, the study also highlighted that other factors may influence the likelihood of conduct disorder, such as the home environment, lifestyle and parenting style. While these were taken into account in the study, they may not have been fully adjusted for.

This study benefited from a long follow-up of children whose mothers had consumed alcohol during pregnancy. But owing to the small size of the study, further research is needed to assess how drinking alcohol during pregnancy is associated with conduct disorder. Regardless, it is recommended that women avoid drinking alcohol during pregnancy for a number of other health reasons.

Analysis by Bazian
Edited by NHS Choices