I read this article in The West a little while ago, and it hit my bad-science peeve button. “Half of WA women drink throughout pregnancy“. This beat-up is certainly not alone in drawing conclusions that do not follow from the data, but it’s a particularly egregious example.
The newspaper reports on a study that looked at data from the Western Australian Midwives Notification System for 4719 births to non-indigenous women in 1995-1997. Bear with me while I pull it to pieces. [Emphasis is mine.]
Almost half of all Western Australian mothers-to-be drink alcohol throughout their pregnancy, and some even admit to bingeing in the final months before giving birth.
The women who were prepared to drink on were also more likely to smoke during their pregnancy, according to a study of 4700 mothers in the State. Doing so placed these women at the most risk of having their baby prematurely, according to the research by WA’s Telethon Institute for Child Health Research.
“Our research shows pregnant women who drink more than one to two standard drinks per occasion – and more than six standard drinks per week – increase their risk of having a premature baby,” says institute researcher Colleen O’Leary. This was the case even if the women stopped drinking before their second trimester, Ms O’Leary said.
“The risk of preterm birth is highest for pregnant women who drink heavily or at binge levels, meaning drinking more than seven standard drinks per week, or more than five drinks on any one occasion.”
The study focused on a random selection of non-indigenous women who gave birth between 1995 and 1997, and they were quizzed on their pregnancy and pre-pregnancy drinking habits.
Fewer than 20 per cent of women abstained during the pre-pregnancy period, but this increased to 57 per cent in the first two trimesters before settling to 54 per cent in the third trimester.
“Low” or “moderate” drinking came in at 44 per cent during the third trimester, while more than two per cent of women admitted to “binge” or “heavy” drinking in the final months before birth.
The study found a low birth weight was more likely to be caused by a mother’s smoking rather than drinking.
And, while there was no difference for women who abstained or drank low levels of alcohol, it said abstinence was still the safest option. […]
So, do the conclusions quoted in this newspaper report reflect the reality of the study? I tracked the full study down (Hat tip: Clive). Here are the main points from the abstract:
“The effect of maternal alcohol consumption on fetal growth and preterm birth”
(CM O’Leary et al, BJOG 116(3), pp 390 – 400, doi 10.1111/j.1471-0528.2008.02058.x)
Objective: To investigate the relationship between prenatal alcohol exposure and fetal growth and preterm birth and to estimate the effect of dose and timing of alcohol exposure in pregnancy. […]
Results: The percentage of SGA infants and preterm birth increased with higher levels of prenatal alcohol exposure; however, the association between alcohol intake and SGA infants was attenuated after adjustment for maternal smoking. Low levels of prenatal alcohol were not associated with preterm birth; however, binge drinking resulted in a nonsignificant increase in odds.
Preterm birth was associated with moderate and higher levels of prenatal alcohol consumption for the group of women who ceased drinking before the second trimester. This group of women was significantly more likely to deliver a preterm infant than women who abstained from alcohol (adjusted OR 1.73 [95% CI 1.01–3.14]).
Conclusions: Alcohol intake at higher levels, particularly heavy and binge drinking patterns, is associated with increased risk of preterm birth even when drinking is ceased before the second trimester. This finding, however, is based on small numbers and needs further investigation. Dose and timing of prenatal alcohol exposure appears to affect preterm delivery and should be considered in future research and health education.
There is no room for doubt here at all – the study did not, just plain did not, find that low or moderate levels of prenatal alcohol intake were associated with preterm birth. It’s as simple as that. Generalising from “high levels of alcohol intake in a certain pattern during pregnancy may cause problems” to “Women should never drink in pregnancy, not a drop; any level of alcohol is dangerous” is just plain wrong. It’s just plain bad science.
There is also a word change in the abstract conclusions and in the newspaper report that catastrophically misrepresents the data. “This was the case even if the women stopped drinking before their second trimester” was not actually what they found. Here is the money quote from the full study:
“Moderate to heavy alcohol intake resulted in an increased risk of preterm birth only in women who stopped drinking before the second trimester.”
“Only“. That’s right. The study failed to find any association even of heavy or binge drinking with preterm birth, until they plucked out one particular group of women: those who drank heavily in the first trimester, then stopped.
These researchers had to work damn hard to pick out a group for which alcohol intake seemed to make any difference at all – and in that group, the confidence interval extended to 1.01. The researchers are going to have to work quite a bit harder to convince me that this association is sufficiently distinguishable from cherry-picking, a chance result of analysing and re-analysing the data in a desperate search for something “statistically significant” – especially given the complete lack of so much as a trend in the other groups. The authors claim a trend – but I can’t see it. Look for yourself.
Click through for the larger version of this table. Look down the second-to-right column, which is the fully adjusted odds ratio (OR) for preterm birth for the various alcohol categories. For those who aren’t familiar with this type of data presentation, Any OR with a confidence interval (CI) encompassing 1.0 is not statistically significant.
On small for gestational age (SGA) babies : No association at all was found between SGA and alcohol intake, once smoking was controlled for. Not “a low birth weight was more likely to be caused by a mother’s smoking”, but there was no association at all with drinking as an independent factor. (You can see those data here).
There are a couple of methodological issues worth pointing out. The sensationalist opener “Almost half of all Western Australian mothers-to-be drink alcohol throughout their pregnancy” is based on data in which the “low” category extends right down to include women who drink one unit of alcohol less than once a month. It seems, from my reading of the full study, that one single drink per trimester would get you categorised as someone who “drank alcohol throughout her pregnancy”. Count me in to those stats!
There was also no attempt to control for nutritional status, previous late-term abortion, other past cervical procedures, or infections that may be related to preterm birth. Many other potential confounding factors were controlled for, though perhaps not always in particularly effective ways. (How many new mothers disclose every bit of illicit drug use? I sure wouldn’t have, had I been in that category.)
The “We must tell them there’s no safe dose” mentality comes from a place of infantilisation of women. It is assumed that we’re stuck in the pre-rational child phase, and that any ‘concessions’ by those who think they are in power will lead us to ‘take’ far more than we’re entitled to. (Inch, mile, etc).
It is assumed that if women are told the actual data – that there is no evidence that having a couple of standard drinks harms a fetus – that we will immediately rip off our matronly long skirts and whoop it up in nights out on the town. It is assumed that we’re just hanging back from boozing it up because we’re waiting for “permission”. It is assumed that we just couldn’t possibly understaaaaaand the reality of the dose being the poison. It is assumed that it is doctors’ and midwives’ responsibility to shield us from reality and give us “unambiguous” absolutist messages about abstinence.
It is assumed that we as women, as incubators, cannot take responsibility for ourselves and our decisions. So someone else must take it for us, and that those people need to twist and simplify the message in order for us to comprehend it with our tiny hormone-soaked inferior brains.
So don’t believe everything you read in the paper. I’m Lauredhel, and I’m sick of this shit.