WashU researchers use genetics to find psychopathology risks

When trying to understand how genetic influences factor into youth behavior, researchers at Washington University in St. Louis have taken the “big trawl” approach, casting their net wide to pull in all the measured traits, behaviors and environments that make up who we are and examine associations with the genetic building blocks comprising risk for mental health problems.

This cutting-edge methodology has turned up valuable new insights into factors related to psychopathological genetic risk, such as stressful life events and screen time. Although the results, published in Nature Mental Health, are unable to say if one causes the other, the findings provide promising leads to understand the nature of psychiatric disorders emerging during adolescence.

“We’re catching all the little fish here,” said Nicole Karcher, assistant professor of psychiatry at WashU Medicine, likening their genetic screening tools to trawling the ocean.

“But now we get to wade through the fish that we caught, and future steps include understanding the extent to which these are meaningful in terms of their ability to reduce risk for mental health concerns.”

An innovative approach to “catching” risk factors

Much of what we know about links between the genome and behavior come from Genome-wide Associations Studies (GWAS), which identify links between specific genetic variants across the genome and a feature of interest, also known as a phenotype. Phenotypes can range from physical characteristics to psychiatric disorders (e.g., depression, anxiety).

Many behavioral disorders are correlated at the genetic level. Results from a GWAS scanning for genetic links to depression, therefore, may also reflect genetic associations with frequently co-occurring conditions such as anxiety.

“We know that one behavioral variable is not going to be the only association with genetic risk, so we were interested in taking a more agnostic, data-driven approach to the wealth of information that is available in large datasets,” said Karcher.

Doing so would hopefully identify not only expected associations between genetic risk and psychiatric symptoms, but also potential novel associations that could improve insight into how psychiatric disorder risk may unfold.

So senior author Karcher and first author Sarah Paul, a graduate student in Ryan Bogdan’s Behavioral Research and Imaging Neurogenetics Laboratory at Art & Sciences, ran what’s called a phenome-wide association study (PheWAS) that inverts the GWAS.

Rather than starting with the psychiatric condition and looking for associated genetic variants, their PheWAS started with genetic variants known to be linked with mental health disorders and examined their relationship to hundreds of measured variables reflecting behavior, symptoms, environments, health problems and other phenotypes. They included approximately 1,300 to 1,700 phenotypes in total from the Adolescent Brain Cognitive Development (ABCD) Study.

“We took a pretty broad approach,” said Paul, describing different phenotypes as “anything from impulse control problems and problematic behavior or psychotic-like experiences to screen time, to how much caffeine they consumed.”

Think of it as fishing with a big net.

That means they want to identify associations between genetic predisposition and potentially modifiable risk factors that can be potentially addressed before the onset of psychopathology, Bogdan, the Dean’s Distinguished Professor of Psychological & Brain Sciences in Arts & Sciences, said.

What they caught

The results of the PheWAS show some surprises and confirm some of what they already know about genetic risks and behaviors that are associated with mental health disorders in youth.

The WashU researchers took 11 GWAS and created four broad genetic risk factors, or polygenic scores: neurodevelopmental, internalizing (e.g., depression, anxiety), compulsive and psychotic. Below are some of the associations they found in those categories:

*Genetic risk for neurodevelopmental psychopathology (predominantly ADHD and Autism Spectrum Disorder, as well as Major Depressive Disorder and problematic alcohol use) was associated with some 190 phenotypes including inattention and impulsivity issues, as well as total screen time, sleep problems and psychotic-like experiences. Even environmental conditions like neighborhood crime rates and lower parental monitoring are associated with neurodevelopmental genetic risk.

*Genetic risk for internalizing behavior (Major Depressive Disorder, Generalized Anxiety Disorder, PTSD, as well as problematic alcohol use) were broadly associated with some 120 phenotypes such as depression, stressful life events, psychotic-like experiences and screen time. 

*Psychotic risk (predominantly Schizophrenia and Bipolar Disorder) had few phenotype associations aside from lower school involvement and more consumption of energy drinks.

Karcher said it was somewhat surprising that “genetic liability” for mental health concerns may manifest through potentially modifiable behaviors in childhood and early adolescence.

The research sorted hundreds and hundreds of variables potentially associated with genetic risk, and the results highlighted several associations, including the association between neurodevelopmental genetic risk and screentime, she added. 

“The PheWAS was able to point out these pockets of associations that may not have been found otherwise,” she said.

One such pocket was the association between psychotic disorder genetic risk and energy drink consumption. These studies are looking at correlation, not causation, so they cannot conclude that energy drink consumption causes psychotic disorders. It could be that there are genetic components that make these individuals more at risk for psychotic disorders and those same components might make these individuals more likely to consume caffeinated beverages.

A similar phenomenon could be a play with the strong association between screen time and neurodevelopmental risk.

The point of the PheWAS is not to sort those details of causation but get pointed in the right direction with “a 20,000-foot view of the associations,” Karcher said.

Time will tell as the ABCD kids get older and genomic databases get more diverse.

“Following these youth into early adulthood will help better inform how genetic risk is associated with things like screen time, psychopathology, symptoms, and sleep over the course of adolescence into early adulthood,” Paul said. “That will help paint a clearer picture of how these links between your overall genetic risk and your behavior and traits change or don’t change over time.”

Overall, the present work illustrates how the PheWAS technique can be used to identify potential targets for future prevention and early intervention strategies, with this study identifying several potentially modifiable targets, such as screen time and caffeinated beverage consumption, that could represent early “catches” for reducing risk for developing mental health concerns.

Previous genome-wide studies of psychiatric diagnoses/phenotypes make use of data from individuals most genetically similar to European reference populations, with limited well-powered GWAS available for other populations in the world. So, one major limitation of this study was that because the GWAS predominantly used data from European reference populations, only ABCD data from individuals with European ancestry could be used in the PheWAS.

“That really limits the generalizability of these findings,” Paul said, “but as more GWAS become available in individuals genetically similar to other reference populations, and as more sophisticated polygenic score approaches are developed, we should be able to expand the study population to be more inclusive.”

 

################

Paul SE, Colbert SMC, Gorelik AJ, Hansen IS, Nagella I, Blaydon L, Hornstein A, Johnson EC, Hatoum AS, Baranger DAA, Elsayed NM, Barch DM, Bogdan R, Karcher NR. Phenome-wide Investigation of Behavioral, Environmental, and Neural Associations with Cross-Disorder Genetic Liability in Youth of European Ancestry. Nat. Mental Health (2024). https://doi.org/10.1038/s44220-024-00313-2

 

Data for this study were provided by the Adolescent Brain Cognitive Development (ABCD) study , which was funded by awards U01DA041022, U01DA041025, U01DA041028, U01DA041048, U01DA041089, U01DA041093, U01DA041106, U01DA041117, U01DA041120, U01DA041134, U01DA041148, U01DA041156, U01DA041174, U24DA041123, and U24DA041147 from the NIH and additional federal partners (https://abcdstudy.org/federal-partners.html). This study was supported by R01 DA054750. Authors received funding support from NIH: SEP was supported by F31AA029934. NRK was supported by K23MH12179201. AJG was supported by NSF DGE-213989. ECJ was supported by K01DA051759. ASH was supported by K01AA030083. DMB (R01-MH113883; R01-MH066031; U01-MH109589; U01-A005020803; R01-MH090786), RB (R01-DA054750, R01-AG045231, R01-AG061162, R21-AA027827, R01-DA046224, U01-DA055367). NME was supported by NSF DGE-1745038.