2026-07-23

114_ Polygenic effects of impulsivity on cognition and substance abuse

Research Question and Aims

Impulsivity is a heritable trait implicated across a wide range of psychiatric disorders including ADHD (Jackson & Mackillop, 2016), bipolar disorder (Santana et al., 2022), schizophrenia (Bielecki et al., 2024), and substance abuse (Riley et al., 2016). As such, some contemporary theories have proposed dysregulation in impulsivity as a core functional process underlying general psychopathology (Carver et al., 2018; Southward et al., 2023). These models put forward that impairments in self-regulatory processes such as impulsivity constitute a common vulnerability underlying, not just specific diagnoses, but rather, diverse psychiatric conditions. Lending support to these proposals, evidence has suggested that impulsivity is genetically correlated with higher-order mental activities such as executive function, information processing, and attention (Malloy-Deniz et al., 2008). As such, these cognitive domains may represent intermediate phenotypes through which genetic risk for impulsivity manifests. In addition to cognition, research has also associated impulsivity with behavioral outcomes such as substance misuse (e.g., Vilar-Ribo et al., 2025), which constitutes a clinically relevant form of behavioral dysregulation. Further adding to this, certain gene loci implicated in impulsivity, such as CADM2 and TCF4, have also been associated with neurobiological processes such as synaptic function and neurodevelopment (Ibrahim-Verbaas et al., 2015; Davis et al., 2023), thus providing a potential mechanistic explanation as to how genetic risk for impulsivity may contribute, not only to cognitive deficits, but also to dysfunctional behavior.
Recent genome-wide association studies (GWAS) have identified genetic SNP variants associated with impulsivity, thus enabling the quantification of genetic liability toward impulsiveness using polygenic risk scores (PRS) (Sanchez-Roige et al., 2023). However, it remains unclear how this genetic liability relates to outcomes such as cognitive functioning and behavioral outcomes in clinical cohorts and, if such an association exist, whether these are observable across diagnostic categories. The primary research question is therefore whether polygenic risk for impulsivity is associated with cognitive performance in the PsyCourse Study (Budde et al., 2018). The project will further investigate whether impulsivity PRS is associated with behavioral dysregulation, as indexed by substance abuse variables collected in PsyCourse. Finally, sensitivity analyses will examine whether any potentially observed associations found in the total PsyCourse sample also can be found in diagnostic subgroups (i.e. the previously defined broad diagnostic groups of affective, psychotic and neurotypical individuals). These sensitivity analyses would provide supporting evidence to theories proposing impulsivity-related self-regulatory dysfunction as a core mechanism contributing to general psychopathology. The project may in the future also be extended to PsyCouse longitudinal data, and to determining the overlap between previously calculated PRS of a Common Executive Function (cEF) factor.
The PsyCourse Study is particularly suited to address these questions, considering the transdiagnostic design of the study, and the use of deep phenotyping. The present study will consider only baseline variables (Visit 1 or Vistit 2 for the Verbal Learning and Memory test (VLMT), see below).
The specific aims are: 1. To calculate polygenic risk scores for impulsivity in PsyCourse participants using published GWAS (Sanchez-Roige et al., 2023) summary statistics
2. To test associations between impulsivity PRS and cognitive domains related to self-regulation, including executive functioning (TMT-A and B and a previously calculated phenotypic latent common EF), short-term and working memory (digit-span forward and backwards), processing speed (digit-symbol test), and crystallized intelligence (MWT-B). The VLMT will be included as a non-executive test, to assess the specificity of the associations.
3. To investigate whether impulsivity PRS is associated with substance abuse in PsyCourse participants, i.e. either with “ever smoked”, “ever illicit drugs”, “alcohol dependence”, or “ever heavy substance user”.
4. To assess whether any potential arising associations remain after controlling for diagnosis, thereby supporting a transdiagnostic interpretation aligned with contemporary models of psychopathology (Carver et al., 2018; Southward et al., 2023)

Hypothesis 1: Higher impulsivity PRS is associated with poorer outcomes in at least one of the above-mentioned cognitive tests assessed in the PsyCourse Study.

Hypothesis 2: Higher impulsivity PRS is associated with higher rates of substance abuse in the PsyCourse Study in at least one of the associated variables.

Hypothesis 3: Associations between impulsivity PRS and cognitive performance and substance abuse are largely transdiagnostic, i.e. remain when adjusting for diagnostic status.

Analytic Plan

PRS for impulsivity will be calculated for PsyCourse participants using summary statistics from a recent GWAS of impulsivity traits (Sanchez-Roige et al., 2023) and the PRS-CS toolbox. These summary statistics use 23andMe applicants and are thus not freely downloadable (only the top 10,000 SNPs are). We have already applied for these summary statistics. If our application is unsuccessful or delayed, we will use the freely downloadable top 10,000 SNPs. We will use PsyCourse genomic data (GSA chip) that were already quality-controlled by Sergi Papiol. Impulsivity PRS will then be calculated to genotyped participants in the PsyCourse study.
H1/H2: Associations between impulsivity PRS and cognitive domains/substance use mentioned above, will be examined using linear regressions or linear mixed models, if the latter improve model fit. Regression models will be adjusted for a minimal set of relevant covariates (e.g., age, sex, ancestry PCAs). The results of the regression models will be adjusted for multiple comparisons using the False-Discovery-Rate (FDR).
H3: To evaluate the transdiagnostic nature of any potentially observed associations, we will analyze whether impulsivity PRS effects on cognition remain after adjusting for diagnostic category.
Importantly secondary analyses will explore substance-specific variables.

Resources needed

v1_id
v1_stat
v1_center
v1_interv_date
v1_sex
v1_age
v1_marital_stat
v1_partner
v1_liv_aln
v1_school
v1_prof_dgr
v1_ed_status
v1_curr_paid_empl
v1_disabl_pens
v1_spec_emp
v1_cntr_brth
v1_cur_psy_trm
v1_outpat_psy_trm
v1_dur_illness
v1_1st_ep
v1_Antidepressants
v1_Antipsychotics
v1_Mood_stabilizers
v1_Tranquilizers
v1_Other_psychiatric
v1_fam_hist
v1_ever_smkd
v1_age_smk
v1_no_cig
v1_alc_pst12_mths
v1_alc_5orm
v1_lftm_alc_dep
v1_evr_ill_drg
v1_sti_cat_evr
v1_can_cat_evr
v1_opi_cat_evr
v1_kok_cat_evr
v1_hal_cat_evr
v1_inh_cat_evr
v1_tra_cat_evr
v1_var_cat_evr
v1_evr_hvy_usr
v1_pst6_ill_drg
v1_scid_dsm_dx
v1_scid_dsm_dx_cat
v1_gaf
v1_nrpsy_com
v1_nrpsy_lng
v1_nrpsy_mtv
v1_nrpsy_tmt_A_rt
v1_nrpsy_tmt_B_rt
v1_nrpsy_dgt_sp_frw
v1_nrpsy_dgt_sp_bck
v1_nrpsy_dg_sym
v1_nrpsy_mwtb
v2_nrpsy_vlmt_check
v2_nrpsy_vlmt_corr
v2_nrpsy_vlmt_lss_d
v2_nrpsy_vlmt_lss_t
v2_nrpsy_vlmt_rec