Wednesday June 20 2012
Claims that alcohol is safe in pregnancy are misleading and potentially damaging
“Pregnant women can binge drink safely,” according to a report in today’s Metro. Expectant mothers should be able to “down up to 12 alcoholic beverages a week knowing it will have no ill effect on their offspring before the age of five”, the paper continued. Reports in several other papers were in agreement, with the Daily Mail claiming that a drink a day would not harm the baby’s development and the Daily Express reporting that 12 drinks a week is safe in pregnancy. So should pregnant women heave a sigh of relief and down a large glass of Chardonnay? Unfortunately, no.
The newspaper reports are based on a series of studies of women and their five-year-old children. The studies looked at the effect of different drinking patterns in early-to-mid pregnancy on the child’s intelligence, attention and other mental functions, such as planning and organisational ability.
An analysis of the five studies found that, overall, there was no effect of low-to-moderate weekly alcohol consumption on children’s neurodevelopment at the age of five. The research also didn’t find any association between binge drinking – defined here as five or more drinks on a single occasion – and children’s neurodevelopment. However, one of the studies did find a significant association between nine or more drinks a week and the risk of a low attention score in the child.
These studies are a valuable addition to the research on drinking in pregnancy and appear to provide some evidence that the occasional drink may not affect these particular neurodevelopmental outcomes in the child. However, they do not give the green light to an alcoholic free-for-all during pregnancy, as much of the media seems to imply. The research is in only a relatively small sample of pregnant women, and has not examined the wide breadth of potential adverse effects that alcohol could have upon the developing baby.
Heavy drinking in pregnancy is known to increase the risk of problems such as miscarriage, and is also known to increase risk of foetal alcohol syndrome (FAS) or its milder form of foetal alcohol spectrum disorder (FASD). FAS and FASD can encompass a wide range of problems in the child, from birth defects to learning and behavioural difficulties and problems with movement and co-ordination.
There is still uncertainty about what constitutes a "safe" level of alcohol during pregnancy, if any amount is safe. The research does not alter the current advice in the UK for women who are pregnant or planning to get pregnant. The Department of Health (DH) advises that alcohol is to be avoided in pregnancy if possible, while the independent National Institute for Health and Clinical Excellence (NICE) specifically advises women to avoid alcohol in the first three months of pregnancy because of the risk of miscarriage. If women choose to drink alcohol during pregnancy they are advised to drink no more than one to two UK units once or twice a week. The DH advises that women should not get drunk, which NICE also says may be harmful. A unit equals half a pint of standard-strength lager or beer, or one shot (25ml) of spirits, while one small (125ml) glass of wine is equal to 1.5 UK units.
Where did the story come from?
The five studies that form the basis for these reports were carried out by US, Norwegian and Danish researchers. Funding was provided by the US Centers for Disease Control and Prevention (CDC) and various other organisations. The studies were published in the peer-reviewed British Journal of Obstetrics and Gynaecology (BJOG). The studies comprised research into:
- the effects of low-to-moderate alcohol in early pregnancy upon IQ in five-year-olds
- the effects of binge drinking in early pregnancy upon general intelligence in five-year-olds
- the effects of low-to-moderate alcohol or binge drinking in early pregnancy upon selective and sustained attention in five-year-olds
- the effects of low-to-moderate alcohol or binge drinking in early pregnancy upon executive function (for example, ability to plan or organise) in five-year-olds
- a combined analysis looking at the effect of different alcohol consumption during early-to-mid pregnancy and effects upon the child’s intelligence, attention and executive function
This Behind the Headlines appraisal focuses on the last study of these, which summaries the other four reports.
Coverage in the media was confusing, potentially misleading and damaging. Several papers, such as the Metro and the Mail, claim that binge and heavy drinking during pregnancy is safe, while the BBC and the Telegraph report that low or moderate drinking does “no harm” to the child. The claim made by the Express and the Mail that pregnant women can safely consume 12 alcoholic drinks a week is particularly worrying. Heavy drinking is known to have a risk of affecting the development of the foetus, and one of the studies did demonstrate that nine or more drinks a week was associated with a lower attention span in five-year-olds.
While the details of this study were generally reported fairly accurately, the general conclusion in all national media sources – that there is a safe level of alcohol in pregnancy – is not borne out by the evidence on miscarriages and foetal alcohol syndrome. The research simply shows that in this small Danish sample, children surviving to the age of five show no significant neurodevelopmental damage at that age.
What kind of research was this?
This was a prospective cohort study that included 1,628 Danish women and their five-year-old children. It aimed to assess the association between drinking patterns in early-to-mid pregnancy and various cognitive abilities of the children at the age of five. The authors point out that while heavy drinking is known to affect neurodevelopment, less is known about the effects of low-to-moderate alcohol consumption, while the results of studies of moderately high alcohol intake during pregnancy have been inconsistent.
Cohort studies are useful for looking at possible associations between various lifestyle factors (such as drinking alcohol) and health outcomes (such as a child’s neurodevelopment). They enable researchers to follow large groups of people for many years, but they cannot establish cause and effect.
What did the research involve?
The women in the study, who were drawn from the Danish National Birth Cohort, were recruited between 1997 and 2003, at their first antenatal visit to a GP. The women were asked about their drinking patterns in early and mid pregnancy during interviews undertaken at the time. They were asked about the number of alcoholic drinks they consumed weekly, including beer, wine and spirits. The definition of a drink in this research comes from the Danish National Board of Health, which states that one standard drink is equal to 12 grams of pure alcohol. In the UK, the volume of alcohol in a drink is measured in units and one unit of alcohol is defined as 7.9 grams.
From this information, researchers than categorised the maternal average alcohol intake into four groups:
- women who did not drink at all, who were the reference group (or control)
- women who drank one-to-four drinks a week
- women who drank five to eight drinks a week
- women who drank nine drinks a week or more
They also obtained information on episodes of binge drinking, defined as five or more drinks on a single occasion (in the UK this would be about seven units).
Mothers who had been interviewed about drinking when they were pregnant were then invited to participate in the study by letter about four-to-six weeks before their child’s fifth birthday. Those who wished to do so were sent a questionnaire about their child’s general postnatal health and development as well as their own lifestyle. Between 2003 and 2008 their children were tested on measures of their general intelligence, attention and "executive" function (for example, their ability to plan or organise), using established neuropsychological tests. Their mothers were also tested for IQ. All tests were carried out by trained psychologists.
The researchers used standard statistical methods to analyse associations between drinking patterns and the child’s neurodevelopment, taking into account factors that might affect this, including parental education, maternal age and IQ, and prenatal maternal smoking. Their overall analysis combined the results of the tests on the children, to give a general measure of neurodevelopment.
What were the basic results?
A total of 1,628 women took part in the study. The average age in pregnancy was nearly 31; about half were first time mothers; about 12% were single and nearly one-third reported smoking during pregnancy.
Overall, in the combined analysis the researchers found no statistically significant association between average weekly alcohol consumption and the intelligence, attention and executive function of the children at the age of five. Nor was there any association between binge drinking and these outcomes.
The results were demonstrated in each of the separate papers looking at alcohol consumption and each of the three measures of neurodevelopment. However, one of the separate analyses found a significant association between maternal consumption of nine or more drinks a week and increased risk of the child having a low overall attention score (odds ratio 3.50, 95% confidence interval 1.15 to 10.68), while another paper found an association between nine or more drinks a week and increased risk of the child having a low full-scale IQ score (odds ratio 4.6, 95% confidence interval 1.2 to 18.2) and low verbal IQ score (odds ratio 5.9, 95% confidence interval 1.4 to 24.9).
How did the researchers interpret the results?
The researchers say they found no significant association between low-to-moderate average weekly alcohol consumption and any binge drinking during early-to-mid pregnancy and the neurodevelopment of children at five years old. They did not say that alcohol consumption or binge drinking in pregnancy were "safe" for women or their unborn children.
This study had strengths in that it recruited women from a national birth cohort, collected detailed information about women’s patterns of drinking in early pregnancy and also tested several different aspects of children’s neurodevelopment using validated tests. It appears to provide some evidence that low and moderate levels of drinking in the first half of pregnancy may not affect these particular neurodevelopmental outcomes in the child. However, the study had some limitations.
Pregnancy is a very common event. Therefore, for this particular issue this sample of 1,628 Danish women is too small to base any firm conclusions, without replication in other population samples.
Low confidence in results
Most of the calculated risk associations have very wide confidence intervals. For example, although the risk increase of low verbal IQ score with more than nine drinks was 5.9, the confidence interval is 1.4 to 24.9, which means the true risk figure could lie anywhere in between these numbers. Having such wide confidence intervals means we can have less confidence in the accuracy of these risk measures. This may be a reflection of the relatively small sample size, which means that the study was not able to detect subtle neurodevelopmental effects associated with alcohol consumption.
Unknown confounding factors
Despite the authors’ efforts to adjust their results for confounders, it is always possible that other unmeasured lifestyle, social and demographic factors could influence the results.
Accuracy questions over self-reporting of alcohol consumption
It is also possible that women did not accurately report their drinking patterns, which could affect results.
Limited participation by women
Only 51% of the initial number of women invited to take part participated in the study, which introduces the possibility of bias. For example, those who chose not to participate may have included women who drank excessively in pregnancy, or who had children with neurodevelopmental problems.
Single early measures of neurodevelopment taken
The children’s cognition was only measured once at five years old, which, as the authors concede, is a relatively early stage in neurodevelopment.
Possible underestimation of binge-drinking risks
Binge drinking as defined in this study constituted five or more drinks on one occasion. Among women who were categorised as binge drinkers in this study, 69% reported only one episode of bingeing in early pregnancy. Therefore, this may not give a reliable indication of the risks of binge drinking in pregnancy.
Of critical importance is that this study has not examined the wide breadth of potential adverse effects that alcohol could have upon the developing baby. Heavy drinking in pregnancy is known to increase the risk of problems such as miscarriage, and is also known to increase risk of foetal alcohol syndrome (FAS) or its milder form, foetal alcohol spectrum disorder (FASD). FAS and FASD can encompass a wide range of problems in the child, from birth defects to learning and behavioural difficulties and problems with movement and co-ordination.
In conclusion, the research does not alter the current advice in the UK for women who are pregnant or planning to get pregnant. The Department of Health advises that alcohol is avoided in pregnancy if possible. The National Institute for Health and Clinical Excellence (NICE) advises women who are pregnant to avoid alcohol in the first three months in particular, because of the increased risk of miscarriage. If women choose to drink alcohol during pregnancy they are advised to drink no more than one to two UK units once or twice a week. A unit equals half a pint of standard-strength lager or beer, or one shot (125ml) of spirits, while one small (125ml) glass of wine is equal to 1.5 UK units.
Further research is required in this important area. The key point is that there is still uncertainty about what constitutes a "safe" level of alcohol during pregnancy.