Sir

Perhaps the high tide of genetic determinism in all things biological is beginning to recede a little at last. I was relieved to read of the caution advised by the UK bioethics committee with regard to genetic screening as a means of predicting the susceptibility of individuals to mental disorders (Nature 395, 309; 1998).

Even the most fervent supporters of the contentious idea that there is a significant genetic component to mental disorders would probably be prepared to admit that the correlation between the occurrence of a gene and that of a disorder in these cases is statistical. There are plenty of people with the gene, but not the disorder, and plenty of others with the disorder, but not the gene. This is also true of the much publicized genes ‘for’ heart disease and breast cancer, among others.

What does the individual do with the knowledge that they carry such genes and what advice can genetic counsellors give them? Don't smoke, don't drink too much, be careful about what you eat, take a little exercise, avoid stress and, you there with that gene, even more so? This is stretching the concept of genetic determinism beyond utility.

At a recent conference on the commercial potential of genomics I heard a representative of those who wish to offer such screening to all individuals on a commercial basis concede that their counsellors sometimes had difficulty in communicating the importance of the information to the recipients. He appeared to think that this was a problem of education and perhaps he was right, though the problem may be his, not theirs. The recipients may have been sufficiently well educated to realize that statistics are properties of populations, not individuals, and, even if the assumptions about the genetic component of the disease were actually correct, the information they were receiving was absolutely useless to them.