Genome-wide association study identifies eight risk loci and implicates metabo-psychiatric origins for anorexia nervosa

I know people who have or have had anorexia who would absolutely say that their thoughts and beliefs are highly involved in anorexia.
- one who doesn't eat because she believes that she doesn't deserve to eat
- two who became anorexic due to seeking something to control during a time of turmoil in their lives.

(I am oversimplifying their stories here, because I don't have the right to describe in detail their state of mind.)

They would absolutely say that they needed what you call "psychiatric rubbish" to help them.

 

And I'm not denying that someone suffering from anorexia will have their thoughts and beliefs heavily influenced by their condition, and if the people you know felt some benefit from seeking psychiatric help then I'm glad for them.

What I'm calling psychiatric rubbish is the drivel that proclaims that psychiatric "explanations" are sufficient to explain diseases that we currently can't explain biomedically. The repeating pattern that we see is that more and more conditions once thought of as psychiatric are found to have a solid biomedical basis for them, and my prediction is that this will be the case for anorexia.

 

In a documentary I saw about anorexia nervosa, it seemed to me like the patients were just parroting what their doctors/therapists told them about the cause of their anorexia. Personality, trauma, etc.

Which is, of course, a similar narrative to what the BPS crowd tell ME/CFS patients.

Wouldn't the affect of calorie deprivation on cognitive function, make patients more susceptible to brainwashing?

I found it particularly chilling watching the therapist being interviewed too, as they clearly believed their model of treatment to be gold standard, even though the patients would be in and out of the clinic their whole lives. The bps model didn't cure them, but it did guarantee income for the clinic.

I remember thinking that the chances of anorexia ever escaping the stranglehold of the BPS model was pretty much zero, so it's very refreshing to see this study. Thanks for posting!

 

I went through the experience of anorexia but didn't get to the severe skin and bone situation but I did get very thin. I was 15 years old.

From my own experience something very subtle happens to the brain and very quickly too. It only took 3 months to lose the weight by eating healthy food, calculating the calorie intake and exercising more. But there was a huge trap for me and obviously others too who go on to get anorexia and that is I lost perception of where I was in the whole process.

I seemed to not know where or when to stop. I just keep continuing on and I was still seeing myself as fat in the mirror when in fact I was very thin. Fortunately a friend helped me to recognize that I was very thin and needed to put on weight. But it took me more than 2 years to stop counting calories and stop the whole process I had started. It was hard to break the pattern. I see the trap because I have experienced it. It's a real worry/concern especially for young people who start out on these diets to lose fat. Something happens to brain for those who are susceptible. It was easy to lose weight but there is such a huge risk of not being able to stop the monitoring of food intake.

I don't know enough science wise to know what is causing this change. But I agree with you that help is needed to get the right perception back, very difficult and takes a long time. I think for some it never comes back because some die trapped in this condition.

 

One of the Canadian Consensus Criteria is "marked weight change."

In the first couple of months of the illness I lost about 15% of my body weight. I just seemed to have lost my appetite, perhaps influenced by the fact that eating often resulted in a "pounding heart" within just minutes. My heart rate increased somewhat, but the beat was very forceful. I'd liken it more to a food poisoning reaction than to something related to sugar.

The odd thing was that it came on so fast. My glucose tolerance tests were all normal - except that my heart pounded during them, too.

Fortunately, this reaction lessened over time.

 

Hello Esther

The headline is imo problematic and dualist. If systemic illness causes mainly psycho-behavioural problems , then the illness can still be deemed psychiatric. In itself the term "psychiatric" does not exclude systemic physical. It is only when, in the pursuit of psychiatric categorisation, the systemic physical is ignored (esp. on dogmatic grounds), is fallaciously "explained" through purely psychological mechanisms or misrepresented as sth which it is not, or when poor trials are presented as good ones and issues of subgroups ignored, that psychiatry oversteps the mark. Paying attention to, finding actual mechanisms and representing truly may of course remove the condition from the province of psychiatry. We know all about this, but the headline is not thought through and can be represented as showing a naive and in in a sense anti-psychiatry attitude which PBSers then use to "demonstrate" how their critics are all misleading/misled fools.....Guardian (and BBC) should do better than to promote false notions as to what psychiatry is.

(There is, of course, the pragmatic problem of what happens when a system is shown to be involved, in which your average psychiatrist might have no expertise, which a-n may now represent).

As to a-n, I think people are different and I would keep all channels open to help sufferers in a way that best suits them as individuals. I bet epigenetics, evolutionary biology (starving=dearth= need to starve more), starvation induced low levels of appetite inducing nutrients e.g. zinc, biotin and other matters play a role and I bet culture does too.

 

Fully agree with this. Between the media headlines and the confusion between anorexia and anorexia nervosa this coverage and some reactions to it have not been good and pretty insulting at times.

I’m someone who is naturally thin, rarely hungry and that has to make a conscious effort to eat to keep my BMI in the green (but with no body image problems) so I find the research interesting in itself.

 

Yes, which is a bit amusing given Ian Sample's smears against ME/CFS patients who dare point out the problems with Larun's work.

 

Interesting tangent:

https://www.theguardian.com/lifeand...repeat-the-shocking-rise-of-muscle-dysmorphia

 

Genome-wide association study identifies eight risk loci and implicates metabo-psychiatric origins for anorexia nervosa

This is a good example of the potential of genome-wide association studies to transform the understanding of an illness — especially one without good treatments.

From the conclusion:

Low BMI has traditionally been viewed as a consequence of the psychological features of anorexia nervosa (that is, drive for thinness and body dissatisfaction). This perspective has failed to yield interventions that reliably lead to sustained weight gain and psychological recovery7. Fundamental metabolic dysregulation may contribute to the exceptional difficulty that individuals with anorexia nervosa have in maintaining a healthy BMI (even after therapeutic renourishment). Our results encourage consideration of both metabolic and psychological drivers of anorexia nervosa when exploring new avenues for treating this frequently lethal illness.​

Abstract:

These results further encourage a reconceptualization of anorexia nervosa as a metabo-psychiatric disorder. Elucidating the metabolic component is a critical direction for future research, and paying attention to both psychiatric and metabolic components may be key to improving outcomes.​

Brief commentary on the study​

The study expands on a 2017 study with 3,500 cases. The new study pulled in a lot of other anorexia nervosa GWAS to give almost 17,000 cases and 55,000 controls. Given that anorexia nervosa is highly heritable (50 to 60%, according to twin studies), where genetic defects will be bigger, that's a big sample.

There isn't reliable data on ME/CFS heritability, but the estimates that exist put it at around 20%.

The study identified eight different risk loci (chromosomal locations), each with multiple hits.

As with all GWAS, the significant SNPs are effectively markers, flagging up sections of DNA that increase or decrease disease risk. However, because DNA sequences that are close together tend to be inherited together, GWAS aloe can't be precise in identifying exactly which genes are involved.

The authors wnet on to use a range of techniques to identify the genes most likely to be causally involved, identifying 133 candidate genes. A handful of genes looked to be particularly strong candidates.

Instead, the main way the study tried to identify what underlying genes are involved was to see which of the SNPs that were significant in this GWAS were also significant in GWAS for other diseases and traits.

First they found that SNPs that were risk factors for AN are also risk factors for psychological illnesses including obsessive-compulsive disorder (most strongly), anxiety and schizophrenia. They noted that these genetic patterns reflected comorbidities seeing in clinical and epidemiological studies.

The authors also found that quite a few of the anorexia nervosa SNPs predicted "good" metabolic outcomes, protecting against insulin resistance and type II diabetes, for instance.

Similarly, anorexia nervosa SNPs also predit low body fat percentage, low-fat mass, small waist conference, low BMI and reduced risk of obesity.

Note that genetic risk factors exist before anorexia started, so it is not a case of low BMI simply being a consequence. Instead, being genetically prone to having a low BMI is a risk factor for anorexia nervosa. Even so, the authors controlled for the effects of these BMI SNPs and found that there was still an independent association of anorexia nervosa with genetic variants linked to metabolism.

Systems biology approaches looking at cell types showed higher SNP concentration in the central nervous system. Interestingly, further analysis suggested that the neurons involved include those linked to feeding behaviours, including food motivation and reward. But this is a tentative observation.

 

That’s very interesting and I wonder if they had any more success treating AN in the past. I also tought that @Woolie made a good point in reply:

"What's changed since 1960 is the high social desirability of female thinness, I'd expect that nowadays, there'd be proportionally more cases where this type of reward played a role - at least in triggering the initial behaviours that led to the problem."

To be fair, the evidence from this study is of both psychological and metabolic causal factors playing a role. Granted, psychological approaches have a poor track record but that doesn't automatically me mean that metabolic approaches will work. It may even be that a combined approach will work, or perhaps it will depend on the particular patient.

As the conclusion argues, because the negative correlation is with genetic risk of low BMI (which exists before the low BMI of anorexia [it's pathognomonic, says the paper!]) it is surprising, suggesting that low BMI might play a causal role.

I would say that the study tells us a great deal, though, as you say, it doesn't pinpoint the precise causes. What it does is highlight that metabolic factors probably play a causal role. That is quite a turnaround for an illness like this. GWAS can never provide final answers, but they can show researchers very promising areas to explore.

 

Didn't know that. Thanks.

 

Interesting study.

I think Figure 2 shows the correlation between the SNPs that were significant in the anorexia cohort and those associated with other phenomena and diagnoses. If I've got that right, then what the Figure is showing is other observable features and diagnoses that share some of same SNPs hat are implicated in anorexia nervosa in this study.

Anyone know if I’ve got this right?

If so, then the approach is quite an interesting way of attacking the question of causation in anorexia nervosa (AN).

What I read from the Figure is that the SNPs that were significant for this (AN) cohort and also heavily implicated in OCD. This is interesting, as it suggests people who develop AN might share some temperamental characteristics that are also common in OCD. Maybe the same intense, focussed type of temperament.

There are a lot of other associations in the figure that don’t tell us much about causation. It could be that those temperamental characteristics that make people good at weight loss (a very focussed style) are also useful for achieving well in school/university.

As @strategist says, the metabolic correlations seem unlikeky to be causes. AN (both fully diagnosed and subclinical variants) could be a major cause of low BMI, and high levels of physical exercise, and if so, the same SNPs would obviously be implicated in both. Ditto for insulin resistance, variables associated with diabetes, cholesterol, etc.

Some here have said this evidence demonstrates that AN is not “psychological”. I think it shows that certain temperamental characteristics – that without question have a biological underpinning – heighten your likelihood of getting AN. But its not the full story. Any complete explanation of AN has to address the very specific role played by thinness as both a motivator and a source of reward. An explanation phrased entirely at the biological level would lose this important feature.

Anorexia is, to me, a really good example of a condition where a psychological explanation is not only supported by some real positive evidence, but that explanation plays a crucial role in understanding and treating the disease.

I don’t mean to say that anyone who has trouble keeping weight on has anorexia nervosa (it can happen for all sorts of other reasons, and is often misdiagnosed). Just that there is definitely a condition where the extreme weight loss is connected with the explicit goal of thinness.

 

You have.

Interestingly, there was no correlation between anorexia nervosa and SNPs associated with IQ.

From the evidence presented we can't tell, but my reading is that it is more likely than unlikely — and further analysis (and research!) could sort this out.


Diagnosed cases of anorexia nervosa will be excluded from the healthy controls, though these could include subclinical cases. Whether it's a significant proportion or not, there's no way of knowing. I might be more typical of low BMI cases in that I like food, eat well and think I am probably too thin.


Also, the study found that the association with metabolic, lipid and anthropometric traits was independent of common variants associated with BMI.

This might also be relevant:​

Second, we investigated bidirectional causality using generalized summary data-based Mendelian randomization (GSMR)18. This indicated a significant bidirectional causal relationship such that anorexia nervosa risk-increasing alleles may increase the risk for low BMI, and BMI-lowering alleles may increase the risk of anorexia nervosa (Supplementary Table 16). It is important to note that having only eight genome-wide significant loci for anorexia nervosa render this analysis marginally powered in the direction of anorexia nervosa to BMI, although this analysis is well-powered in the direction of BMI to anorexia nervosa.​

I am assuming the "not causal" argument is that the SNPs are not picking up genetic differences due to metabolic et cetera characteristics, but some other aspects of anorexia (eg that leads to change in physical activity or to low calorie/restricted diets). Please shout if I have got the wrong end of the stick.


There is not enough information in the paper to be sure about causality. But there are reasons to think that at least some of the SNPs involved signify changes to/affecting metabolic genes.


I am pretty sure that HDL cholesterol levels are strongly influenced by genetic factors (probably more so than diet).


Similarly, while obesity is a major driver of type II diabetes, and many SNPs are risk factors for obesity, many SNPs also independently predict T 2D, and quite a few of these have been linked to causal genes, such as those affecting insulin secretion in pancreatic islet cells and various insulin action-associated loci in monocytes, adipocytes and hepatocytes. If the relevant SNPs were involved here (or cholesteryl, or any of the metabolic factors) it would point to a metabolic causal role in anorexia nervosa.


Anyway, the only way to establish the role (or not) of metabolic factors in anorexia nervosa is further research. It certainly seems a route worth pursuing.

 

Thanks, @Simon. I found it really interesting thinking on the value of GWAS more generally.

I understand that people like yourself might be accidentally misdiagnosed as having anorexia. But you don't have it, no way. Real anorexia nervosa has nothing to do with just being thin. Its about hard work and intense vigilance. Its very scary for a person with anorexia to imagine what would happen if they let go for a bit.

I discussed our different views with some others. One suggested you could reconcile them like this:

- anorexia could be seen as a form of OCD in which the goals/behaviours are centered on becoming or remaining lean (this fitd with the GWAS findings).
- the content of any sort of OCD-like behaviour is determined by the person's values and their capacities to gain feelings of agency from that behaviour
- people with the SNP's associated with low BMI (etc.) are very good at becoming and remaining thin
- therefore, these people are more likely to make becoming/remaining lean a central goal and focus, and to gain some sense of agency from those activites.

 

Sort of like inverse obesity with grehlin and leptin signalling gone awry?

 

That's pretty much what I was thinking too. So like a high jumper, its "easy" for them to jump high - relative to shorter people - but then there's a lot of work and effort required to actually win a competition.

No, I think such explanations fail to account for the main features (which is that the behaviours in AN are very much goal-oriented require incredible self-control).

But them @Amw66, some other biological changes might occur when BMI goes below a certain level, and these might contribute to the behaviours.

 

My husband's family are thin. His female cousins are twig like naturally.

I used to think I was terribly greedy because if we missed a meal I would be ravenous but he would not be bothered. Then I realised that he just was not hungry. There was no will power involved. He eats what he wants when he wants (not nearly as healthily as me) and has put on a bit over the years but not to a great extent.

We think that his liver releases glucose easily or some other biological process.