Pregnant women are under fire again, and this time it's just for having a job. Aminatta Forna reports
Last week's news that pregnant women who worked more than 40 hours a week risked having a premature baby was designed to chill the heart of many expectant mothers. The story, which first appeared a few weeks ago, was still running early this week and within a short space of time had made the leap from the science pages to the women's sections.

The angle of the headlines was virtually uniform: "Pregnant career women are more likely to have a premature baby" declared the Mirror. "Career women run an increased risk of losing their babies", said the Sunday Times, going a step further, linking figures for premature births to infant mortality. In the Mail a penitent career woman was put on display. In redemptive tones, the former "City high flier" confessed: "I wanted it all. But the stress of work could have killed my unborn babies".

What readers were not told - or, if so, only much lower down the page - was that the key factor was stress, and that other forms of stress identified by researchers at the University of Perugia in Italy included unemployment, manual or repetitive work, work which required a lot of standing and poverty. In fact, the women most at risk of going into labour prematurely were more likely to be factory workers or women without a job at all. Daphne Metland, editor of Pregnancy Plus, the magazine of The National Childbirth Trust, was dismayed: "A spin was put on the stories which focused on professional women, who are not the ones most at risk."

By the end of the week, most GP's were telling their pregnant patients to go on as before. The National Childbirth Trust are not planning on giving any new advice to mothers-to-be. But it will only be a matter of time before the next scientific scare targeted at pregnant women and mothers comes around. Only days before the latest reports, the evening news carried reports that drinking alcohol causes infertility. Already deluged with improbable amounts of often conflicting information, pregnant women are beginning to feel under siege.

Partly it is the reporting of science, twinned with a popular understanding of science as absolute, which leads to the ability of reports to turn into scares. Studies may be small scale and isolated, findings may be preliminary and therefore far from conclusive. Also, according to the Royal College of Gynaecologists and Obstetricians, people need to understand the difference between a demonstrated correlation and actual proof.

Proof exists when a direct cause and effect is established. To use an absurd example - a correlation exists between death and walking down the middle of a road. But it is not walking down the middle of the road which cause death, but being hit by a car. A correlation may exist between stress and going into labour prematurely, but scientists still do not know what actually causes childbirth to occur prematurely. The research also indicates likely causes of stress, but stress might equally exist at home by an unhelpful husband or caring for other small children. Only once a study has been subject to rigorous analysis, including being repeated successfully by other scientists, can it be accepted as anything akin to "proof".

But the latest talk within the scientific community itself is that there may be a flaw in scientific thinking, specifically the measure of what constitutes a "significant" scientific finding. It's not clear from available information whether this relates to the Italian research on premature births, but it is true that it is the reason why, from a layman's point of view, major scientific breakthroughs seem to be hailed one after the other, yet one often hears nothing more.

One example given by Robert Matthews at Aston University, who recently published a paper for the European Science and Environment Forum, relates to pre-eclampsia, a potentially extremely dangerous condition that can affect 1 in 7 pregnant women which was found, in the early 1990's, to be preventable by taking low doses of aspirin. Within a short space of time it was suggested that expectant mothers should add taking aspirin to their already inflated list of "do"s and "don't"s. Yet by 1994, scientists had dismissed the claims of the earlier study after it failed to be verified by a major international study. Matthews and others argue that the determinant of what constitutes a "significant" finding - worthy of publication and regarded as scientific evidence - is far too low, boosting rather petty findings into something altogether more consequential.

But many reporters, particularly science correspondents, are capable of working out much of this for themselves. The truth is that when it comes to motherhood, the politics of blame is never very far from sight, especially when it relates to working mothers. Diane Eyer, an American academic and psychologist, is the author of Motherguilt. She argues that women's work, as the biggest social factor contributing to feminine independence, is constantly targeted as the cause of children's problems and has been blamed for failure to bond, attachment problems, delinquency and family breakdown.

The way newspaper and magazine coverage of stories is handled gives away the real agenda. For example, "the working mother is usually portrayed wearing a three piece suit and toting an executive briefcase, even though she's most likely to be a secretary, clerk, teacher or nurse." The Daily Mail's example was an investment banker! The latest report on premature childbirth could have been reported as an issue for employers or the government; in fact, in exactly the same week, the EC issued a directive aimed at producing more family-friendly work policies.

When it comes to reporting stories of male reproductive responsibility, it is a very different matter. Cynthia Daniels, a lawyer who teaches at Rutgers University and an expert in the politics of pregnancy, has reviewed quantities of coverage for her own study. "Male reproductive exposures (to alcohol, smoking, drugs or environmental toxins) are now proven or strongly suspected of causing not only fertility problems but also miscarriage, low birth weight, congenital abnormalities, cancer, neurological problems and other childhood health problems," she says. "Yet men have been spared the retribution aimed at women."

It is no surprise that few people realise this, considering the subject is barely covered in the press, and those articles which do tackle the issue generally do so in a straight-forward or even light-hearted way (with illustrations of besieged, embattled little sperm with cartoon faces). Men are neither blamed for their lifestyles nor held responsible for their children's health.

Daniels has also investigated how, in America, concern for unborn children has been systematically used to bar women from certain kinds of work, usually at the instigation of male-dominated unions. In January, a similar case came to light in the UK when a pregnant police recruit was ordered off a training course by her superiors After weighing all the evidence, the industrial tribunal upheld her complaint.

Daniels calls the entire phenomenon "the production of truth"; starting with the choice of research subject by scientists chasing funding and publicity, to the reporting of stories which frame public understanding, to the eventual framing of policies which then re-shape the working of society. And if that sounds too far-fetched, remember that the reason that Britain is alone in Europe in still having virtually no state provided childcare is that in the 1950's a British psychologist claimed that working mothers produced delinquent children.