Thread by @AlexanderHatoum

Preprint alert! TL;DR We find that a common genetic mechanism underlies Substance Use Disorder behaviors, and remains when controlling for genetics of substance use. Suggesting an #addictions genetic mechanism, what we call the a(g)-Factor, a thread! 1/8 medrxiv.org/content/10.110…

Polysubstance use or substitution behaviors are the norm among cases. With more lifetime SUDs comorbidity we often see more severe presentations of Substance Use Disorders (SUDs). Our current nosologies are defined by the drug, but there are efforts to find deeper mechanisms. 2/8

Further, any evidence of overlap among substance use disorders should also be aware of the high comorbidity with psychopathology and the division between SUDs and normative substance use behaviors like initiation and frequency of substance use. 3/8

I.e. drinking more is not necessarily an SUD. While use and use disorder are inherently linked, it is possible to separate their risk with genetic designs. Genetic correlations with (lower diagonal) and without (upper diagonal) accounting for substance use below 4/8

Genetic correlations are some of the highest in psychiatric genetics and they form a well fitting factor model, a(g), that accounts for most of the genetics of opioid use disorder and cannabis use disorder (when controlling for substance use), offering power for illicit SUDs 5/8

As a confirmation, a(g) relates to the processes we expect SUDs to relate to. Independent associations with deficits in executive function (shameless plug: biorxiv.org/content/10.110…), Neuroticism, and risk-taking behaviors (when accounting for substance use).

Finally, and perhaps most importantly, the linear combination of psychopathology factors and substance use does not account for a(g), instead SUDs genetics is a distinguishable unique dimension. This uniqueness is predicted by deficits in executive function.

We think that this structure will refine psychiatric nosologies and inform future GWAS and predictive modeling. This is the first of series of papers, including a GWAS and some ML work! @apadivision50 #PsychTwitter, #MedTwitter

Also, would like to thank the coauthors who are on twitter! @e_covey_johnson, @RenatoPolimanti, @HEdenberg2 @hangzhou202 @raymondkwalters, the PGC-SUD working group, Joel Gelernter, Ryan Bogdan, and Arpana Agrawal.