Don't overstate the risks of drinking while pregnant

It is no surprise that the recently published research into the epigenetic effects of alcohol consumption during pregnancy (1) has received considerable media attention. Practically all pregnant women will worry - understandably - about the health of their future children, and will quickly seize on information or reassurance from credible-seeming sources - a fact well recognised by our commercially-motivated press.

Last year, I found myself embroiled in something of a media stir (2-4) after suggesting that overstating the dangers of drinking while pregnant was both ethically problematic and possibly counterproductive (5). Needless to say, it was not my intent to encourage pregnant women to indulge in unrestrained binge-drinking. The evidence we currently have certainly suggests that heavy drinking is a bad idea (and not only for pregnant women!)

My concern was then, and remains, about the medical establishment's approach to less extreme behaviour. When research from University College London (UCL) appeared to confirm earlier findings that light-to-moderate alcohol consumption in pregnancy had no discernable harmful effects on resulting children (6) a spokesperson for the British Medical Association (BMA) immediately expressed concern that the findings might 'lull women into a false sense of security and give them the green light that there is no problem with drinking during pregnancy.' That being so, and while the effects of light drinking during pregnancy remained uncertain, the BMA continued to recommend that 'the simplest and safest advice is for women not to drink alcohol during pregnancy' (7).

At first glance, such an approach appears prudent. If we don't know for sure whether a safe drinking level exists, still less what it might be, there is obvious appeal to an approach that appears to err on the side of safety. On closer inspection, though, the flaws in the BMA's approach become apparent. For one thing, a policy of preaching 'abstinence only' seems more than a little paternalistic - not to say, patronising. The BMA's own general advice about information disclosure is that '[d]octors should respond honestly to direct questions from patients and, as far as possible, answer questions as fully as patients wish' (8). Where the available evidence doesn't allow a straightforward, unambiguous answer to a question like 'how much is it safe to drink during pregnancy?', an honest response would be to admit as much to a patient.

The BMA's Vivienne Nathanson's responded to the UCL study by noting that 'so-called 'heavy' and 'moderate' drinking harm the unborn baby. Very light drinking may or may not'. So what is the problem with telling pregnant women this? Such uncertainties are part and parcel of every adult's life. We know that many aspects of our diets, our hobbies, and our lifestyles are not entirely 'safe', in the sense of exposing us to no risk whatever, and we - or at least the reasonable among us - do not demand of our doctors precise guarantees about safe levels of absolutely everything. Why, then, is it assumed that pregnant women, uniquely among competent adults, are unable to deal with the same uncertainty? Why does the BMA assume that admitting that no one knows precisely how much it is 'safe' for her to drink will throw a pregnant woman into a tizzy of confusion, or be interpreted as a 'green light' for unrestrained, bacchanalian indulgence?

Of course, simple half-truths may be easier for doctors to communicate than complex realities. But a doctor-patient partnership based on mutual trust, respect and communication is not consistent with substituting a true, complicated account of risk with a more simplistic, but less accurate, one. The days when doctors routinely withheld information - about risks, alternatives, side effects or prognoses - on the grounds that patients would become confused and make bad decisions are, supposedly, long gone. Except, it seems, for pregnant women.

It is quite possible that all of this is borne of paternalistic instinct that, while out of keeping with recent trends in medical ethics, is borne of a genuine benevolence. The problem here is that benevolence of intent does not always translate into benevolence of outcome. If the medical establishment is seen to be exaggerating risks that recent (and well-publicised) studies have shown to be negligible, then a real risk arises that their advice on genuine dangers will carry less authority. Trust in expert advice is a resource that is easily squandered, but not easily restored. Professing pseudo-certainties as a means to manipulating patient behaviour is both insulting to patients, and potentially damaging to that trust. It is also worth considering whether constant additions to the list of dos and don'ts may result in a sort of 'advice fatigue' whereby, in the face of unduly onerous requirements, pregnant women abandon reliance on medical advice and epidemiological research altogether, and simply fall back on anecdote or instinct.

If this approach seems to view pregnant women as too stupid to be trusted with nuanced information, the alternative interpretation is even more pernicious. We may be witnessing a revival of some kind of aura of sacrifice around motherhood, whereby a women proves her 'good mother' credentials by showing how much she will give up for her child - even if her sacrifice doesn't benefit that child in the slightest. Hence, pregnant women who drink are routinely depicted as hedonistic and selfish, even when there is practically no evidence that they are doing any harm.

This is the attitude, perhaps, that led Ellen C G Grant to suggest in her BMJ Rapid Response that it was time that: 'society grew up and put the health of future generations before profiteering and hedonism' (9). And that saw 'Helen' from Norwich to respond to the Daily Mail's coverage of my article earlier this year with the rhetorical 'So it is better to risk the health of the unborn baby than to perhaps offend a woman who cannot control her urge to drink alcohol?'

To fall foul of this new socio-ethical imperative, it seems, a woman need not give more weight to her own interests than those of her potential future child. Even placing her interests on the scales is interpreted as indicative of questionable prematernal commitment.

So how does the discovery that alcohol seems to affect gene expression affect my earlier argument? Not much, really. My contention was, and remains, that pregnant women - no less than any other class of patients - should be offered the most honest and reliable advice and answers. If a credible link were ever established between light alcohol consumption and developmental defects then it would be at least as patronising, at least as incompatible with patient autonomy, to withhold or distort that information - even if doing so made some women feel better about their choices.

The Queensland study reported in BioNews 542 (1) may go a long way to helping us understand the mechanisms by which heavy alcohol use affects development in utero. It may, for example, help explain why - as its authors note - only around 5 per cent of women who drank heavily while pregnant give birth to children with Fetal Alcohol Syndrome. It is less likely, however, to tell us anything new about safe levels of consumption; the mice who were the study's subjects had a blood-alcohol level 150 per cent of the UK's legal driving limit, which is a good deal more than the 'one or two units' previously regarded as acceptable.

While the situation remains uncertain, many pregnant women will continue to worry that they may be harming their future children somehow. The BMA's increasingly extreme position on drinking risks adding unnecessarily to those worries, or even alienating those women, most who are at very low risk - while having presumably negligible impact on high-risk drinkers who ignored the previous guidelines anyway.