How behavior and biology combine to drive addiction

A massive genomic study uncovers how shared behavioral traits and substance-specific biology together shape addiction risk, offering new directions for prediction, prevention, and treatment.

Study: Multivariate genetic analyses of 2.2 million individuals reveal broad and substance-specific pathways of addiction risk. Image credit: OlgaKan/Shutterstock.com

Substance use disorders (SUDs) cause immense personal, societal, and economic losses. They involve both genetic and environmental predispositions. A recent study in Nature Mental Health found that improved detection of genetic risk factors for SUD was achieved through the integrated analysis of genes associated with externalizing traits and SUDs.

Shared genetic risk links multiple substance use disorders

Twin studies suggest that SUD risk overall is heavily influenced by shared genes. Up to 80 % of genetic influences on alcohol use disorders (AUDs) are shared with other SUDs, and up to 74 % for other SUDs. Thus, the remaining genetic influences are substance-specific, rather than representing total non-genetic risk.

Further support for this view comes from SUD-focused genome-wide association studies (GWAS) covering a large number of genes. This showed a common genomic risk for such disorders.

SUDs form one component of the externalizing spectrum of behaviors and disorders characterized by behavioral disinhibition. This also includes childhood conduct disorders or adult antisocial behavior. Multiple SUDs and other externalizing traits often occur together.

All these share a common underlying genetic framework, with a latent externalizing liability estimated to be up to 80 % heritable, far exceeding the magnitude of gene associations with any individual disorder. Despite this, existing genetic research often focuses on identifying genes that confer risk for single SUDs.

Previous research by the same authors suggested that there was a large genetic overlap between SUDs and other externalizing traits, but some factors that conferred risk for specific SUDs. The current study sought to determine whether genes associated with externalizing are distinct from those that affect SUD risk or are shared across the externalizing spectrum. This would help refine study methods dealing with SUD origin.

Study tests whether SUD risk is distinct or shared

The study focused on two alternative latent genomic models: a single externalizing factor including SUDs, and a two-factor model separating behavioral disinhibition and SUD while allowing them to be highly correlated. The researchers used multivariate genomic analyses on data from over 2.2 million people, with over 5.9 million genetic variants.

They carried over the two models from their earlier work. One model treats SUDs as part of the externalizing spectrum, while the second treats them as distinct from externalizing behaviors characterized by behavioral disinhibition, yet highly correlated with them. This would help clarify the role of externalizing traits in the origin of SUDs.

Substance-specific and shared genetic pathways both identified

The researchers compared results from these models to determine whether separating SUD-specific genetic risk improved discovery beyond shared externalizing liability. They found that analyzing SUDs alongside other externalizing traits helped them uncover the neurobiological pathways and genetic architecture underlying the risk of SUDs as a whole and of specific SUDs. This advance in knowledge did not come at the expense of the specificity of genetic markers for SUD risk.

Notably, models focusing only on SUDs did not identify novel genetic signals, underscoring the importance of incorporating broader externalizing traits to improve discovery.

For instance, the first (externalizing spectrum) model yielded 708 genetic loci, of which 26 % were novel in their association with externalizing traits or addiction risk. Moreover, 57 % of them were associated with a substance use trait for the first time.

The researchers also identified genomic risk loci for the residual genetic risk for specific SUDs, such as problematic alcohol use or tobacco use. Many of these were genes involved in the metabolism of these substances, suggesting a genetically-encoded sensitivity to each substance that increases addiction risk.

The second model included two factors: disinhibition and SUD. The genomic analysis identified 631 and 48 genomic risk loci associated with these factors, respectively.

The researchers also identified genes unique to each genomic factor:

• 37 for externalizing traits – over 80 % already linked to substance use phenotypes, emphasizing the common underlying genetic architecture
• 21 for behavioral disinhibition
• 3 for SUDs – one of which is linked to multiple SUDs

They also identified genes for residual problematic tobacco and alcohol use, most of which were already known to be linked to SUD risk. However, some of them were selectively associated with a specific SUD. This type of residual risk was captured by polygenic scores (PGS) that best predicted the specific SUD.

This highlights the potential translational utility of broad and specific PGS whereby a broader metric of risk captures an individual’s general liability to addiction, whereas specific metrics can provide insight into risk for problems with specific substances.

In addition, residual genetic effects exist that act through non-externalizing mechanisms, indicating the need to look for internalizing and thought disorders as well in patients with SUDs.

The results also uncovered neurobiological pathways involved in the externalizing and behavioral disinhibition network, both of which contained higher numbers of genes linked to psychiatric illness and substance abuse.

In addition, the scientists identified >100 druggable targets (genes that could be targeted by a pharmaceutical agent) suitable for drug repurposing, especially those approved for medication-assisted treatment (MAT) of SUDs. However, these represent candidate targets rather than immediately validated therapies.

Addiction risk reflects shared traits and substance sensitivity

The genetic risk for SUD may be accounted for by a combination of increased behavioral disinhibition via a shared set of genetic loci linked to broad externalizing behavior and a biological response that increases individual sensitivity to a specific substance.

This work highlights the need to recognize the key role of externalizing in SUDs, both in clinical practice and in research. Conversely, the shared genetic architecture makes it likely that people with SUD risk or current SUD may develop other externalizing disorders. This knowledge will improve the prevention and treatment of both types of disorder.

Integrated genetic analyses could help detect more genes linked to SUD susceptibility, their mechanisms, and associated risks, with evidence that broader externalizing-based models improve gene discovery more than SUD-only analyses.

Study limitations

The study included only participants of European ancestry, limiting its generalizability to broader populations. Age at first intercourse was a factor in the GWAS analysis, which might introduce bias across populations. Residual risk analysis depends on the statistical power of the underlying GWAS and the remaining variance after accounting for shared genetic liability. Finally, SUDs are interwoven with non-externalizing forms of mental illness, which were not modeled here.

Externalizing traits and biology jointly shape addiction risk

The study suggests that the genetic risk of SUD acts through both broad externalizing and substance-specific genetic variants. The integrated approach to SUD genetic study, using externalizing trait networks, improves gene-detection power and expands current knowledge of the underlying neurobiological pathways and genetic mechanisms involved in these disorders.

Download your PDF copy by clicking here.