I’m always reading newspaper stories reporting the discovery of “a gene for” traits, such as depression, schizophrenia or homosexuality. In every case, the discovery is a false dawn, subsequently unreported in the papers to correct the earlier claim.

The Human Genome Project (completed in 2003) aimed to change all that by examining genes directly. The good news is that it looks as if it may succeed - although not in the direction the geneticists were hoping.

So far, in all fields of human psychology, the lack of findings proving a significant role for genes has been so resounding that it has led the most influential and widely-respected figures to reverse their earlier claims. They now admit that, after all, there is no “gene for” any of the major mental illnesses. This is good news because there is stacks of evidence that, whether it be parents, teachers or children themselves, people prosper better if they believe in nurture rather than nature.

The story of the most promising gene so far illustrates how elusive their influence is proving to be. A New Zealand study found that certain genetic variations create the vulnerability to depression. If you have the wrong genes, good early care can protect you, but bad care combined with these genes means depression in later life. Conversely, the presence of good genes is protective: with them, even if you have bad care - up to and including abuse - you are less likely to be depressed.

But subsequent studies have called these findings into question. True, four studies partly replicated the original finding, but these were small samples. They categorised the gene in different ways and in two of them, the replication was only for women, not men. Three other small studies provide partial support. But as a whole, this evidence is far from conclusive.

On the other hand, three studies have not found that genes have any effect on depression. In one - based on a larger sample than all the supportive studies - people with the genetic variation that should have made them vulnerable to depression actually turned out to be less vulnerable.

Furthermore, in large-sample studies of depressed patients, their genes do not vary in the predicted manner compared with non-depressed people.

On a wider scale, there are reasons to question the role of this genetic variation. Strong predictors of who gets depressed in a developed nation are being on low income and being a woman - they are twice as likely as the rich and males to be depressed. Two studies have found no greater occurrence of the supposedly depression-inducing genetic variation in low- income people. Nor is the gene found more often in women than men.

Taken overall, it is quite possible that the genetic variation may play no part in causing depression. The key factor is whether the environment is adverse and especially whether there has been childhood maltreatment.

Even if genes do, in the end, turn out to be important, it may still be best to believe they are not. For instance, if you suffer a mental illness, believing it’s down to genes means you are less likely to recover, probably because you feel there’s nothing you can do. Likewise, if you are a parent and you believe your child’s development is basically due to nature, you are significantly more likely to maltreat them, up to and including abuse.

Secondary pupils who take a flexible view of their abilities get significantly higher grades than ones who believe they are fixed. When samples of teens and undergraduates are taught to think of themselves as being malleable rather than fixed, they get better grades as a result of the tuition. The mechanism is clear: change the belief from fixed to malleable and you change the motivation, which improves the grades.

Other evidence shows how critical it is that parents and teachers do not regard genes as fixing children’s capabilities: if either group has negative expectations of the child, academic performance suffers.

Both parents and teachers can have a huge impact on children if they get it right. As a society, the sky is the limit as to how mentally healthy and capable we can enable our children to become.

- A New Zealand study .

Caspi, A., et al (2003) Science, 301, 386-9

- Four studies have partially replicated .

Wilhelm, K., et al (2006) British Journal of Psychiatry, 188, 210-15

Eley, T.C., et al (2004) Molecular Psychiatry, 9, 908-915

Kendler, K.S., et al (2005) Archives of General Psychiatry, 62, 529-35

Zalsman, G., et al, in press, American Journal of Psychiatry

- Three other small studies .

Grabe, H. J., et al (2005) Molecular Psychiatry, 10, 220-4

Fox, N., in press, Psychological Science

Kaufman, J., et al (2004) Proceedings of the National Academy of Sciences USA, 101, 17316-21

- On the other hand, three have not .

Gillespie, N.A., et al (2005) Psychological Medicine, 35, 101-11

Willis-Owen, S.A., et al (2005) Biological Psychiatry, 58, 451-6

Surtees, P.G., et al (2006) Biological Psychiatry, 59, 224-229

- One of these had a much larger sample .

Surtees et al, 2006

- Furthermore, in large studies of samples of depressed patients .

Mendlewicz, J., et al (2004) European Journal of Human Genetics, 12, 377- 82

Lasky-Su, J.A., (2005) American Journal of Medical Genetics, 133B, 110- 5

- Two studies have found no greater .

Manuck, S.B., et al (2004) Psychoneuroendocrinology, 29, 651-68

Manuck, S.B., et al (2005) Psychological Medicine, 35, 519-28.