HEALTH / Common procedures: Choosing a baby's sex

POWERFUL men have always wanted a son to whom they could pass on their business - or their kingdom. Mothers of sons may long for a daughter. Many couples would like one child of each sex. Folk healers, wise women, and dishonest charlatans have claimed to be able to guarantee the unborn child being male by methods ranging from lying on the right side after sex to the more drastic surgical removal of the left testis. With an unavoidable 50 per cent success rate, these salesmen have been able to offer 'money back if not satisfied' and still become rich.

Medical science is at last taking an interest in helping couples to choose the sex of their children - for two reasons. Firstly, several of the most common serious genetic disorders are sex-linked: haemophilia and Duchenne muscular dystrophy, for example, occur only in boys. Parents who know they carry the abnormal genes may reasonably ask for help in tipping the balance in favour of a newly conceived child being a girl. Secondly, family planning programmes in developing countries which are trying to persuade couples to have only two children are more likely to succeed if couples can have two of the sex they want.

The starting point is that about 106 boys are born for every 100 girls (but boys die more often in infancy, childhood, and young adult life, so that the numbers even out). Making the sums easy and assuming a 50 per cent chance of a boy or girl, then the chance of a couple with two children having one boy and one girl is also 50 per cent. If a couple with two boys - or three, or four - try again for a girl, their chances are still 50 per cent, as in betting on red or black at roulette. How can these chances be improved?

Attempts have been made to change the natural sex ratio by advising parents to have sex at a particular time in the woman's menstrual cycle (the closer to ovulation the better the chances for a girl) or by advising changes to the diet (eating acid foods is supposed to make a girl more likely). There is no good evidence that any of these methods has more than a marginal effect.

The most recent and apparently the first really effective method uses artificial insemination with sperm from the father-to-be which has been processed to increase the proportion of sperm carrying chromosomes for the sex the parents want. Data from 65 clinics around the world using this technique have been published in the research journal Fertility and Sterility; these showed that 72 per cent of 1,034 couples wanting a boy were successful, as were 69 per cent of 193 wanting a girl.

The failure rate of this method is still, therefore, substantial. It is possible to improve on the figures by using in vitro fertilisation and testing the embryo before implanting it in the mother's uterus, but this is expensive high-technology medicine that will only be used when there is a risk of a serious sex-linked genetic disorder. Doctors in most countries refuse - on ethical grounds - to terminate pregnancies when the foetus is of the unwanted sex.

For couples who simply have a preference for the sex of their next baby, there is still no easy, reliable method. If one were to be found, the evidence suggests that it might lead to more boys being chosen than girls.