In medicine, twin studies use comparisons between identical and non-identical twins to explore the contributions of genes and environment to health and disease.

130 years ago, Darwin's cousin Francis Galton (the pioneer of twin research) had already recognised that twins provide a 'naturally occurring experimental design', a 'living laboratory' in which to study the effects of nature and nurture. And he had also noted similarities in their constitutions, their tendency to develop similar diseases.

Since then, twin studies have fallen in and out of favour as determinist and environmentalist models of human nature have waxed and waned. Today, medical researchers are increasingly recognising the power of the twin study model, combined with new gene technologies, to tease out the genetic bases for a range of common but complex diseases, such as cancer, heart disease, arthritis and diabetes. Modern twin studies can also help to locate the genes and explore the metabolic pathways involved, opening up the possibility of finding new ways to prevent or treat such illnesses.

In any trait, both genes and environment are involved and it is notoriously difficult to separate out their influences. Twin studies do so by comparing identical with non-identical twins. Comparing the two groups produces information on the relative contribution of genes and environment, and how the two interact.

Studying identical twins alone is less useful, since they share not only all their genes but also their, birthdate, uterine and early environment, schooling and family background. Non-identical twins act as a control, as they too share the same early environmental factors, but are no more genetically alike than any other siblings.

The researchers look for traits that show a greater similarity in identical twins than in non-identical twins - this greater similarity must indicate a shared genetic basis for the trait. This is expressed as a percentage known as 'heritability' - for example, asthma has a heritability of 60 per cent, and insulin-dependent diabetes mellitus 70 per cent, indicating a strong genetic predisposition.

Heritability and what it measures is often misunderstood. It is a population-based statistic, a measure of variance in a group of people. It is not a measure of the relative contribution of genes in an individual, or a measure of their risk. For example, a heritability estimate of 70 per cent for 'obesity' does not mean that 70 per cent of any one person's obesity is somehow due to his/her genes and the other 30 per cent to his/her environment or that they have a 70 per cent risk of disease.

Another common confusion is to equate the heritability of a trait and the number of genes that may influence it - as is often done in headlining the results of twin research. The discovery of a degree of heritability gives clues to how easy genes will be to uncover and is only a first step in understanding the complex relationship between genes and environment in health and disease.