2026-09-02

116_ miR-10a-5p and transdiagnostic dimensions of impulsivity and motivation

Research Question and Aims

Impulsive and compulsive behaviors are key transdiagnostic dimensions that cut across affective disorders, psychotic disorders and behavioral addictions, and they strongly affect real life functioning and treatment response 1–3. Experimental and clinical work suggests that microRNAs, including miR-10a-5p, contribute to the regulation of fronto-striatal circuits involved in reward processing, motivation and inhibitory control 4–6, but their role in human psychiatric populations remains insufficiently understood. Building on preclinical findings obtained by our group that link miR-10a-5p dysregulation in brain tissue and plasma samples to impulsive behavior, we aim to examine whether circulating microRNA profiles in the PsyCourse Study are associated with dimensional measures of (1) impulsivity, (2) compulsivity and (3) amotivation, and whether these molecular markers help explain inter individual differences in the course of illness. In a preliminary analysis (performed by a member of the PsyCourse Core Team), it was challenging to model a latent compulsivity construct in PsyCourse phenotype data, as this factor remained highly correlated with the latent impulsivity construct. We will therefore focus on the two constructs impulsivity and amotivation.
The primary aim of our proposal is to test whether baseline levels of circulating miR-10a-5p (and closely related microRNAs implicated in addictions) are associated with validated measures of impulsivity and motivational functioning in PsyCourse participants, regardless of clinical/neurotypical status. We will address this question in the PsyCourse baseline sample (V1). In secondary analyses, we will also examine whether the aforementioned changes can also be observed in broad diagnostic groups (neurotypical/affective/psychotic), and explore the correlated nature of the impulsivity and compulsivity constructs. Future aims will be to determine whether these microRNA measures covary with proteomic changes, and if these predict longitudinal changes in the selected behavioral dimensions, but these data will be requested in an Amendment.

Analytic Plan

Overview and hypotheses
As our primary hypothesis is that higher circulating miR 10a 5p expression is associated with higher impulsivity and amotivation. A further hypothesis is that these associations will hold across traditional diagnostic categories, supporting a dimensional, transdiagnostic perspective.

Phenotypes
In a feasibility approach, we have tried to extract latent factor scores for Impulsivity, Compulsivity and Amotivation from the PsyCourse dataset (v6.0), using different sets of items. Unfortunately, the Compulsivity and Impulsivity factors were highly correlated, and Compulsivity is not assessed that specifically in the PsyCourse Study. We have therefore decided to concentrate only on Amotivation and Impulsivity, using factor scores from a structural equation model (SEM) that fits the data well (Figure 1). The latent Impulsivity phenotype is extracted from the following items: v1_panss_g14 (poor impulse control), v1_panss_p4 (excitement) and v1_ymrs_itm8 (content). The latent Amotivation phenotype is extracted from the following IDSC-30 items: item 19 (interest), item 20 (energy/tiredness), and item21 (enjoyment).

MicroRNA data and quality control
We request access to small RNA/microRNA data (raw counts and, if available, normalized data) at baseline. Our primary molecular variable will be miR 10a 5p. Exploratory analyses will consider other microRNAs that are strongly correlated with miR 10a 5p and/or previously implicated in motivational alterations: miR 212 7, miR 132 8, miR 9 9, miR 34a 5p 10 and miR 30a 5p 11.
MicroRNA expression data will be processed using established pipelines that have already been applied in PsyCourse, including filtering out very lowly expressed miRNAs, normalization with count based methods such as DESeq2, and explicit adjustment for technical factors such as sequencing batch and run. We will also check for outlier samples and use false discovery rate correction for all genome wide miRNA tests.

Statistical methods
We will first perform cross-sectional analyses at baseline to test whether miR 10a 5p levels are associated with the main dimensional phenotypes of interest, including motivational impairment, behavioral disinhibition, affective instability and global functioning. Depending on the distribution of each outcome, we will use linear or generalized linear models with normalized miR 10a 5p expression as the main predictor, adjusting for key confounders such as age, sex, study site, diagnosis, illness duration, medication exposure and other relevant clinical variables. For exploratory analyses involving multiple microRNAs, p-values will be corrected using the Benjamini-Hochberg false discovery rate approach.

Resources needed

v1_id
v1_stat
v1_interv_date
v1_sex
v1_age
v1_marital_stat
v1_partner
v1_liv_aln
v1_school
v1_prof_dgr
v1_curr_paid_empl
v1_spec_emp
v1_dur_illness
v1_Antidepressants
v1_Antipsychotics
v1_Mood_stabilizers
v1_Tranquilizers
v1_Other_psychiatric
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_panss_p1
v1_panss_p2
v1_panss_p3
v1_panss_p4
v1_panss_p5
v1_panss_p6
v1_panss_p7
v1_panss_n1
v1_panss_n2
v1_panss_n3
v1_panss_n4
v1_panss_n5
v1_panss_n6
v1_panss_n7
v1_panss_g1
v1_panss_g2
v1_panss_g3
v1_panss_g4
v1_panss_g5
v1_panss_g6
v1_panss_g7
v1_panss_g8
v1_panss_g9
v1_panss_g10
v1_panss_g11
v1_panss_g12
v1_panss_g13
v1_panss_g14
v1_panss_g15
v1_panss_g16
v1_idsc_itm19
v1_idsc_itm20
v1_idsc_itm21
v1_ymrs_itm1
v1_ymrs_itm2
v1_ymrs_itm3
v1_ymrs_itm4
v1_ymrs_itm5
v1_ymrs_itm6
v1_ymrs_itm7
v1_ymrs_itm8
v1_ymrs_itm9
v1_ymrs_itm10
v1_ymrs_itm11
v1_cgi_s
v1_gaf
v1_med_pst_wk
v1_med_pst_sx_mths
v1_big_five_extra
v1_big_five_neuro
v1_big_five_openn
v1_big_five_consc
v1_big_five_agree
gsa_id