Are you Hangry?...

I hesitate to go after studies that have a ridiculous enough premise to start with, but after critiquing GWAS studies, they are all starting to sound a bit ridiculous and I think we have a bit of a slippery slope, where there is this idea that a gene can be found for just about anything you can conceive of.  Do you like raisins?  Maybe there's a gene for that.  Do you think elephants are simply beautiful?  Perhaps it's a gene that makes you think that...
So here is a real study that was done to determine whether there is a gene for getting "hangry."  What is hangry, you might ask?  Well, of course it is the propensity to get angry when you get hungry.  Get it?  Hangry...   Here is a link to it so that you don't think I'm making it up
In any case, I'll take it on...
The study starts by justifying the use of the term, "hangry."

Being hungry doesn’t put anyone in a good mood, but have you ever felt hangry? “Hangry,” a mashup of hungry and angry, entered the popular lexicon in the 1990’s and was formally added to the Oxford English Dictionary in 2018.

I've never heard the term before this groundbreaking study, and I'm a bit surprised to hear that it was actually included in the dictionary.  I am pretty sure that it is not included in the DSM-V, but they then make an attempt to make it a psycho-physical  construct to justify it as a real disease? trait? disorder?

Several studies have found that low blood sugar is associated with personality changes, including higher aggression between spouses.  In another study comparing judicial decisions by time of day, judges handed down harsher sentencing decisions before mealtime breaks compared with after. Feel an argument coming on? Try a snack first. Up for trial? Bring the judge a sandwich.

heh heh...   At this point, it's worth noting that the author doesn't seem to know whether to turn this whole thing into a joke or take it as a serious study.  Eventually, he/she opts for the latter.  It seems that studies coming out almost need a silly hook like this in order to get a little attention.  Such was the case for a recent, poorly designed study for genetics and cognitive ability (my critique of which can be found here), that devolved into a media frenzy because the authors included something about a genetic association for poor eyesight and a need for glasses.  If you do a search on that study, that is mostly what you will see.  Lucky for them to have the diversion.

So here is how this study was done.  It comes from 23andMe, with their endless database to play with, so they asked  100,000 people:  “How often do you feel angry or irritable when you are hungry?” Okay, so even if you wanted to take a "study" like this seriously, this question could be answered in one of  five ways, "Always," "most of the time," "half of the time," "some of the time," and "never."  This is going to create some confusion.  You go from either/or to 5 possibilities when you are looking at "hangry".  Is there a different gene for being "hangry" half the time vs. all the time or some of the time, for example?  In any case, who gets to ask 100,000 people if they get hangry?

The results showed that 75% of people get hangry at least some of the time.  Another perhaps more interesting finding is that women were significantly more likely to get "hangry."  Do women have more hangry genes?  Probably not, so this might have been an interesting finding if they had stopped there and wondered what that was all about instead of reducing it to some genetic linkage.  For example, might women be more likely to get "hangry" because of the immense social pressure on them to remain thin?  Perhaps they are angry due to the battle they have to fight when feeling hungry and the pressure not to eat.  I might suggest to the authors that a path like this might be more fruitful for studying if they aren't married to genetic science.  In any case, it doesn't appear that their GWAS results were broken down by sex.  Here is the result:

To investigate the underlying biology of feeling angry and irritable when hungry, we ran a genome-wide association study among unrelated Europeans, controlling for age, sex, genetic ancestry, and the version of genotyping platform.Two regions reached genome-wide significance, which are indicated in red in Figure 4. This graph is called a Manhattan plot, and it shows how strong the associations are between answers to the hangry question and different genetic variants (represented by dots). Red dots show genetic variants that are “statistically significantly” associated with getting hangry — which means they’re most likely to be real associations (as opposed to random noise).

Sure, more likely to be real associations like 10,000 to 1 instead of 20,000 to 1, I suppose.  I will once again point out that these studies really need a randomized control and I would ask the authors to do so for any GWAS study like this so that we can determine how many "significant" associations we would get at random when we have a 100,000 person study like this.  Feel free to be the first to use my own devised method for this, which can be found here.  I venture to guess that you will come up with two variants, or so.

The next part of the study is an attempt to correlate the two variants they found to some kind of phenotype.  This is a common method of justifying the likely random results by connecting them to some sort of similar function.

One region was located in the vaccinia-related-kinase 2 gene (VRK2), and the other in the exoribonuclease 1 gene (ERI1). Both of these loci have previously been associated with a range of personality and neuropsychiatric conditions. Specifically, the VRK2 gene has been associated with schizophrenia, depression, epilepsy, and multiple sclerosis.

I'm not really sure if they think that associations with these things bolsters their case.  I'll simply point out that none of these alleged associations have been replicated, either.  The case for the ERI1 was slightly better:

In two publications using 23andMe data, ERI1 has been associated with subjective well-being, irritability, and neuroticism.

Slightly better, that is, if you ignore the contradictory nature of "subjective well-being" with the other traits.  In any case, again, none of these associations has ever really been replicated.

The silver lining with a study like this is that it highlights some of the weaknesses of all the GWAS studies related to mental/emotional/behavioral traits.  The very definition of the trait is debatable, the results are not compared to a randomized control, attempts are made to bolster the validity of the results by comparing them to marginally related traits,  and the hype created because no one really reads the study and just accepts it as fact.   And, of course, the results will never be replicated.

I think that, barring anyone with credence calling this stuff out (in other words, someone other than this one man, anarchist blog), we are going to see more and more of these studies with vaguer and more dubious traits, creating the illusion that we live in this world in which genes have been discovered for every imaginable trait.  In my view, this is creating  a real dumbing down of some of our smartest and brightest.