But for the fifth of all pregnant women on antidepressants there is the added concern about the risk their medication poses to their unborn child. Experts agree that a severely depressed mother can be a greater risk to her unborn child than antidepressants, as she might fail to eat properly or engage in harmful behaviour. Indeed, babies of very depressed mothers can have lower birth weights and more pre-term complications.

But there is also the potential for the antidepressant to harm the unborn child. Two weeks ago, The New England Journal of Medicine reported a study which suggested that pregnant women who took antidepressants, such as Prozac, after the 20th week of pregnancy, were six times more likely to have to babies with persistent pulmonary hypertension, a rare but sometimes fatal breathing problem.

Just under 20 per cent of pregnant women in UK take some kind of antidepressant, according to recent estimates.

“It's a not insignificant number,” says Dr Margaret Oates, a specialist in Perinatal Psychiatry at Queen’s Medical Centre, Nottingham. “Most of the prescriptions for antidepressants in early pregnancy are probably unnecessary. Because for mild depression, there is no evidence that antidepressants are any more effective than psychological treatment.”

Dr Oates believes that, anyway, women are often protected from depression during pregnancy by the surge of hormones, particularly oestrogen.Her advice to patients who do need to take antidepressants is that “the evidence for foetal abnormality is there but in terms of absolute risk . . . probably fairly low”. However, she tapers the drugs so that mothers will be off medication at the time of delivery.

She also cautions against stopping treatment abruptly when women learn that they are pregnant, because many will suffer “the mother and father of all withdrawal effects”.

Lauren Slater, a psychologist and author who has documented her own travails with depression, discovered this. Four days after learning that she was pregnant she found herself lost in a pile of conflicting medical reports and stopped taking the Prozac that had kept her stable for ten years. Six weeks later she recognised the signs of relapse, as she describes in her book Love Works Like This: “It is difficult to move; my limbs say no. I stand at the top of the stairs, looking down and considering. I spend a long time weighing the pros and cons of descending. The decision feels tricky and enormous.” Slater was quickly put back on antidepressants.

Dr Oates says that psychiatric specialists have become more cautious about prescribing selective serotonin reuptake inhibitors, or SSRIs, the class of antidepressants that includes Prozac and Seroxat, because of concern that they might cause neonatal problems.

Newborns exposed to SSRIs can experience “withdrawal” symptoms that include persistent crying, tremors, sleep problems and gastrointestinal problems, although this is temporary. Dr Oates is particularly cautious about GlaxoSmithKline’s Paroxetine, known as Seroxat, since the US Food and Drug Administration reported early findings of two studies that suggested that the drug may cause heart defects in infants when women take it during the first three months of pregnancy.

However, Dr Lee Cohen, a perinatal psychiatrist at Massachusetts General Hospital in Boston, who runs one of a handful of specialist maternity psychiatry units in the US, believes that risks from SSRIs to the infant are small and benign. He points to a recent study in the Archives of Pediatrics, which reported the same “withdrawal” symptoms as previous studies, but found that most children needed no medical treatment.

Dr Cohen does not believe in reducing medication as the pregnancy develops. “The dose that got you well, keeps you well,” he says. He led a study, released earlier this month, which indicated that 68 per cent of 207 women who discontinued their medicine during pregnancy relapsed into depression.

Relapse close to the end of a pregnancy is a concern, because women are ten times as likely to suffer from depression in the first year after they have a child than at other times in their lives, and those with a history of depression stand an even greater chance of becoming ill.

Dr Cohen stresses that the Paroxetine studies flagged by the FDA are as yet unpublished, and added that they indicated a tiny risk. One of them gave warning of a risk of heart defects in babies whose mothers had taken Seroxat of about 1.5-2 per cent, compared with a 1 per cent risk in the whole population.

Dr Cohen believes that the debate exists only because of the stigma associated with mental illness. “We would never be having this discussion if you were talking about hypertension, diabetes, thyroid disease.” He adds that far more is known about antidepressants than many other medicines taken by women during pregnancy, such as antacids and sleep aids.

(Indeed, far less is known about the effects of alcohol on the unborn child. Experts know that alcohol can harm the foetus, but don’t know how much is too much.) Slater’s medication had no adverse effect on her daughter, indeed she even asks whether Prozac did her good.

“I sometimes wonder now if the Prozac was not a little like a vitamin. I spent nine months of pregnancy and so many months thereafter thinking of it as a toxin, but who’s to say those little spheres did not in some way add folds to your brain, give greenness to your eyes.”