2019-04-23

009_ Analyzing cognitive performance of diagnostic groups on the affective-psychotic continuum and its association with polygenic risk scores for schizophrenia, bipolar disorder and major depressive disorder

Research Question and Aims

Cognitive impairments in individuals with major psychiatric disorders like schizophrenia (SZ), schizoaffective disorder (SZA), bipolar disorder (BD) and major depressive disorder (MDD) play a crucial role for the affected individuals. Cognitive deficits are not only present in the acute phase of the disorders but also in remission[1,2,3,4]. Those disorders not only overlap regarding cognition but also in symptomatology and genetics. Consequently, there is a hypothesized continuum from BD over SZA to SZ[5]. A corresponding continuum is found for cognition: individuals with BD perform significantly better than individuals with SZ, while individuals with SZA are in between[6,7]. For example, Lynham et al.[8] tested this hypothesis of increased cognitive impairment from BD (n=78) to SZA bipolar type (n=76), to SZ (n=558) and SZA depressive type (n=112) (controls: n=103). However, Lynham et al.[8];, considered neither MDD, a disorder on the affective end of the affective-to-psychotic continuum, nor genetic aspects.
Regarding genetics, there is much evidence that different diagnostic groups share underpinnings. For example, in the latest GWAS from the BD PGC, the MDD-PRS were greater in BD II vs. BD I cases, whereas greater SZ-PRS were found in BD I vs. BD II cases. Those PRS findings provide support for a continuum from MDD to BD I to BD II to SZA and SZ based on genetic effects[9].
However, results on association between polygenic risk score (PRS) for SZ (SZ-PRS) and cognitive performance are heterogeneous10. The focus is mainly on SZ and its corresponding PRS, while BD, MDD, and their corresponding PRSs are neglected. Therefore, in the present study, we would also like to research the relationship between PRS for SZ, BP and MDD, and cognitive performance in the trans-diagnostic PsyCourse sample.

This project
- is a partial replication of Lynham's approach (at least 4 out of 7 cognitive domains)
- differentiates, unlike Lynham's study, between BD I and BD II
- includes MDD as further disorder
- presents the cognitive profile of the PsyCourse cohort at visit 2
- considers a possible association of polygenic risk for SZ, BD, and MDD and cognitive performance.

The main aims of our study are:
1) Test whether the continuum of diagnostic groups based on specific cognitive domains (collected via specific neurocognitive tests) holds in the PsyCourse sample
2) Analyze a possible transdiagnostic association between cognitive performance and SZ-, BD- and MDD-PRS

Analytic Plan

H1:Cognitive performance decreases on a continuum from controls over affective to psychotic disorders
H2:There is an association between the different PRS for SZ, BD and MDD with cognitive performance

Participants:
PsyCourse 3.0
Participants at V2 with cognitive data (between N=951 and N=862)

Phenotype:
Cognitive data: TMT A/B, digit symbol test, verbal digit span forward and backward, and the verbal learning and memory test. The results of the MWT-B test of crystallized intelligence will be compared between groups.
Furthermore: see "Resources Needed"

Genomic data:
- SZ-, BD- and MDD PRS
- Ancestry principal components

Analytic methods
a) Comparing cognition between diagnostic groups
1: ANCOVA for each neurocognitive test (DV) and across diagnostic groups (IV); age, sex, medication, inpatient status as covariates; Tukey's HSD for pairwise comparisons and Bonferroni correction
2: Linear Mixed Model (LMM) to compare profiles of cognitive performance (DV) between the diagnostic groups (IV); medication, symptoms as covariates
b) Examining cognition as a dimension across diagnostic groups
For continuum analysis, cognition as dimensional phenotype: ordinal regression ? outcome: diagnosis (recoded as ordinal); predictor: single neurocognitive tests and composite cognition score (based on principal component analysis); covariates: age & sex
c) Exploring cross-disorder indicators and cognitive performance
Linear regression ? outcome: composite cognition; predictor: symptoms (PANSS, IDS-C, YMRS, GAF), lifetime severity (opcrit), lifetime presence of psychosis; cross-diagnostic Principal component analysis of cognitive data
d) Association of clusters with polygenic risk scores
Multinomial regression model to test for a possible association between cognitive performance and SZ, BD- and MDD-PRS at several p-value thresholds

Resources needed

Genetic data:
SZ-, BD and MDD-PRS
Ancestry principal components

Variables:

Socio-demographics
v1_id
v1_center
v1_sex
v1_ageBL
v1_age_1st_out_trm
v1_age_1st_inpat_trm
v1_dur_illness
v1_1st_ep
v1_fam_hist
v1_lftm_alc_dep
v2_school
v2_prof_dgr
v2_curr_paid_empl
v2_disabl_pens
v2_cur_work_restr

Diagnose
v1_scid_dsm_dx_cat
v1_scid_ever_psyc

Treatment
v2_clin_ill_ep_snc_lst
v2_cur_psy_trm
v2_Antidepressants
v2_Antipsychotics
v2_Mood_Stabilizers
v2_Tranquilizers
v2_Other_psychiatric

Symptomatology
v2_panss_sum_pos
Version 1.1
Page 7 of 9
v2_panss_sum_neg
v2_panss_sum_gen
v2_panss_sum_tot
v2_idsc_sum
v2_ymrs_sum
v2_gaf
v4_opcrit

Neurocognitive tests
v1_nrpsy_mwtb
v2_nrpsy_lng
v2_nrpsy_mtv
v2_nrpsy_tmt_A_rt
v2_nrpsy_tmt_B_rt
v2_nrpsy_dg_sym
v2_nrpsy_dgt_sp_frw
v2_nrpsy_dgt_sp_bck
v2_nrpsy_vlmt_corr
v2_nrpsy_vlmt_lss_d
v2_nrpsy_vlmt_lss_t
v2_nrpsy_vlmt_rec