Twin studies

If a disease runs in your family this may not mean that it is genetic. Families share some of their genes, but they also often share many of their lifestyle habits, their exposure to pollution due to where they live and work, and other social factors that affect health, such as unemployment. If you and your father both get heart disease, for example, this could be due to chance, shared environmental or lifestyle factors, or shared genes. Or it could be a complex combination of all these different things.

Scientists have been puzzled why genetic studies have not yet found many of the genes which they thought would explain a major part of why many common diseases run in families. They call this the missing heritability.

In 2006, GeneWatch's Director published a scientific paper which looks at how important genes are in explaining why diseases run in families. It concludes that much of the 'missing heritability' of common diseases might not exist. This argument has been made by many other scientists previously using several different methods: they all question how studies of twins are analysed based on assumptions originally made by the eugenicist Ronald Fisher in 1918. This webpage explains the argument and why it is important.

What is heritability?

Scientists think of the risk of a particular disease as a normal distribution (the famous 'bell curve'), with most people at average risk (in the middle of the curve) and a small number in the tails (at very high or low risk). The spread of the bell curve is measured by the variance (a measure of how wide or narrow the bell curve is). The bell curve can also be applied to characteristics such as height, body mass index (a measure of whether you are the right weight for your height), and performance in a wide variety of tests, including IQ tests or psychological tests.

The measured variance in any of these characteristics is thought to be explained by a combination of genetic and environmental factors. The proportion of the variance that is thought to be due to inherited genetic factors is known as the 'heritability'. This heritability is calculated from studies of twins. It is different in different populations because it depends on the environmental and genetic risks in that population.

Hunting for the missing heritability

Many common diseases have an estimated heritability of about 50% (although this varies quite a lot for different diseases and in different studies). However, only about 1-5% of the variance in risk for any common disease is caused by the genes that have been discovered so far. This is much less than was expected.

Recently, scientists have been surprised that they have not found this 'missing heritability' even by doing new genetic studies (known as 'genome wide association studies') in big groups of people.

Some geneticists argue that they will find more genetic factors by making these studies even bigger, and that these will explain the missing heritability. Others argue that bigger studies would be a waste of money and smaller, more detailed studies, which perhaps look at a person's whole genetic make-up, might find the missing genetic differences.

Does the missing heritability exist?

Other scientists have long criticised twin studies and think that some or all of this 'missing heritability' might not exist.

The scientific argument about twin studies has been going on ever since 1918, when the eugenicist Ronald Fisher devised the method used to do the calculations. The argument has not been resolved because twin studies alone cannot really show how much of the risk of a disease is inherited. In practice, to work out the 'heritability', Fisher made several assumptions which are still disputed. It is possible to show mathematically that, at best, these assumptions mean that Fisher's method gives an upper limit to the genetic component of the variance. So, if the assumptions used in twin studies are wrong, the environment or chance must play a bigger role than expected and the calculated 'heritability' will be too high. Genetic factors would then explain less of the differences between individuals, and be less important than expected in explaining why the disease or characteristic being studied runs in families.

When the bell curve is applied to human behaviours or characteristics such as intelligence, this scientific dispute is known as the 'nature versus nurture' debate - where nature means your genes and nurture means your family environment. Most scientists now believe that both genes and environment play a role in most diseases, but their relative importance is still disputed.

How do twin studies work?

Most twin studies involve comparing pairs of identical and non-identical twins. Identical (called 'monozygotic') twins come from the same egg, so they share all of their genes - this is why they look so similar. Non-identical ('dizygotic') twins come from two different eggs, even though they shared their mother's womb and were born at the same time. This means that they share only half their genes.

The idea behind twin studies is that if there is an inherited component to a particular disease, identical twins will be more likely to develop the same disease than non-identical twins are, because they share more of their genes. For common diseases, twin pairs won't always develop the same disease even if they are identical - this shows that these diseases are not completely determined by our genes.

To do a twin study, scientists must first ideally make sure that they have identified all the twins in a given population and then find out how many have developed the disease they want to study. In some countries, particularly in Scandanavia, all twins are registered at birth, which makes twin studies easier. As in all scientific studies, there can be some errors in the data collected from the twins. However, the main controversy is about how this data is interpreted.

If the likelihood of having a twin with the same disease is the same, whether your twin is identical or not, this means that the study has not found any inherited component. If the likelihood of developing the same disease is higher for identical twins than for non-identical twins, it is usually assumed that this difference is all due to inherited genetic factors.

However, there is some dispute about whether it is possible to tell how much - if any - of this difference is really due to genes. For example, if identical twins share more of the same environment or lifestyle factors than non-identical twins do, this could be an alternative explanation for any differences. If multiple genes and environmental factors interact together to cause the disease, this also makes explaining any differences less straightforward than is usually assumed, so the calculated 'heritability' could again be exaggerated.

Why is the debate important?

The debate is important whoever is right, for several different reasons.

Firstly, if the missing heritability does exist, people who buy commercial genetic tests are being told only a tiny part of their genetic risk for most common diseases. This means that the marketing is premature because new genes will be discovered that are likely to substantially change what they are being told today. Better tests might become available in future, once all or most of the missing heritability has been discovered.

However, an important caveat is that these tests might still not be useful to decide who needs to change their lifestyle or environment, or to take preventive medication. This does not just depend on how important genes are for predicting disease risk, but also on whether people at high genetic risk have more to gain than others by taking particular medical advice. Separate studies would be needed to find out if this was true.

On the other hand, if the missing heritability does not exist, then gene tests will never be very useful to predict the risk of common diseases in most people. This might mean that other factors - such as the environment, poverty, or diet - are much more important. Or, it might mean that such diseases are not very predictable: because where you are on the bell curve could occur mainly by chance. If this were the case, finding some genes might still help scientists to understand diseases, because small differences in risk can provide some useful clues about biology. But funding big studies to try to predict genetic risk would be a waste of money, and the growing genetic testing market would be built on a false premise.

What does GeneWatch think?

GeneWatch thinks that too much money has been spent on a race to find genes for common diseases without first looking more closely at the evidence from twins, families and environmental studies. The idea that testing your genetic make-up will be good for health has been widely promoted, even though we do not yet know if this is true, except in rare special cases.

Studies in twins and families give clues about how important genetic factors are, and how useful knowing your genetic risk might be, but it is important to interpret them correctly. It seems likely that chance, choice and complexity (interactions between different genetic and environmental factors) are all more important than Fisher assumed, but noone yet knows how important.

For some diseases, such as lung cancer, no twin study has ever found a significant heritable component. In this case the assumptions made do not matter because they do not change the answer. However, a lot of money has been spent in hunting for genes for lung cancer even though they cannot be important in influencing smokers' risk.

For other conditions, such as schizophrenia, it is already known that the twin study assumptions must be wrong and that something more complicated must be happening. Again, until this was discovered, many claims to have found the 'gene for schizophrenia' were made.

For diseases such as breast cancer, it is not yet possible to tell how much of the missing heritability exists. If the twin study assumptions are correct, shared genes, which have mostly not yet been discovered, would be the main explanation for why breast cancer runs in families. But if the twin studies are wrong, it could be that shared environmental factors are really more important.

↑ Top