2026-03-09
111_ Assessing the Long-Term Stability of microRNAs in neurotypical individuals
Research Question and Aims
MicroRNAs are relatively stable in acellular environments and can be detected in peripheral fluids, offering insight into brain-related processes. This makes them a promising low‑cost and non‑invasive alternative to other approaches, with potential both for diagnostic and therapeutic targets. Although circulating microRNAs show considerable variability across individuals, across time, and under different pathophysiological conditions, small-scale studies suggest that some microRNAs remain relatively stable longitudinally (PMID: 38272952, here 3-months period and 22 individuals). MicroRNA dysregulation has been documented in several psychiatric disorders, further underscoring their relevance as potential biomarkers (PMID: 40505822, PMID: 39291752, PMID: 40263528).
In this study, we will apply Generalized Linear Mixed Models (GLMMs) with both random-intercepts and random-slopes, to longitudinally study microRNAs in the PsyCourse Study. First, we will quantify temporal stability and identify microRNAs whose expression remains constant or varies significantly over time in neurotypical individuals. Second, we will analyze whether the microRNAs that vary over time show temporal correlations with phenotypes most of which have previously been investigated in neurotypical individuals (PMID: 35232513), namely GAF, BDI, MSS, YMRS, IDS-C, and the SF-12. Third, we will analyze how individual levels of microRNAs that remain constant over time covary with polygenic risk scores (PRS) for SZ, BD and MDD. Genomic ancestry principal components will be included in the latter analyses.
Analytic Plan
Quality control has already been carried out on these data by one of our researchers, yielding 432 microRNAs measured in 226 healthy individuals, 298 individuals with affective disorders, and 284 individuals with psychotic disorders. To quantify the temporal stability of microRNA expression, we will fit general LMMs in R (e.g.: using the lme4 package):
- For each microRNA, quality‑controlled expression values will serve as the outcome variable – we assume that its ln-transformation is sufficient to use GLMMs without further link functions as this will approximate the normality of residuals.
- Each model will include fixed effects for time (visit), allowing detection of an overall temporal trend.
- Random effects will be used to account for individual‑level variability in measurements, and the within person standard deviation will be calculated to identify participants whose measurements may exhibit unusually high noise.
Additional covariates in the model may include age, sex, smoking habits, change in medication use, substance abuse, and genomic ancestry. Overall temporal stability will be evaluated through hypothesis testing of the fixed time effect. First, we will assess whether some miRNAs show a quadratic temporal course; then for the remaining ones we will see which miRNAs show a linear trend and which stay stable. For the subset of microRNAs identified as stable, expression levels will be regressed on PRS scores using linear regression.
Resources needed
v1_id
v1_stat
v1_center
v1_sex
v1_school
v1_prof_dgr
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
gsa_id
visit
age
alc_5orm
alc_pst6_mths
asrm_itm1
asrm_itm2
asrm_itm3
asrm_itm4
asrm_itm5
asrm_sum
bdi2_itm1
bdi2_itm10
bdi2_itm11
bdi2_itm12
bdi2_itm13
bdi2_itm14
bdi2_itm15
bdi2_itm16
bdi2_itm17
bdi2_itm18
bdi2_itm19
bdi2_itm2
bdi2_itm20
bdi2_itm21
bdi2_itm3
bdi2_itm4
bdi2_itm5
bdi2_itm6
bdi2_itm7
bdi2_itm8
bdi2_itm9
bdi2_sum
gaf
idsc_itm1
idsc_itm10
idsc_itm11
idsc_itm12
idsc_itm13
idsc_itm14
idsc_itm15
idsc_itm16
idsc_itm17
idsc_itm18
idsc_itm19
idsc_itm2
idsc_itm20
idsc_itm21
idsc_itm22
idsc_itm23
idsc_itm24
idsc_itm25
idsc_itm26
idsc_itm27
idsc_itm28
idsc_itm29
idsc_itm3
idsc_itm30
idsc_itm4
idsc_itm5
idsc_itm6
idsc_itm7
idsc_itm8
idsc_itm9
idsc_itm9a
idsc_itm9b
idsc_sum
marital_stat
med_pst_sx_mths
med_pst_wk
medchange
mss_itm1
mss_itm10
mss_itm11
mss_itm12
mss_itm13
mss_itm14
mss_itm15
mss_itm16
mss_itm17
mss_itm18
mss_itm19
mss_itm2
mss_itm20
mss_itm21
mss_itm22
mss_itm23
mss_itm24
mss_itm25
mss_itm26
mss_itm27
mss_itm28
mss_itm29
mss_itm3
mss_itm30
mss_itm31
mss_itm32
mss_itm33
mss_itm34
mss_itm35
mss_itm36
mss_itm37
mss_itm38
mss_itm39
mss_itm4
mss_itm40
mss_itm41
mss_itm42
mss_itm43
mss_itm44
mss_itm45
mss_itm46
mss_itm47
mss_itm48
mss_itm5
mss_itm6
mss_itm7
mss_itm8
mss_itm9
mss_sum
panss_g1
panss_g10
panss_g11
panss_g12
panss_g13
panss_g14
panss_g15
panss_g16
panss_g2
panss_g3
panss_g4
panss_g5
panss_g6
panss_g7
panss_g8
panss_g9
panss_n1
panss_n2
panss_n3
panss_n4
panss_n5
panss_n6
panss_n7
panss_p1
panss_p2
panss_p3
panss_p4
panss_p5
panss_p6
panss_p7
panss_sum_gen
panss_sum_neg
panss_sum_pos
panss_sum_tot
pst6_ill_drg
sf12_itm0
sf12_itm1
sf12_itm10
sf12_itm11
sf12_itm12
sf12_itm2
sf12_itm3
sf12_itm4
sf12_itm5
sf12_itm6
sf12_itm7
sf12_itm8
sf12_itm9
ymrs_itm1
ymrs_itm10
ymrs_itm11
ymrs_itm2
ymrs_itm3
ymrs_itm4
ymrs_itm5
ymrs_itm6
ymrs_itm7
ymrs_itm8
ymrs_itm9
ymrs_sum