The Genetics of Anorexia Nervosa

Is it the culture of thinness, obsession with dieting or just bad mothering? When it comes to determining the causes of anorexia nervosa, the answer appears to be none of the above. Increasingly, the evidence is pointing to genetics playing an important role in predisposing individuals to anorexia nervosa. Among clinicians and researchers, the notion that genetic factors are important in the development of anorexia nervosa seems uncontested. In this short review, Dr. Cynthia Bulik and colleagues summarize some of the findings in the genetics of anorexia nervosa.

Currently (DSM-IV), to be diagnosed with anorexia nervosa, a patient must show:

• An inability to maintain normal weight (<85% of what is expected)
• Intense fear of weight gain and/or becoming fat, though underweight
• Obsession with body weight and shape, giving it undue importance in evaluating self-esteem/self-worth
• Amenorrhea (missing 3 or more consecutive periods)
• There are two AN-subtypes: restricting type (AN-R) and bingeing/purging type (AN-BP)

Bulik et al. outline some well-known personality characteristics and co-morbid psychiatric disorders commonly found in anorexia nervosa patients. AN patients tend to be perfectionists, exhibit a high degree of obsessionality, harm avoidance and anxiety. Many also struggle with major depression, anxiety disorders, and obsessive-compulsive disorder. These often pre-date the eating disorder and persist after “recovery”. Overall, AN has a fairly low lifetime prevalence of ~0.5%, but that figure depends a lot on the criteria used in the studies. The authors stress that when comes to anorexia, genetics and the environment both play important roles.

Furthermore, the precise influence of genes to environment will vary from person to person. For some patients, genetics may be the main factor, while for others, the environment might play a more important role in leading to anorexia. Naturally, this complicates the picture, in terms of research but also in terms of treatment. Perhaps – and this is just my hypothesis – this may explain why patients respond differently to various treatments. Perhaps a large part of it is due to differences in the degree to which genetic and environmental factors played a role in the development and maintenance of the patients eating disorder. Perhaps if we were able to determine the extent of, say, genetic influence in any given patient, we would be able to tailor treatment for that person. Admittedly, that seems too far into the future, and practically, too difficult.

There are several well-established risk factors for anorexia, which Bulik el al. list: female sex, history of childhood eating problems, sexual abuse, negative self-evaluation, other psychiatric disorders (anxiety, depression, OCD), family history. The authors mention important limitations of this (and other similar) data: (1) the rare occurrence of this disorder makes prospective studies difficult and (2) it is often difficult to differentiate between a risk factor and an early symptom, such as  in the case of dieting.

Family studies have shown that first-degree relatives have a six to ten times greater lifetime risk of developing AN than relatives of healthy controls (Bulik et al. report 10, but I’ve seen six reported as well). They also have a greater likelihood of suffering from disordered eating (other EDs, sub-threshold eating disorders, meaning they almost but don’t quite meet the diagnostic criteria for anorexia or bulimia nervosa) versus specifically AN. This reflects that many anorexia nervosa patients later fulfill the diagnosis for bulimia nervosa (a topic I will cover in an upcoming post).

Research from twin studies has shown that anorexia nervosa is significantly heritable. This means that additive genetic traits that can be passed on from parents to offspring account for a large percentage of the traits seen in anorexia nervosa. The caveat is that often these values are taken from studies of mostly European populations. It is possible that the influence of genetic versus environmental factors varies in different cultures and racial backgrounds. There are some important things to keep in mind when thinking about heritability, as outlined in the link above:

Heritability and environmentability are abstract concepts. No matter what the numbers are, heritability estimates tell us nothing about the specific genes that contribute to a trait…

Heritability and environmentability are population concepts. They tell us nothing about an individual. A heritability of .40 informs us that, on average, about 40% of the individual differences that we observe in, say, shyness may in some way be attributable to genetic individual difference. It does NOT mean that 40% of any person’s shyness is due to his/her genes and the other 60% is due to his/her environment.

Heritability depends on the range of typical environments in the population that is studied. If the environment of the population is fairly uniform, then heritability may be high, but if the range of environmental differences is very large, then heritability may be low..

Heritability is no cause for therapeutic nihilism. Because heritability depends on the range of typical environments in the population studied, it tells us little about the extreme environmental interventions utilized in some therapies.

What about the genes that underlie this highly heritable disorder? While family and twin studies have shown that genetics most certainly plays a large role in the development of eating disorders, in particular anorexia nervosa, they do not shed light on what specific genes and gene variants (alleles) contribute to its development. Human molecular genetic studies, such as genome-wide linkage studies and association studies have attempted to find genes responsible for increasing the vulnerability to AN.

Genome-wide linkage studies identify regions of the genome that may contain genes that either increase or decrease the risk of anorexia nervosa. These studies are done without any prior assumptions about specific genes. Bulik et al. summarize some findings of studies with large sample groups (reducing the likelihood of false positives due to random chance). These findings should be viewed as preliminary data and as pointed out in the paper, many of the linked regions cover very large portions of the genome. Overall, chromosome 1 shows linkage with restricting-subtype of anorexia; incorporating measures of “drive for thinness” and “obsessionality” increases the linkage to chromosomes 2 and 13.

Association studies compare anorexia nervosa patients with a control groups for variation in specific candidate genes that are thought to be involved, or could be involved in AN based on prior knowledge of pathophysiology. Most studies in AN have focused on serotonin and dopamine neurotransmitter pathways. Bulik et al. are quick to point out that a lot of initial reports are not replicated in later studies, and focus on the few studies that are large enough to provide meaningful data. Before proceeding to summarize the findings mentioned in the review, I wanted to provide some links that explain synaptic communication, neurotransmitters, and  neurotransmitter receptors

• Serotonin, is important for regulating body weight and eating behaviours, among many other roles, and has been implicated in the development of EDs. Among the studies surveyed, two large studies showed a significant association with one serotonin gene (1D, which resides in the previously mentioned linked chromosome 1 region). No association was found for the serotonin 2A receptor in another large study.
• Increased dopamine activity has been implicated in food repulsion, hyperactivity, weight loss, absence of menstrual cycles and obsessive-compulsive disorder, and is known mediate reward states. Thus far, preliminary studies point to the dopamine receptor D2 as potentially playing a significant role in anorexia nervosa, while a protein that breaks down neurotransmitters including dopamine, called COMT, is unlikely to play an important role.
• Neuropeptides, like ghrelin, hypocretin receptor and the opioid receptor delta have also been studied. Thus far, research is pointing to the involvement of the opioid receptor delta (ORD1) in both subtypes of anorexia nervosa (the gene also resides in the region of chromosome 1 that has been implicated by genome-wide linkage studies).
• Other molecules, like BDNF (brain-derived neurotrophic factor) may also play a role in appetite suppression and weight loss, but again, while promising, these findings need to be replicated.

One of the main things to take away from this review is that while the role of genetics in vulnerability to AN is undisputed in the research community, the causative genes and/or gene variants are not yet clear. There are, however, many candidates: genes involved in some way in regulating appetite, eating behaviours, anxiety, depression, pleasure and reward, and other traits commonly found in AN patients (obsessionality, perfectionism, etc…), but, many of the present findings need to be replicated in larger samples.

The authors point out that many studies attempting to pinpoint the genes involved in the etiology of anorexia nervosa suffer from being underpowered, increasing the likelihood of type II error (in other words, failing to identify genes and being overly skeptical) or multiple testing, increasing the likelihood of type 1 error (in other words, a false positive, or identifying genes that are actually not important in AN.) Bulik et al. suggest that along with increasing the size of the studies, focusing on more homogenous populations (such as more stringent inclusion criteria) and by incorporating behaviours associated with anorexia nervosa may aid in identifying genes that contribute to the development of AN.

References

Bulik, C., Slof-Op’t Landt, M., van Furth, E., & Sullivan, P. (2007). The Genetics of Anorexia Nervosa Annual Review of Nutrition, 27 (1), 263-275 DOI: 10.1146/annurev.nutr.27.061406.093713