�鶹��Ѱ�����

Genome-wide association studies identify genetic overlap among disorders, providing evidence that their distinctions may be misleading

One major difference between psychiatric disorders and purely physical diseases is that the former are largely defined by their symptoms. Patient-reported symptoms are also closely associated with physical illnesses, but this is often accompanied by an awareness of underlying, biological causes, which can be confirmed by tests or scans.��

However, because the biological causes of psychiatric disorders have not been comprehensively explained, the boundaries between them can be blurry, especially considering that many people diagnosed with one disorder will be diagnosed with others, too.

As a step toward the long-term goal of explaining these causes, a large number of scientists from across the United States and the world conducted a study in Nature into the genetic associations of 14 disorders. This group includes first author��Andrew Grotzinger, a member of both the �鶹��Ѱ����� Department of Psychology and Neuroscience and the �鶹��Ѱ�����Institute for Behavioral Genetics.

�鶹��Ѱ�����Boulder scientist Andrew Grotzinger and his research colleagues studied how certain psychiatric disorders are influenced by genetic factors.

This study involved genome-wide association studies on the different disorders, followed by an analysis of the results for signs of genetic overlap (pleiotropy). The disorders ranged from attention-deficit/hyperactivity disorder (ADHD) to schizophrenia, and included several substance-use disorders. The study found that these disorders were influenced by five genetic factors, each of which was shared by two or more disorders.

Pleiotropy and genetic association

When multiple measurable and observable (phenotypic) traits are influenced by a gene or genetic variant, it is called pleiotropy. One example is the typical form of albinism, where a mutation of a single gene influences skin pigmentation, eye color and hair color by altering the production of melanin. The study uses the term pleiotropic loci, which refers to areas of chromosomes within which genes influencing multiple phenotypic traits can be found. In this case, those traits are different psychiatric disorders.

While evidence of pleiotropy can be seen in how some traits tend to vary together between individuals, like hair and eye color, it can only be proven by a thorough analysis of a large amount of genetic data. In this case, genome-wide association studies (GWAS) were done for each psychiatric disorder covered. A GWAS attempts to associate common genetic variants with traits by seeing if people who have the trait of interest also have a given genetic variant more often than would be expected based on chance alone.

For example, if people can have either gene A or gene B at a particular location (locus) in their DNA, a GWAS could determine if that genetic variation was associated with a given trait. If the study found that many people who did not have the trait had gene A and many people who did have the trait had gene B, it would conclude that gene B influenced the trait, even though there might be other factors contributing to it.

However, because there are so many genetic variants, and because scientists do not know which are relevant to begin with, a large number need to be studied. According to Grotzinger, this analysis happens across millions of genetic variants. A massive number of participants, both with and without the trait in question, is also necessary to have this statistical power to reliably study these associations. “What’s unique about this study is, in part, just how many people are involved, how many people we had DNA from,” Grotzinger explains.

That being said, data from some groups was more abundant. Only the European genetic ancestry group had enough data to perform analyses for all 14 psychiatric disorders. According to Grotzinger, analyses need to be performed separately by genetic ancestry group for statistical reasons, not because the genes themselves are very different but because the results may not apply equally to all people. “Initial evidence indicates that it may be more applicable for some disorders than others. So, schizophrenia is very much the same across ancestries, whereas depression is a little bit different,” he says. “The only reason we did European-like ancestry here was availability of data, and the hope would be that the next iteration of this study has greater diversity and representation.”

“Most people who are diagnosed with one psychiatric disorder are going to be diagnosed with multiple, and this has led some to theorize that there are genes that just increase your risk for everything,” says �鶹��Ѱ�����Boulder researcher Andrew Grotzinger. (Illustration: Wikimedia Commons)��

Genomic factors

After analyzing the results of the different GWAS, the researchers identified five genomic factors that explained the majority of the genetic variance of the individual disorders. Some of the variance is non-genetic (for example, resulting from different life experiences), but these genetic factors explain on average around two thirds of the common genetic variation caused by people having different genes. The factors were associated with 238 pleiotropic loci.

Factor 1 was most strongly associated with compulsive disorders, factor 2 was associated with schizophrenia and bipolar disorder, factor 3 was tied to neurodevelopmental disorders, factor 4 was connected to internalizing disorders, and factor 5 explained the genetic variance in substance use disorders

“Most people who are diagnosed with one psychiatric disorder are going to be diagnosed with multiple,” Grotzinger explains, “and this has led some to theorize that there are genes that just increase your risk for everything.” But “by and large, genes increase risks for subsets of disorders, and that’s what those factors are indexing.” All five factors showed high genetic correlation; however, there was evidence for even more overlap between the disorders covered by Factor 2 and Factor 4.

Besides genetic overlap within factor groups, the study also found weaker associations between disorders from different factors groups. That is in line with the theory that there are some genes that increase the risk for many psychiatric disorders, Grotzinger says, “and it seems like there are, but those probably map onto really general pathways, like tendency to experience distress. That’s not specific to OCD versus anxiety versus depression.” This overarching factor is called the “p factor” or general psychopathology factor, and is similar in concept to the “g factor” of general intelligence. In this study, the p factor was correlated with all five factors, especially Factor 4 (internalizing disorders).

Relatedly, when researchers analyzed genetic regions instead of the whole genome, they found 101 “hotspots”—regions that demonstrated significant pleiotropy. According to Grotzinger, these genetic regions include a larger group of genes than loci. “You can think of them as operating at a more local level, as opposed to the genome-wide level that examines the average percentage of genetic signal shared across the whole genome,” Grotzinger explains. The most pleiotropic region was on chromosome 11. Genes in this hotspot influence most of the psychiatric disorders, excluding all Factor 1 disorders, opioid- and nicotine-use disorder and autism.

Although these analyses close in on some of the genetic causes for psychiatric disorders, this knowledge cannot be used to diagnose or treat those disorders. This is because the disorders are influenced by thousands of genes, and scientists still do not know which specific genes are relevant, just the area of the genome they are in. “This chromosome 11 hotspot is a really interesting data point, but it is not going to help diagnose anyone,” Grotzinger says. “It is one piece of a 10,000-piece puzzle, at the end of the day. It’s baby steps.”

Diagnostic boundaries

“One thing people say is that our DNA does not read our diagnostic manual,” Grotzinger says. “Our DNA seems to confer risk in a way that transcends the boundaries that we describe in the Diagnostic and Statistical Manual.” In other words, psychiatric disorders are currently defined based on the way that symptoms tend to occur together rather than their biological or genetic causes. Of course, psychiatric disorders have non-genetic causes, but knowing what genes contribute to each disorder would help diagnose and treat them more effectively.

For example, “if you are someone who is diagnosed with multiple disorders,” Grotzinger says, “they may be more biologically similar than they are distinct. I think that increases the optimism for treatment, because you know that you are not dealing with four separate things. I do not think this is sufficient to argue for changing the diagnostic manual,” he continues, “but it is still a very important piece of evidence for considering whether or not to reconceptualize some of these disorders.

Depression and anxiety in particular are an example of disorders that are often diagnosed together, treated using similar methods (e.g., selective serotonin reuptake inhibitors), and appear nearly identical from a genetic perspective, according to Grotzinger. “It begs the question of whether or not we are calling something that is very similar different names. The metaphor I offer with this study is that, if you had a runny nose, a cough, and a sore throat, it would not be appropriate to go to the doctor and get a diagnosis for runny nose disorder, coughing disorder, and sore throat disorder.”

One implication of this, and a potential topic for future research, is that there are subtypes of disorders. While many of the disorders covered by this study overlap with each other, they do not all overlap completely with themselves. What is classified as depression, for example, may have different genetic causes in some cases. “You can have over 10,000 different symptom combinations, all of which meet the criteria for depression,” Grotzinger says. “So one question is, are there subtypes of disorders?” If enough research does support the reclassification of psychiatric disorders, this could involve both merging and splitting current disorders to most accurately reflect the underlying genetic risk factors.

Did you enjoy this article?����Passionate about psychology and neuroscience?��Show your support.

��