New genetic clue in 'puzzle' of anorexia

Monday September 16 2013

“Anorexia could have genetic link,” The Daily Telegraph reports.

The headline might lead you to think that a role for genes in anorexia is a completely new finding. However, studies in twins and families affected by the condition have already suggested such a link.

The news highlights a study in which researchers found that some variants in and around a gene called EPHX2 are more common in people with anorexia nervosa.

They compared 152 genes in more than 1,000 women with anorexia and almost 1,500 women who did not have the condition.

The EPHX2 gene is involved in cholesterol function in the body, and has not previously been linked with anorexia.

In a complex disease like anorexia, multiple genes are likely to play a role, as well as environmental factors. This makes it difficult to identify the genes involved.

This study gives us only a small piece in a large and complex puzzle.   

Where did the story come from?

The study was carried out by researchers from The Scripps Translational Science Institute in the US, and other research centres in the US, Canada, Italy, and the UK. It was funded by the Price Foundation, and the authors also received support from grants from the US National Institutes for Health and other university and charitable organisations.

The study was published in the peer-reviewed journal Molecular Psychiatry.

The Telegraph’s headline implies that a genetic contribution to anorexia had not been suspected previously, which is not the case. Similarly, the Mail Online implies that previous thinking was that anorexia is purely down to social pressures, which is also incorrect.

Many experts believe that the condition is caused by a combination of psychological, environmental and biological factors (including genetics), which lead to a destructive cycle of behaviour.

What kind of research was this?

This was a case-control study looking at whether variations in specific genes might be linked to anorexia. Genetic factors are thought to play a role in people’s risk of developing anorexia, as well as environmental factors.

However, it is likely that multiple genes play a role, and different genes might be playing a role in different people, and that makes it hard to identify these genes. So far, studies have not been able to provide robust evidence about which genes are involved.

In the current study, researchers wanted to focus on a number of genes which they thought might play a role in anorexia, based on what is known about what these genes do. This approach is called a “candidate gene” approach, as the genes are “candidates” for contributing to disease risk. This is one of a number of standard ways to study the genetic basis of a condition.

What did the research involve?

The researchers studied DNA from 1,205 people with anorexia (cases) and 1,948 people without the condition (controls). They looked at the sequence of 152 candidate genes that might play a role in causing anorexia, to try and identify variations in these genes that were more common in cases than controls.

The researchers initially started with 262 white European women with early-onset severe anorexia and 80 matched controls who were not underweight (matched for ethnicity, age and where they were recruited). The cases:

  • had been clinically diagnosed with a history of restricting type anorexia (where a person limits their calorie intake) with or without purging (vomiting) and were an average age of 14 years when they first experienced these symptoms
  • had a body mass index (BMI) of 15 or less during their lifetime, and had an assessment age of 19 years or older (to give evidence of a persisting disease course)

Women who reported regular binge eating were not included in this sample, to ensure women in the sample were all as similar as possible (to try and ensure that participants had anorexia and not another diagnosed eating disorder such as bulimia nervosa). Tightly defining the study population helps researchers to identify genes that might contribute to causing the condition. This is because different eating disorders might be caused by different genes, and this makes teasing out these genes from a mixed sample difficult.

After initially looking at the candidate genes in these women, they went on to test their findings in 500 anorexia cases and 500 controls, and also data on a further 444 cases of anorexia or eating disturbances and 1,146 controls from previous studies. All of these participants were of European ancestry. This method of looking for associations in one sample and then confirming what you find in another sample (called a “replication sample”) is a standard way of increasing confidence that the genes identified are truly associated with the disease.

For some of the cases and controls in previous studies, the researchers had data on BMI and metabolic measurements, such as cholesterol, taken over time. This allowed them to look at whether the variants they identified were related to these changes. They also looked at where the genes they identified as being associated with anorexia were active in the brain.

What were the basic results?

The researchers identified just over 8,000 single nucleotide differences and other small differences in the 152 genes they were interested in, in different individuals.

The two variants which showed the strongest association with anorexia (those that showed the greatest difference between cases and controls in how common they were) were in the Estrogen Receptor Beta gene (ESR2). Previous studies have suggested that oestrogen and oestrogen receptors might play a role in anorexia.

They also used another form of analysis, which looked at groups of variants together, and in this analysis variants in two genes – ITPR3 and EPHX2 – showed the strongest association with anorexia. When they tested these variants in their replication samples, the variants in EPHX2 and ESR2 still showed evidence of an association with anorexia.

The researchers then focused on the EPHX2 gene, as its association with anorexia has not been described previously.

The gene is involved in cholesterol function, and although this may seem counterintuitive, people with anorexia often show high cholesterol levels that return to normal once they return to normal weight.

The researchers found that one of the variants in EPHX2 was related to how BMI and cholesterol levels changed over time. The EPHX2 variants also showed some evidence of association with depressive and anxiety symptoms in women with anorexia, and appeared to affect how BMI related to depressive symptoms. The EPHX2 gene was found to be active in some parts of the brain relating to feeding behaviours, anxiety, and depression.

How did the researchers interpret the results?

The researchers concluded that they had identified “a novel association of gene variants within EPHX2 to susceptibility to [anorexia nervosa] and provide a foundation for future study of this important yet poorly understood condition”.


The current study has identified an association between a gene called EPHX2 and anorexia in women, in what is reported to be the largest DNA sequencing study in anorexia to date. Even so, the researchers note that it will be essential to test their findings in a larger sample for confirmation.

The exact causes of anorexia are not known, although both genetic and environmental factors are thought to play a role. In complex diseases such as anorexia, many genes may potentially play a role, and different genes may play a role in slightly different forms of the condition. This makes identifying the genes involved very challenging.

A strength of the current study from a genetic perspective was it looked at a very well-defined group of women with anorexia However, the downside of this type of study design it that it does limit the extent to which findings can be extended to other groups. For example, it only looked at women and those who did not binge, so results may not apply to men or people who may have another eating disorder diagnosis, such as bulimia.

As the study authors note, more work is needed to confirm a role of this gene in anorexia. There will also be other genes playing a role, so this finding is only one small piece in an evolving puzzle.

Analysis by Bazian
Edited by NHS Choices