Randolph Nesse is on a mission. Normally a professor of psychiatry at the University of Michigan, he is spending a year at University College London, and one of his targets while in town is the "clinician's illusion". By this, Nesse means the belief that, as he puts it: "The symptoms are the disease." In fact, a cough, a fever or an anxiety attack may well be a subtle and well-considered response that the body is right to produce.

Nesse, one of the organisers of today's Academy of Medical Sciences meeting, hopes that this idea will "help physicians take a broader perspective" on illness.

Take anxiety, he says. Our ancestors lived in an environment in which there were many things to fear. "Any looming object in your field of view that is rapidly getting larger will cause fear," he says. "That is absolutely innate. In the past it might have been useful in an environment where there might be a lion by the waterhole. Now it is useful to avoid being knocked over by a car." By contrast, a fear of snakes is not innate but is very easily acquired, while a fear of firearms and knives takes longer to learn. People have not coexisted with either long enough for an innate fear to have developed.

Nesse hopes that the evolutionary approach to disease will allow us to take a longer-term look at illness. "A panic attack in Sainsbury's is probably bad," he says, "but there are other settings in which it is a correct response."

ÖйúAƬ

But it would be wrong to conclude that today's humans are the helpless inhabitants of minds and bodies shaped for a distant existence on the plains of Africa and forced today to cope with living in cities and working at a desk. Human populations, Nesse says, can change some of their attributes, such as height, in a few generations. Interactive relations between hosts and pathogens can also change rapidly. "It used to be thought that these relations were shaped by selection," Nesse says. "In fact we now know that they develop in a particular way. The pathogen becomes as virulent as it can while keeping its chance of being transmitted at an ideal level."

By contrast, Nesse points to dietary preferences in the developed world. "Our diet is killing us and the many publicity campaigns to get us to drink less alcohol or eat less fat are not succeeding. This is because our minds and our motivations are designed for a very different environment." This might sound like a counsel of despair but it is not. Nesse believes that the evolutionary perspective might help us think better about how to design more persuasive public health messages.

ÖйúAƬ

He also hopes that an evolutionary approach might encourage a more sophisticated look at our emerging knowledge of human genes. "We need to know why genes that cause disease persist. In fact it may turn out that these genes are selected for." Our knowledge of the human genome requires a sophisticated idea of how genes work in people and the environment they inhabit, going beyond the idea of good and bad genes that lead to either health or disease.