Skip to main content
NCBI home page
Canadian Journal of Psychiatry. Revue Canadienne de Psychiatrie logoLink to Canadian Journal of Psychiatry. Revue Canadienne de Psychiatrie
. 2025 Mar 29;70(7):552–564. doi: 10.1177/07067437251329073

Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1

Leen Magarbeh 1,2, Samar S M Elsheikh 1, Farhana Islam 1,2, Victoria S Marshe 3, Xiaoyu Men 1,2, Emytis Tavakoli 1, Martin Kronenbuerger 1, Stefan Kloiber 1,2,4,5, Benicio N Frey 6,7, Roumen Milev 8, Claudio N Soares 8, Sagar V Parikh 9, Franca Placenza 10, Stefanie Hassel 11,12, Valerie H Taylor 11, Francesco Leri 13, Pierre Blier 14, Rudolf Uher 15, Faranak Farzan 16, Raymond W Lam 17, Gustavo Turecki 18, Jane A Foster 6,19, Susan Rotzinger 7, Sidney H Kennedy 4,5,10,20,21, Daniel J Müller 1,2,4,5,22,23,✉
PMCID: PMC11955985  PMID: 40156272

Abstract

Objective

The genetic architecture of antidepressant response is poorly understood. This study investigated whether polygenic risk scores (PRSs) for major psychiatric disorders and a personality trait (neuroticism) are associated with antidepressant treatment outcomes.

Methods

We analysed 148 participants with major depressive disorder (MDD) from the Canadian Biomarker Integration Network for Depression-1 (CAN-BIND-1) cohort. Participants initially received escitalopram (ESC) monotherapy for 8 weeks. Nonresponders at week 8 received augmentation with aripiprazole (ARI), while responders continued ESC until week 16. Primary outcomes were remission status and symptom improvement measured at weeks 8 and 16. At week 16, post-hoc stratified analyses were performed by treatment arm (ESC-only vs. ESC + ARI). Eleven PRSs derived from genome-wide association studies of psychiatric disorders (e.g., MDD and post-traumatic stress syndrome (PTSD)) and neuroticism, were analysed for associations with these outcomes using logistic and linear regression models.

Results

At week 8, a higher PRS for PTSD was nominally associated with a lower probability of remission (odds ratio (OR) = 0.08 [0.014–0.42], empirical p-value = 0.017) and reduced symptom improvement (beta (standard error) = −29.15 (9.76), empirical p-value = 0.019). Similarly, a higher PRS for MDD was nominally associated with decreased remission probability (OR = 0.38 [0.18–0.78], empirical p-value = 0.044). However, none of the results survived multiple testing corrections. At week 16, the stratified analysis for the ESC-only group revealed that a higher PRS for MDD was associated with increased remission probability (empirical p-value = 0.034) and greater symptom improvement (empirical p-value = 0.02). In contrast, higher PRSs for schizophrenia (empirical p-value = 0.013) and attention-deficit hyperactivity disorder (empirical p-value = 0.032) were associated with lower symptom improvement. No significant associations were observed in the ESC + ARI group.

Conclusions

These findings suggest that PRSs may influence treatment outcomes, particularly in ESC monotherapy. Replication in larger studies is needed to validate these observations.

Keywords: polygenic risk scores, major depressive disorder, antidepressant, treatment outcomes

Graphical abstract.

Graphical abstract

Open in a new tab

This is a visual representation of the abstract.

Introduction

Major depressive disorder (MDD) is a leading cause of disability worldwide, 1 affecting over 250 million people and contributing significantly to the global mental health burden.2,3 Second-generation antidepressants (ADs), specifically selective serotonin reuptake inhibitors (SSRIs), are first-line interventions for MDD. 4 Despite the widespread use of ADs and their greater efficacy compared to placebo, 5 only about one-third of MDD patients achieve complete remission following their initial prescription.6,7 Another one-third requires treatment change or augmentation, while the remaining one-third experience treatment failure (i.e., they do not respond to at least 2 different AD medications).6,7 This variability in treatment outcomes emphasizes the need to identify clinical and biological predictors of AD response.

Response to ADs has a genetic basis, 8 with 42% (standard error (SE) = 18%; 95% confidence interval [CI], 7% to 77%) of the variance may be explained by common single nucleotide polymorphisms (SNPs). 9 Notably, accumulating evidence from genome-wide association studies (GWASs) suggests that AD response is a complex polygenic trait influenced by many genetic variants or SNPs.9–13 Each of these variants has a small but cumulative effect on treatment response to AD.9–13 The aggregation of those SNPs through polygenic risk scoring (PRS) has emerged as a promising approach to predict AD treatment outcomes in patients with MDD. 14

PRS is a method that calculates an individual's genetic risk for a trait by using genome-wide SNPs, which are weighted according to their effect sizes from GWAS data and corresponding summary statistics. 15 Recent studies using PRS have found genetic links between common psychiatric and non-psychiatric disorders and AD treatment outcomes. 16 These studies indicate that higher genetic loading for MDD,17,18 neuroticism, 17 schizophrenia,14,18 coronary artery disease, 19 or obesity 19 are associated with less favourable AD treatment response. In addition, a higher PRS of attention-deficit hyperactivity disorder (ADHD) was associated with treatment-resistant depression. 20 Therefore, using PRSs could be valuable in identifying MDD patients at risk for poor treatment response or the development of treatment-resistant depression. A recent systematic review, which summarized current evidence of the genetic overlap between AD treatment response and PRSs for 30 traits in patients with MDD, has concluded that the majority of these studies were underpowered and assessed heterogeneous populations. 16 As a result, additional studies using standardized outcome measures are needed to expand the MDD patient cohort and to facilitate meta-analyses, which could identify more robust genetic associations.

In this exploratory study, we assessed the genetic associations between AD treatment outcomes and both psychiatric disorders and non-psychiatric traits in MDD patients from the well-characterized Canadian Biomarker Integration Network For Depression (CAN-BIND-1) cohort. 21 The CAN-BIND-1 cohort included 211 adults treated with escitalopram (ESC) and/or aripiprazole (ARI; an antipsychotic used for augmentation therapy) for 16 weeks. We constructed PRSs for clinical disorders (MDD, 11 bipolar disorder, 22 schizophrenia, 23 ADHD, 24 primary anxiety disorders, 25 post-traumatic stress disorder (PTSD), 26 and substance-use disorders (SUDs) 27 ), a personality trait (neuroticism 28 ), and treatment outcomes (antidepressant non-remission and percentage improvement). 14 We then examined the association of these individual scores with our 2 phenotypes of antidepressant response: (a) remission, a binary outcome (remitter vs. non-remitter) and (b) symptom improvement, a continuous outcome measured as a percentage change in rating scale scores.

Given the complexity of the antidepressant response, which may involve multiple biological pathways and shared genetic influences across psychiatric and nonpsychiatric traits (see Supplemental Figure 1), we further investigated the joint predictive capacity of combining these individual PRSs. 29 This approach, known as the multiple polygenic risk score (mPRS) methodology, has been shown to capture more genetic variance than individual scores.30,31

Methods

Target Dataset

The PRS analysis was conducted on the dataset collected from the CAN-BIND-1 cohort. CAN-BIND-1 was a multicenter discovery study designed to identify predictors of treatment response in MDD participants (ClinicalTrials.gov identifier: NCT01655706). The CAN-BIND-1 sample consisted of 211 (18–61 years old) participants diagnosed with MDD in a current major depressive episode of ≥3 months according to the Diagnostic and Statistical Manual for Mental Disorders IV 32 using the Mini International Neuropsychiatric Interview. 33 In addition, all participants were free of psychotropic medications for at least 5 half-lives prior to the start of the study and had a Montgomery–Asberg Depression Rating Scale (MADRS) score ≥24 at the time of screening. 34 A detailed description of the CAN-BIND-1 protocol, study design, inclusion, and exclusion criteria are available elsewhere.21,35

CAN-BIND-1 consisted of a 2-phase study protocol following screening and baseline visits. During Phase I (weeks 0 to 8), all participants received open-label ESC monotherapy (10–20 mg/day, flexible dosage). Depressive symptoms were assessed using the MADRS every 2 weeks. At the end of Phase I or week 8, participants were classified as responders if they achieved a 50% or greater reduction in their MADRS from baseline, or non-responders if the reduction in their MADRS from baseline was <50%. During Phase II (weeks 8–16), responders continued with ESC monotherapy until the end of the study period, whereas non-responders were given augmentation therapy with ARI (2–10 mg/day, flexible dosage). Treatment remission was defined as a MADRS score of ≤10, assessed at the end of Phase I (week 8) and Phase II (week 16) visits (see Supplemental Figure 2).

Blood samples for genetic analysis were collected in week 4. Genomic DNA was extracted from venous blood samples using a modified version of the FlexiGene DNA kit (Qiagen, Hilden, Germany) and sent for genotyping to the Pharmacogenetics Research Clinic (Centre for Addiction and Mental Health, Toronto, Canada). Genotyping was performed on 190 participants using the Illumina Omni 2.5 BeadChip genome-wide microarray according to the manufacturer's protocol. Imputation was also conducted, see Supplemental Material for detailed methodology. Quality control on the genome-wide array data was performed using PLINK 2.00 36 (https://www.cog-genomics.org/plink2). Variants with missing rate ≥5%, minor allele frequency <1%, and Hardy-Weinberg equilibrium test with p-value <5 × 10−6 were excluded.37,38 Furthermore, we excluded participants with genotyping rate <99% (n = 4), sex discrepancies between self-reported sex and genetic sex (n = 3), abnormal heterozygosity (n = 4), and related participants (identity by descent >0.185, n = 4).37,38 Our final sample included 175 participants who passed genetic data quality control and included 1,575,137 SNPs.

Polygenic Risk Scores (PRSs)

Discovery Cohorts for PRS Derivation

PRSs were constructed on the target dataset using summary statistics from well-powered GWASs. These studies were selected based on their large sample sizes (i.e., mega- and meta-analyses) and the availability of publicly accessible summary statistics from the Psychiatric Genomic Consortium (https://pgc.unc.edu/for-researchers/download-results/). 39 Risk scores were constructed for outcomes across 9 large genome-wide studies, including MDD (with and without the data from 23andMe), 11 bipolar disorder, 22 schizophrenia, 23 ADHD, 24 primary anxiety disorders, 25 PTSD, 26 neuroticism, 28 SUD, 27 and antidepressant non-remission and percentage improvement, 14 see Supplemental Table 1.

Single PRS Calculation and Association With Outcome of Interest

PRSs were constructed using the clumping and thresholding method with the PRSice-2 software (version 2.3.5) 15 across 10 p-value thresholds, ranging from PT = 0.0001 to PT = 1. Only SNPs with a GWAS association p-value below the specified thresholds were included in the PRS calculation; all other SNPs were excluded. The constructed scores were then evaluated with our following primary outcome measures at the end of Phase I (week 8) and Phase II (week 16) using logistic and linear regressions:

  1. Remission status: A dichotomous measure, where participants were classified as remitters if their MADRS score was ≤10, and non-remitters otherwise.

  2. Symptom improvement: Defined as the percentage change in MADRS score from baseline.

In Phase II (week 16), all participants were initially analysed together with the treatment arm (ESC-only or ESC + ARI) included as a covariate. We then conducted a post-hoc stratified analysis by treatment arm (ESC-only vs. ESC + ARI) to further explore potential differences in genetic associations. All regression models were adjusted for covariates including age, sex, baseline MADRS, and the first 3 principal components to account for population stratification. We also conducted 10,000 permutations testing to obtain empirical p-values, accounting for potential inflation and overfitting. The “meff” function from the R package “poolr” 38 was used to calculate the number of independent tests in our target dataset. 40 To correct for multiple testing, we employed Nyholt's method, which adjusts for the nonindependence of the variables in the analysis. 41 As a result, the effective number of independent tests was estimated to be 38, resulting in a corrected p-value threshold of <0.0013 (=0.05/38) for statistical significance. When describing results from this study, nominal significance refers to empirical p-values < 0.05 that did not reach the multiple testing correction threshold.

Post-hoc PRS Score Stratification

To explore whether participants with higher polygenic risk exhibited distinct clinical or demographical profiles, we divided participants into high and low PRS groups based on the median split of individual PRSs. Participants with PRS values above the median were categorized as the high PRS group, while those below the median were categorized as the low PRS group. Differences in clinical and demographic variables between the high and low PRS groups were assessed using chi-square tests for categorical variables and t-tests for continuous variables. See Supplemental Tables 10–13 for detailed results.

Multiple PRSs and Model Evaluation

We explored the collective predictive capability by combining individual polygenic scores using an mPRS methodology. 30 Predictive models were constructed using elastic net regression, with detailed information regarding model building, hyperparameter tuning, and performance evaluation provided in the Supplemental material. Briefly, 2 types of models were developed for each outcome (remission and symptom improvement) at weeks 8 and 16: (1) a clinical model including age, sex, ethnicity, baseline MADRS, and treatment arm, and (2) a full model incorporating all covariates in the clinical model with the addition of PRSs.

Results

CAN-BIND-1 Target Dataset Characteristics

Genome-wide data were available for 190 participants. We excluded 15 participants who failed quality control checks, as described in the “Target Dataset” section, and 27 participants who dropped out prior to week 8, resulting in a lack of MADRS scores for both Phases I and II. Therefore, the final sample size for Phase I included 148 participants from the CAN-BIND-1 target dataset, of whom 70 (47%) continued ESC monotherapy, and 78 (53%) received adjunctive ARI (ESC + ARI) in Phase II. By the end of Phase II (week 16), 12 additional participants dropped out, resulting in a sample size of 136 for Phase II analyses. Demographic and clinical characteristics of the target dataset for both Phases I and II are summarized in Table 1 and Supplemental Tables 2 and 3. Overall, the 148 participants had a mean age of 36.3 (standard deviation (SD) = 13.1) years, of which 58.1% were female (n = 86 out of 148) and 73.6% were considered of European ancestry (n = 109 out of 148), with no significant differences between the 2 treatment arms. The overall remission rates at the end of Phases I and II were 31.8% (n = 47 out of 148) and 55.4% (n = 82 out of 148), respectively. At week 16, the proportion of remitters in the ESC group was significantly higher than in the ESC + ARI group (71.4% vs. 41.0%, p-value < 0.001).

Table 1.

Demographic and Clinical Characteristics of the CAN-BIND-1 Target Dataset Stratified by Treatment Arm.

Overall (N = 148) Phase II Treatment Arm
Variables ESC (N = 70) ESC + ARI (N = 78)
N (%) N (%) N (%) P-value a
Sex 0.659
 Female 86 (58.1%) 42 (60.0%) 44 (56.4%)
Ethnicity 0.868
 Non-White 39 (26.4%) 18 (25.7%) 21 (26.9%)
 White 109 (73.6%) 52 (74.3%) 57 (73.1%)
b Ancestry 0.629
 Admixed 7 (4.7%) 3 (4.3%) 4 (5.1%)
 African 3 (2.0%) 3 (4.3%) 0 (0%)
 American 7 (4.7%) 3 (4.3%) 4 (5.1%)
 East Asian 17 (11.5%) 7 (10.0%) 10 (13.0%)
 European 109 (73.6%) 52 (74.3%) 57 (73.1%)
 South Asian 5 (3.4%) 2 (2.9%) 3 (3.8%)
Remission status at week 8 <0.001
 Remitter 47 (31.8%) 47 (67%) 0 (0%)
 Non-remitter 101 (68.2%) 23 (33%) 78 (100%)
Remission status at week 16 <0.001
 Remitter 82 (55.4%) 50 (71.4%) 32 (41.0%)
 Non-remitter 54 (36.5%) 14 (20.0%) 40 (51.3%)
ESC dosage at week 2 >0.999
 10 mg 147 (99.3%) 70 (100%) 77 (98.7%)
 20 mg 1 (0.7%) 0 (0%) 1 (1.3%)
ESC dosage at week 16 0.207
 10 mg 14 (9.5%) 9 (12.9%) 5 (6.4%)
 15 mg 1 (0.7%) 1 (1.4%) 0 (0%)
 20 mg 119 (80.4%) 54 (77.1%) 65 (83.3%)
ARI dosage at week 10 >0.999
 2 mg 67 (45.3%) 0 (0%) 67 (85.9%)
 5 mg 2 (1.4%) 0 (0%) 2 (2.6%)
ARI dosage at week 16 0.600
 2 mg 14 (9.5%) 0 (0%) 14 (17.9%)
 5 mg 25 (16.9%) 1 (1.4%) 24 (30.8%)
 10 mg 26 (17.6%) 0 (0%) 26 (33.3%)
Mean (SD) Mean (SD) Mean (SD) P-value a
Age 36.28 (13.11) 35.13 (12.77) 37.32 (13.40) 0.310
Baseline MADRS score 30.21 (5.56) 29.07 (5.46) 31.23 (5.48) 0.010
%Δ in MADRS score from Baseline at week 8 46.26 (31.36) 72.93 (15.82) 22.32 (20.63) <0.001
%Δ in MADRS score from baseline at week 16 66.69 (26.50) 77.35 (21.68) 57.21 (26.93) <0.001
Open in a new tab

Note. Mean (with standard deviation) and frequency are displayed for continuous and categorical variables, respectively.

ESC = escitalopram; ARI = aripiprazole; MADRS = Montgomery–Åsberg depression rating scale; SD = standard deviation. Statistically significant differences between treatment arms are indicated in bold.

aKruskal–Wallis rank sum test for continuous variables; Pearson's chi-squared test or Fisher's exact test for categorical variables.

bCategories for ancestry are adapted from the International Genome Sample Resource 1000 Genomes Project (http://www.internationalgenome.org/category/population/).

Association of PRSs With Outcomes of Interest in the Target Dataset

Phase I (Week 8) Results for all Participants (n = 148)

The associations of PRSs with remission status and symptom improvement at week 8 are shown in Figures 1 and 2 and Tables 2 and 3 (see also Supplemental Figures 3 to 6 showing the associations across all GWAS p-value thresholds). At week 8, the PRS for PTSD, MDD, and anxiety were nominally associated with ESC remission and/or symptom improvement. A 1 SD increase in polygenic risk for PTSD was associated with decreased probability of remission (OR = 0.08 [0.014–0.42], p-value = 0.003, empirical p-value = 0.017) and lower symptom improvement (Beta (SE) = −29.15 (9.76), p-value = 0.003, empirical p-value = 0.019). Similarly, the PRS for MDD (excluding 23andMe) was associated with a decreased probability of remission (OR = 0.38 [0.18–0.78], p-value = 0.009, empirical p-value = 0.044). In contrast, a 1 SD increase in polygenic risk for primary anxiety disorders was associated with better symptom improvement (Beta (SE) = 7.76 (2.75), p-value = 0.006, empirical p-value = 0.029). Nonetheless, none of the PRS associations above survived the multiple testing correction threshold of p-value < 0.0013.

Figure 1.

Figure 1.

Open in a new tab

Association of PRSs for various phenotypes with remission status (remitter vs. non-remitter) at week 8 (n = 148) and week 16 (n = 136). A logistic regression was used with covariates age, sex, baseline MADRS scores and the first 3 principal components to account for population heterogeneity. At week 16, the treatment arm was also added as a covariate (as there were individuals who were given escitalopram only and individuals who were given escitalopram + aripiprazole). MDD refers to MDD with 23andMe data. *Empirical p-value < 0.05 before multiple testing correction. All empirical values became nonsignificant after multiple testing corrections at a threshold of p-value < 0.0013.

Note. ADHD = attention-deficit hyperactivity disorder; MDD = major depressive disorder; PTSD = post-traumatic stress disorder; SUD = substance-use disorders; PRS = polygenic risk score.

Figure 2.

Figure 2.

Open in a new tab

Association of PRSs for various phenotypes with symptom improvement (i.e., % change in MADRS; also referred to as a change in the figure) at week 8 (n = 148) and week 16 (n = 136). A linear regression was used with covariates age, sex, baseline MADRS scores, and the first 3 principal components to account for population heterogeneity. At week 16, the treatment arm was also added as a covariate (as there were individuals who were given escitalopram only and individuals who were given escitalopram + aripiprazole). MDD refers to MDD with 23andMe data. *Empirical p-value < 0.05 before multiple testing correction. All empirical values became nonsignificant after multiple testing corrections at a threshold of p-value < 0.0013.

Note. ADHD = attention-deficit hyperactivity disorder; MDD = major depressive disorder; PTSD = post-traumatic stress disorder; SUD = substance-use disorders; PRS = polygenic risk score.

Table 2.

Association of Polygenic Risk Scores with Remission Status (Remitter vs. Non-remitter) at Weeks 8 (n = 148) and 16 (n = 136) for all Participants.

Phenotype Threshold PRSR2 PRS R2 (adj.) Full R2 NullR2 Coef. SE P-value No. of SNPs Empirical P-value OR [95% confidence interval]
Remission at Week 8
Schizophrenia 1.0000 0.008 0.010 0.210 0.200 −0.626 0.642 0.330 128,087 0.675 0.535 [0.152–1.882]
PTSD 0.0500 0.081 0.100 0.300 0.200 −2.581 0.869 0.003 17,950 0.017* 0.076 [0.014–0.415]
MDD with 23andMe 0.0010 0.002 0.003 0.202 0.200 0.105 0.204 0.604 669 0.926 1.111 [0.746–1.656]
SUD 1.0000 0.020 0.025 0.225 0.200 −0.300 0.193 0.126 87,549 0.400 0.743 [0.508–1.081]
Neuroticism 0.0001 0.007 0.008 0.208 0.200 0.175 0.199 0.377 103 0.887 1.192 [0.807–1.759]
MDD excluding 23andMe 0.0500 0.061 0.076 0.276 0.200 −0.976 0.371 0.009 20,793 0.044* 0.377 [0.182–0.780]
Bipolar 0.0010 0.035 0.043 0.243 0.200 −0.453 0.226 0.045 2502 0.148 0.636 [0.408–0.991]
Anxiety a 0.0500 0.042 0.052 0.252 0.200 0.480 0.219 0.028 17,760 0.130 1.616 [1.053–2.481]
Antidepressant percentage improvement 0.0001 0.005 0.006 0.205 0.200 0.149 0.203 0.463 81 0.941 1.160 [0.780–1.726]
Antidepressant non-remission 0.0010 0.007 0.009 0.209 0.200 0.190 0.204 0.350 482 0.841 1.210 [0.811–1.804]
ADHD 0.3000 0.003 0.004 0.204 0.200 −0.161 0.265 0.542 62,272 0.954 0.851 [0.506–1.430]
Remission at Week 16
Schizophrenia 1.0000 0.003 0.004 0.270 0.266 0.415 0.655 0.526 128,087 0.900 1.515 [0.420–5.467]
PTSD 0.0001 0.025 0.029 0.295 0.266 0.448 0.263 0.089 95 0.352 1.565 [0.934–2.623]
MDD with 23andMe 0.0001 0.065 0.077 0.343 0.266 0.620 0.228 0.007 206 0.025* 1.858 [1.188–2.907]
SUD 0.0001 0.006 0.007 0.272 0.266 0.163 0.193 0.400 316 0.848 1.177 [0.806–1.719]
Neuroticism 0.0010 0.010 0.011 0.277 0.266 0.206 0.193 0.286 693 0.776 1.229 [0.842–1.794]
MDD excluding 23andMe 0.0001 0.027 0.032 0.297 0.266 0.370 0.209 0.076 254 0.286 1.448 [0.962–2.181]
Bipolar 0.0001 0.012 0.014 0.280 0.266 0.248 0.207 0.232 859 0.545 1.281 [0.854–1.922]
Anxiety a 0.0100 0.007 0.009 0.275 0.266 0.194 0.208 0.349 4668 0.849 1.215 [0.808–1.825]
Antidepressant percentage improvement 0.0001 0.083 0.099 0.364 0.266 −0.747 0.249 0.003 81 0.015* 0.474 [0.291–0.773]
Antidepressant non-remission 0.0500 0.012 0.015 0.236 0.221 −0.390 0.326 0.232 15,998 0.676 0.677 [0.357–1.283]
ADHD 0.0010 0.020 0.024 0.290 0.266 −0.328 0.211 0.120 1819 0.363 0.720 [0.476–1.089]
Open in a new tab

aPrimary anxiety disorders. Logistic regression was used for analysis. Covariates included age, sex, the first 3 principal component analysis, baseline MADRS and treatment arm (for week 16 only). Empirical p-value: correction for overfitting. *Empirical (EMP.) p-value < 0.05 before multiple testing corrections. All empirical values became nonsignificant after multiple testing corrections at a threshold of p-value < 0.0013. ADHD = attention-deficit hyperactivity disorder; MDD = major depressive disorder; PTSD = post-traumatic stress disorder; SUD = substance-use disorders; MADRS = Montgomery–Åsberg depression rating scale.

Table 3.

Association of Polygenic Risk Scores With Symptom Improvement (Percent Change in MADRS From Baseline) at Weeks 8 (n = 148) and 16 (n = 136) for all Participants.

Phenotype Threshold PRS R2 Full R2 Coef. SE P-value No. of SNPs Empirical p-value b
Change at Week 8
Schizophrenia 0.0100 0.023 0.063 8.526 4.573 0.064 12,107 0.182
PTSD 0.0500 0.0575 0.093 −29.149 9.758 0.003 17,950 0.019*
MDD with 23andMe 0.0100 0.008 0.047 3.143 2.993 0.295 2556 0.606
SUD 1.000 0.027 0.066 −5.264 2.616 0.046 87,549 0.163
Neuroticism 0.0001 0.010 0.050 3.360 2.723 0.219 103 0.668
MDD excluding 23andMe 0.0500 0.028 0.068 −9.635 4.671 0.040 20,793 0.172
Bipolar 0.0100 0.035 0.074 −7.940 3.471 0.024 9024 0.079
Anxiety a 0.0500 0.052 0.091 7.758 2.754 0.006 17,760 0.029*
Antidepressant percentage improvement 0.01 0.004 0.043 1.917 2.603 0.463 4133 0.941
Antidepressant non-remission 0.0010 0.007 0.059 4.777 2.800 0.090 482 0.332
ADHD 0.3000 0.012 0.052 −4.504 3.368 0.183 62,272 0.502
Change at Week 16
Schizophrenia 0.0100 0.004 0.190 3.015 3.871 0.437 12,107 0.824
PTSD 0.0001 0.015 0.200 4.162 2.733 0.130 95 0.470
MDD with 23andMe 0.0001 0.038 0.224 5.555 2.235 0.014 206 0.042*
SUD 0.001 0.007 0.192 2.194 2.104 0.299 1100 0.729
Neuroticism 0.0010 0.023 0.209 4.119 2.128 0.055 693 0.226
MDD excluding 23andMe 0.0001 0.038 0.225 5.387 2.202 0.013 254 0.060
Bipolar 0.0500 0.013 0.198 −4.463 3.135 0.157 23,754 0.394
Anxiety a 0.0010 0.007 0.193 −2.321 2.150 0.282 616 0.753
Antidepressant percentage improvement 0.0001 0.037 0.223 −5.597 2.274 0.015 81 0.068
Antidepressant non-remission 0.0001 0.001 0.187 −0.972 2.439 0.691 58 0.997
ADHD 0.0100 0.400 0.226 −7.337 2.873 0.012 7381 0.049*
Open in a new tab

aPrimary anxiety disorders. Null R2 at week 8 is 0.0395 and at week 16 is 0.186. Linear regression was used for analysis. Covariates included age, sex, baseline MADRS, the first 3 principal component analyses and the treatment arm (for week 16 only).

bEmpirical p-value: adjusted p-value after 10,000 permutation tests to correct for overfitting.

*Empirical p-value < 0.05 before multiple testing correction. All empirical values became nonsignificant after multiple testing corrections at a threshold of p-value < 0.0013. ADHD = attention-deficit hyperactivity disorder; MDD = major depressive disorder; PTSD = post-traumatic stress disorder; SUD = substance-use disorders; MADRS = Montgomery–Åsberg depression rating scale; SE = standard error.

Phase II (Week 16) Results for all Participants (n = 136)

The associations of PRSs with remission status and symptom improvement at week 16 are shown in Figures 1 and 2 and Tables 2 and 3 (see also Supplemental Figures 3–6 showing the associations across all GWAS p-value thresholds). Specifically, the PRS for MDD (with 23andMe) was nominally associated with increased probability of remission (OR = 1.86 [1.19–2.91], p-value = 0.007, empirical p-value = 0.025) and better symptom improvement (Beta (SE) = 5.56 (2.24), p-value = 0.014, empirical p-value = 0.04). In contrast, the PRS for antidepressant percentage improvement was nominally associated with decreased probability of remission (OR = 0.47 [0.291–0.773], p-value = 0.003, empirical p-value = 0.015). The PRS for ADHD was nominally associated with lower symptom improvement (Beta (SE) = −7.34 (2.87), p-value = 0.012, empirical p-value = 0.049). None of the PRS associations above survived the multiple testing correction threshold of p-value < 0.0013.

Phase II Post-hoc Stratified Analyses by Treatment Arm (ESC vs. ESC + ARI)

The associations of PRSs with remission status and symptom improvement stratified by treatment arm at week 16 are present in Supplemental Tables 4 and 5. For the ESC-only group (n = 64), the PRS for MDD (with 23andMe) was associated with an increased probability of remission (OR = 3.69 [1.40–9.73], p-value = 0.008, empirical p-value = 0.034) and better symptom improvement (Beta (SE) = 7.65 (2.68), p-value = 0.006, empirical p-value = 0.02). Furthermore, the PRSs for schizophrenia (Beta (SE) = −23.69 (7.78), p-value = 0.003, empirical p-value = 0.013), ADHD (Beta (SE) = −9.52 (3.4), p-value = 0.007, empirical p-value = 0.032) and antidepressant percentage improvement (Beta (SE) = −10.21 (2.59), p-value = 0.0002, empirical p-value = 0.002) were associated with lower symptom improvements. For the ESC + ARI group (n = 72), there were no significant associations present.

Post-hoc PRS Score Stratification

Demographic and clinical characteristics were compared between high and low PRS groups for MDD, PTSD, Anxiety, and ADHD (see Supplemental Tables 10 to 13). No significant differences were observed in age, sex, baseline MADRS scores, or treatment arm across PRS groups (p-value > 0.05). However, a significant difference in ethnicity distribution was observed for all PRS types (p-value < 0.001), with a higher proportion of White individuals in the low PRS group. Furthermore, high PRS groups for PTSD and ADHD had lower rates of family history of psychiatric illness and reported fewer previous MDD episodes (p-value < 0.05).

Predicting Treatment Outcomes Using the mPRS Model

We evaluated the predictive capacity of the 11 risk scores when added to a clinical-only model for both remission status and symptom improvement at weeks 8 and 16. For symptom improvement at week 16, the mPRS model achieved similar significant predictive performance to the clinical-only model (root mean square error = 22.08, R2 = 0.13, mean above error = 19.43, Pearson's ρ = 0.36 [0.04, 0.61], p-value = 0.028), see Supplemental Tables 6 to 9.

Discussion and Conclusion

In this exploratory study, we investigated the association between PRSs for major psychiatric disorders and the personality trait neuroticism with ESC and ARI treatment outcomes in patients with MDD from the CAN-BIND-1 clinical trial. We have identified nominal associations of PRSs for PTSD, MDD, anxiety, ADHD, and schizophrenia with remission and symptom improvement.

PTSD is a common, prevalent, and debilitating mental psychiatric disorder that affects individuals following traumatic events. In our study, a higher polygenic risk for PTSD was nominally associated with a lower probability of remission and reduced symptom improvement to ESC monotherapy at week 8. Studies exploring the genetic overlap between PTSD and other psychiatric disorders have reported a moderate shared genetic liability with MDD, with common genetic influences explaining up to 58% of the genetic variation in PTSD.42–44 Furthermore, one GWAS reported a strong positive genetic correlation between PTSD and more severe depressive symptoms (rg = 0.8), as well as between PTSD and MDD (rg = 0.62). 26 This shared genetic liability between both disorders may contribute to a more complex clinical presentation of MDD, affecting both the underlying biology of the depression and variability in treatment response. 14 Interestingly, this nominal association did not persist at week 16, suggesting a potential time-dependent effect of genetic predisposition to PTSD on early antidepressant outcomes.

Furthermore, our results showed a nominal association between higher PRS for MDD and decreased probability of remission at week 8. This is in line with previous findings from various studies showing that higher polygenic loading for MDD is associated with less favourable treatment outcomes. 16 For example, a meta-analysis combining 3 cohorts and including 760 patients reported that a higher PRS for MDD was associated with poorer response to SSRIs, as reflected by the smaller reduction in depression scores after 4 and 8 weeks of antidepressant treatment. 17 Additionally, studies using antidepressant prescription data from population cohorts and the UK Biobank have shown a positive association between PRS for MDD and prescription of more than 2 ADs, indicative of treatment resistance.45,46 Higher PRS for MDD may affect treatment response by influencing symptom severity, with individuals having a higher genetic risk for MDD potentially presenting more severe depressive symptoms.46,47 Interestingly, at week 16, post-hoc stratified analyses revealed an opposite effect: in the ESC-only group, a higher PRS for MDD was significantly associated with an increased probability of remission and greater symptom improvement. These contrasting results suggest that genetic predisposition to depression may influence treatment outcomes differently based on treatment strategy and duration. Specifically, individuals with higher genetic risk may initially present with more severe symptoms and require a longer duration of consistent treatment to achieve meaningful improvement.

Another interesting finding was that a higher PRS for primary anxiety disorders was nominally associated with better symptom improvement (empirical p-value = 0.029) at week 8. Such genetic correlations between PRS for anxiety and SSRI response have not been extensively studied. Notably, a GWAS by Li et al. 48 reported no significant association between PRS for primary anxiety disorders and esketamine response or symptom improvement in 527 patients with treatment-resistant depression. Differences in the patient population and the antidepressant used may explain the discrepancy in results. Since anxiety and depression are highly comorbid 49 and both show a strong positive genetic correlation (rg = 0.78),50,51 genetic liability to anxiety may influence treatment response through shared biological pathways. These pathways mainly involve the serotonin (5-HT) neurotransmitter system targeted by SSRIs, and the hypothalamic–pituitary–adrenal axis regulated by stress-induced corticotropin-releasing factor (CRF). 52 Stress-induced CRF release positively enhances the 5-HT receptor sensitivity, modulating anxiety-related behaviours. 52 Therefore, it could be postulated that individuals with a higher PRS for anxiety, due to their heightened sensitivity to these mechanisms, may respond better to antidepressant medications, as demonstrated in our study.

Our post-hoc stratified analysis by treatment arm at week 16 revealed significant associations in the ESC-only group but not in the augmented group, suggesting a differential genetic influence based on treatment type. In the ESC-only group, the PRSs for schizophrenia and ADHD were significantly associated with lower symptom improvements. Schizophrenia is known to share genetic factors with MDD, and prior studies reported a suggestive association between PRS for schizophrenia and poor antidepressant treatment response.14,16 Similarly for ADHD, a higher polygenic risk for the disorder has been associated with being prescribed multiple ADs. 45 In the recent systematic review by Meerman et al., 16 it was suggested that ADHD could potentially be a useful predictive factor for poor response to ADs and thus warrants further investigations. We also observed that a high PRS for antidepressant percentage improvement (based on Pain et al. 14 GWAS) was negatively associated with remission and symptom improvement at week 16. This could reflect that the Pain et al. GWAS captured genetic variants linked to short-term symptom relief. Additionally, the small sample size of the Pain et al. GWAS may have limited its ability to fully capture the genetic architecture of antidepressant response, contributing to variability in the observed effects. 14 As for the PRSs for bipolar disorder, neuroticism, and SUDs, no significant associations were found, consistent with prior studies.18,45,46,53

The strength of our study is using a well-defined target sample of adults with MDD from the CAN-BIND-1 clinical trial to explore PRS associations with response to ADs. While we observed nominal associations of PRSs for PTSD, MDD, anxiety, schizophrenia, and ADHD with antidepressant treatment outcomes, these findings should be interpreted with caution given these several limitations. First, due to our relatively small sample size (n = 148), we were unable to validate these findings through cross-validation or using an independent sample. 15 However, we did use empirical p-values to mitigate potential overfitting. Second, we recognize that the GWAS base data used to generate PRS may have limitations in terms of phenotype characterization and population diversity. Since most of the GWAS data were derived from European populations, the generalizability of our findings to other ancestries may be limited. Finally, we acknowledge that MDD is a highly heterogeneous disorder, and additional clinical characteristics could influence genetic and biological mechanisms. Future studies incorporating additional clinical risk factors and non-genetic biomarkers of treatment response could help define more homogeneous subtypes of depression, potentially enhancing the predictive utility of PRS.

In conclusion, further research using the PRS approach may identify individuals genetically predisposed to limited antidepressant response. Once validated, PRSs could serve as a tool to inform pharmacotherapy decisions for individuals with MDD.

Supplemental Material

sj-docx-1-cpa-10.1177_07067437251329073 - Supplemental material for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1
sj-docx-1-cpa-10.1177_07067437251329073.docx (1.5MB, docx)

Supplemental material, sj-docx-1-cpa-10.1177_07067437251329073 for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1 by Leen Magarbeh, Samar S. M. Elsheikh, Farhana Islam, Victoria S. Marshe, Xiaoyu Men, Emytis Tavakoli, Martin Kronenbuerger, Stefan Kloiber, Benicio N. Frey, Roumen Milev, Claudio N. Soares, Sagar V. Parikh, Franca Placenza, Stefanie Hassel, Valerie H. Taylor, Francesco Leri, Pierre Blier, Rudolf Uher, Faranak Farzan, Raymond W. Lam, Gustavo Turecki, Jane A. Foster, Susan Rotzinger, Sidney H. Kennedy and Daniel J. Müller in The Canadian Journal of Psychiatry

sj-docx-2-cpa-10.1177_07067437251329073 - Supplemental material for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1
sj-docx-2-cpa-10.1177_07067437251329073.docx (16KB, docx)

Supplemental material, sj-docx-2-cpa-10.1177_07067437251329073 for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1 by Leen Magarbeh, Samar S. M. Elsheikh, Farhana Islam, Victoria S. Marshe, Xiaoyu Men, Emytis Tavakoli, Martin Kronenbuerger, Stefan Kloiber, Benicio N. Frey, Roumen Milev, Claudio N. Soares, Sagar V. Parikh, Franca Placenza, Stefanie Hassel, Valerie H. Taylor, Francesco Leri, Pierre Blier, Rudolf Uher, Faranak Farzan, Raymond W. Lam, Gustavo Turecki, Jane A. Foster, Susan Rotzinger, Sidney H. Kennedy and Daniel J. Müller in The Canadian Journal of Psychiatry

Acknowledgements

We want to acknowledge and thank the study participants who gave their time and trust to research and without whom this study is not possible. We would like to thank the Canadian Biomarker Integration Network in Depression (CAN-BIND) research team who contributed to participant recruitment and data collection. We would also like to thank our CAMH Biobank and Molecular Core Facility for sample genotyping.

Footnotes

Data Availability: Select CAN-BIND-1 data can be accessed through the Home | Brain-CODE Portal (braincode.ca).

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Stefan Kloiber has received an honorarium for past consultation from EmpowerPharm. Roumen Milev has received consulting and speaking honoraria from AbbVie, Allergan, Eisai, Janssen, KYE Pharmaceuticals, Lallemand, Lundbeck, Otsuka, and Sunovion, and research grants from CAN-BIND, Canadian Institutes of Health Research (CIHR), Janssen, Lallemand, Lundbeck, Nubiyota, Ontario Brain Institute (OBI). Sagar V. Parikh has received research support or consulting income from Aifred, Assurex (Myriad), Janssen, Mensante, Otsuka, Sage, Compass, Abbvie Axsome, and Boehringer Ingelhiem. Pierre Blier received honoraria for participation in advisory boards, giving lectures, and/or expert consultation from Allergan, Bristol Myers Squibb, Janssen, Lundbeck, Otsuka, Pierre Fabre Medicaments, Pfizer, and Sunovion; he received grants from Allergan, Janssen, and Lundbeck/Otsuka. Faranak Farzan received funding from the Michael Smith Foundation for Health Research, Natural Sciences and Engineering Research Council of Canada Discovery, and CIHR. Raymond W. Lam has received honoraria for ad hoc speaking or advising/consulting, or received research funds, from Abbvie, Asia-Pacific Economic Cooperation, BC Leading Edge Foundation, CIHR, Canadian Network for Mood and Anxiety Treatments, Carnot, Genome BC, Grand Challenges Canada, Healthy Minds Canada, Janssen, Lundbeck, Michael Smith Foundation for Health Research, MITACS, Neurotorium, Shanghai Mental Health Center, OBI, Otsuka, Pfizer, Unity Health, and VGH-UBCH Foundation. Gustavo Turecki has received an investigator-initiated grant from Pfizer Canada, and honoraria from Bristol-Meyers Squibb Canada and Janssen Canada. Jane A. Foster has served on the Scientific Advisory Board for MRM Health NL and has received consulting/speaker fees from Klaire Labs, Takeda Canada, WebMD and Rothman, Benson, and Hedges Inc. Susan Rotzinger has received grant funding from the OBI and CIHR, and holds a patent Teneurin C-Terminal Associated Peptides and methods and uses thereof. Sidney H. Kennedy has received research funding or honoraria from the following sources: Abbott, Alkermes, Allergan, Bristol-Myers Squibb, Brain Canada, Canadian Institutes for Health Research (CIHR), Janssen, Lundbeck, Lundbeck Institute, OBI, Ontario Research Fund, Otsuka, Pfizer, Servier, and Sunovion. Daniel J. Müller reports to be a coinvestigator on 2 pharmacogenetic studies where genetic test kits were provided as in-kind contributions by Myriad Neuroscience. All other authors have no conflicts of interest to declare.

Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: CAN-BIND is an Integrated Discovery Program carried out in partnership with, and financial support from, the OBI, an independent nonprofit corporation, funded partially by the Ontario government. The opinions, results and conclusions are those of the authors and no endorsement by the OBI is intended or should be inferred. Additional funding is provided by the CIHR, Lundbeck, and Servier. Funding and/or in-kind support are also provided by the investigators’ universities and academic institutions. All study medications are independently purchased at wholesale market values.

Institutional Review Board Statement: CAN-BIND-1 was conducted ethically in accordance with the World Medical Association Declaration of Helsinki of 1975, as revised in 1983. Written informed consent was obtained from all participants, and all study procedures were approved by the ethical review board at each of the participating sites.

Supplemental Material: Supplemental material for this article is available online.

References

  • 1.Kassebaum NJ, Arora M, Barber RM, et al. Global, regional, and national disability-adjusted life-years (DALYs) for 315 diseases and injuries and healthy life expectancy (HALE), 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388(10053):1603–1658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.WHO. Depression: fact sheet. Geneva: World Health Organisation; 2023. https://www.who.int/news-room/fact-sheets/detail/depression .
  • 3.Ferrari AJ, Charlson FJ, Norman RE, et al. Burden of depressive disorders by country, sex, age, and year: findings from the global burden of disease study 2010. PLoS Med. 2013;10(11):e1001547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Gautam S, Jain A, Gautam M, Vahia VN, Grover S. Clinical practice guidelines for the management of depression. Ind J Psychiatry. 2017;59(Suppl 1):S34–S50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Cipriani A, Furukawa TA, Salanti G, et al. Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis. Lancet. 2018;391(10128):1357–1366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Rush AJ, Trivedi MH, Wisniewski SR, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006;163(11):1905–1917. [DOI] [PubMed] [Google Scholar]
  • 7.Trivedi MH, Rush AJ, Wisniewski SR, et al. Evaluation of outcomes with citalopram for depression using measurement-based care in STAR*D: implications for clinical practice. Am J Psychiatry. 2006;163(1):28–40. [DOI] [PubMed] [Google Scholar]
  • 8.Fabbri C, Crisafulli C, Calabro M, Spina E, Serretti A. Progress and prospects in pharmacogenetics of antidepressant drugs. Expert Opin Drug Metab Toxicol. 2016;12(10):1157–1168. [DOI] [PubMed] [Google Scholar]
  • 9.Tansey KE, Guipponi M, Hu X, et al. Contribution of common genetic variants to antidepressant response. Biol Psychiatry. 2013;73(7):679–682. [DOI] [PubMed] [Google Scholar]
  • 10.Garcia-Gonzalez J, Tansey KE, Hauser J, et al. Pharmacogenetics of antidepressant response: a polygenic approach. Prog Neuropsychopharmacol Biol Psychiatry. 2017;75:128–134. [DOI] [PubMed] [Google Scholar]
  • 11.Wray NR, Ripke S, Mattheisen M, et al. Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet. 2018;50(5):668–681. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Investigators G, Investigators M, Investigators SD. Common genetic variation and antidepressant efficacy in major depressive disorder: a meta-analysis of three genome-wide pharmacogenetic studies. Am J Psychiatry. 2013;170(2):207–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Biernacka JM, Sangkuhl K, Jenkins G, et al. The International SSRI Pharmacogenomics Consortium (ISPC): a genome-wide association study of antidepressant treatment response. Transl Psychiatry. 2015;5(4):e553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Pain O, Hodgson K, Trubetskoy V, et al. Identifying the common genetic basis of antidepressant response. Biol Psychiatry Glob Open Sci. 2022;2(2):115–126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Choi SW, Mak TS, O'Reilly PF. Tutorial: a guide to performing polygenic risk score analyses. Nat Protoc. 2020;15(9):2759–2772. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Meerman JJ, Ter Hark SE, Janzing JGE, Coenen MJH. The potential of polygenic risk scores to predict antidepressant treatment response in major depression: a systematic review. J Affect Disord. 2022;304:1–11. [DOI] [PubMed] [Google Scholar]
  • 17.Ward J, Graham N, Strawbridge RJ, et al. Polygenic risk scores for major depressive disorder and neuroticism as predictors of antidepressant response: meta-analysis of three treatment cohorts. PLoS One. 2018;13(9):e0203896. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fanelli G, Benedetti F, Kasper S, et al. Higher polygenic risk scores for schizophrenia may be suggestive of treatment non-response in major depressive disorder. Prog Neuropsychopharmacol Biol Psychiatry. 2021;108:110170. [DOI] [PubMed] [Google Scholar]
  • 19.Marshe VS, Maciukiewicz M, Hauschild AC, et al. Genome-wide analysis suggests the importance of vascular processes and neuroinflammation in late-life antidepressant response. Transl Psychiatry. 2021;11(1):127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Fabbri C, Kasper S, Kautzky A, et al. Genome-wide association study of treatment-resistance in depression and meta-analysis of three independent samples. Br J Psychiatry. 2019;214(1):36–41. [DOI] [PubMed] [Google Scholar]
  • 21.Kennedy SH, Lam RW, Rotzinger S, et al. Symptomatic and functional outcomes and early prediction of response to escitalopram monotherapy and sequential adjunctive aripiprazole therapy in patients with major depressive disorder: a CAN-BIND-1 report. J Clin Psychiatry. 2019;80(2):18m12202. [DOI] [PubMed] [Google Scholar]
  • 22.Mullins N, Forstner AJ, O'Connell KS, et al. Genome-wide association study of more than 40,000 bipolar disorder cases provides new insights into the underlying biology. Nat Genet. 2021;53(6):817–829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Trubetskoy V, Pardinas AF, Qi T, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. 2022;604(7906):502–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Demontis D, Walters GB, Athanasiadis G, et al. Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains. Nat Genet. 2023;55(2):198–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Otowa T, Hek K, Lee M, et al. Meta-analysis of genome-wide association studies of anxiety disorders. Mol Psychiatry. 2016;21(10):1391–1399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Nievergelt CM, Maihofer AX, Klengel T, et al. International meta-analysis of PTSD genome-wide association studies identifies sex- and ancestry-specific genetic risk loci. Nat Commun. 2019;10(1):4558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Hatoum AS, Colbert SMC, Johnson EC, et al. Multivariate genome-wide association meta-analysis of over 1 million subjects identifies loci underlying multiple substance use disorders. Nat Ment Health. 2023;1(3):210–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Genetics of Personality Consortium, de Moor MH, van den Berg SM, Verweij KJ, et al. Meta-analysis of genome-wide association studies for neuroticism, and the polygenic association with major depressive disorder. JAMA Psychiatry. 2015;72(7):642–650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Milaneschi Y, Lamers F, Peyrot WJ, et al. Polygenic dissection of major depression clinical heterogeneity. Mol Psychiatry. 2016;21(4):516–522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Krapohl E, Patel H, Newhouse S, et al. Multi-polygenic score approach to trait prediction. Mol Psychiatry. 2018;23(5):1368–1374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Kautzky A, Baldinger P, Souery D, et al. The combined effect of genetic polymorphisms and clinical parameters on treatment outcome in treatment-resistant depression. Eur Neuropsychopharmacol. 2015;25(4):441–453. [DOI] [PubMed] [Google Scholar]
  • 32.American Psychiatric Association. Diagnostic and statistical manual for mental disorders: DSM-IV-TR. Washington, DC: American Psychiatric Association; 2000. [Google Scholar]
  • 33.Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry. 1998;59(Suppl 20):22–33; quiz 4-57. [PubMed] [Google Scholar]
  • 34.Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979;134:382–389. [DOI] [PubMed] [Google Scholar]
  • 35.Lam RW, Milev R, Rotzinger S, et al. Discovering biomarkers for antidepressant response: protocol from the Canadian biomarker integration network in depression (CAN-BIND) and clinical characteristics of the first patient cohort. BMC Psychiatry. 2016;16:105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015;4:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Anderson CA, Pettersson FH, Clarke GM, Cardon LR, Morris AP, Zondervan KT. Data quality control in genetic case-control association studies. Nat Protoc. 2010;5(9):1564–1573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Clarke GM, Anderson CA, Pettersson FH, Cardon LR, Morris AP, Zondervan KT. Basic statistical analysis in genetic case-control studies. Nat Protoc. 2011;6(2):121–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Sullivan PF, Agrawal A, Bulik CM, et al. Psychiatric genomics: an update and an agenda. Am J Psychiatry. 2018;175(1):15–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Cinar O, Viechtbauer W. The poolr package for combining independent and dependent $p$ values. J Stat Softw. 2022;101(1):1–42. [Google Scholar]
  • 41.Nyholt DR. A simple correction for multiple testing for single-nucleotide polymorphisms in linkage disequilibrium with each other. Am J Hum Genet. 2004;74(4):765–769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Koenen KC, Fu QJ, Ertel K, et al. Common genetic liability to major depression and posttraumatic stress disorder in men. J Affect Disord. 2008;105(1-3):109–115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Duncan LE, Ratanatharathorn A, Aiello AE, et al. Largest GWAS of PTSD (N = 20 070) yields genetic overlap with schizophrenia and sex differences in heritability. Mol Psychiatry. 2018;23(3):666–673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Mundy J, Hubel C, Gelernter J, et al. Psychological trauma and the genetic overlap between posttraumatic stress disorder and major depressive disorder. Psychol Med. 2021;52(16):1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Fabbri C, Hagenaars SP, John C, et al. Genetic and clinical characteristics of treatment-resistant depression using primary care records in two UK cohorts. Mol Psychiatry. 2021;26(7):3363–3373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Wigmore EM, Hafferty JD, Hall LS, et al. Genome-wide association study of antidepressant treatment resistance in a population-based cohort using health service prescription data and meta-analysis with GENDEP. Pharmacogenomics J. 2020;20(2):329–341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Halldorsdottir T, Piechaczek C, Soares de Matos AP, et al. Polygenic risk: predicting depression outcomes in clinical and epidemiological cohorts of youths. Am J Psychiatry. 2019;176(8):615–625. [DOI] [PubMed] [Google Scholar]
  • 48.Li QS, Wajs E, Ochs-Ross R, Singh J, Drevets WC. Genome-wide association study and polygenic risk score analysis of esketamine treatment response. Sci Rep. 2020;10(1):12649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Davies MR, Kalsi G, Armour C, et al. The Genetic Links to Anxiety and Depression (GLAD) study: online recruitment into the largest recontactable study of depression and anxiety. Behav Res Ther. 2019;123:103503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Levey DF, Gelernter J, Polimanti R, et al. Reproducible genetic risk loci for anxiety: results from approximately 200,000 participants in the million veteran program. Am J Psychiatry. 2020;177(3):223–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Purves KL, Coleman JRI, Meier SM, et al. A major role for common genetic variation in anxiety disorders. Mol Psychiatry. 2020;25(12):3292–3303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Magalhaes AC, Holmes KD, Dale LB, et al. CRF Receptor 1 regulates anxiety behavior via sensitization of 5-HT2 receptor signaling. Nat Neurosci. 2010;13(5):622–629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Tansey KE, Guipponi M, Perroud N, et al. Genetic predictors of response to serotonergic and noradrenergic antidepressants in major depressive disorder: a genome-wide analysis of individual-level data and a meta-analysis. PLoS Med. 2012;9(10):e1001326. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

sj-docx-1-cpa-10.1177_07067437251329073 - Supplemental material for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1
sj-docx-1-cpa-10.1177_07067437251329073.docx (1.5MB, docx)

Supplemental material, sj-docx-1-cpa-10.1177_07067437251329073 for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1 by Leen Magarbeh, Samar S. M. Elsheikh, Farhana Islam, Victoria S. Marshe, Xiaoyu Men, Emytis Tavakoli, Martin Kronenbuerger, Stefan Kloiber, Benicio N. Frey, Roumen Milev, Claudio N. Soares, Sagar V. Parikh, Franca Placenza, Stefanie Hassel, Valerie H. Taylor, Francesco Leri, Pierre Blier, Rudolf Uher, Faranak Farzan, Raymond W. Lam, Gustavo Turecki, Jane A. Foster, Susan Rotzinger, Sidney H. Kennedy and Daniel J. Müller in The Canadian Journal of Psychiatry

sj-docx-2-cpa-10.1177_07067437251329073 - Supplemental material for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1
sj-docx-2-cpa-10.1177_07067437251329073.docx (16KB, docx)

Supplemental material, sj-docx-2-cpa-10.1177_07067437251329073 for Polygenic Risk Score Analysis of Antidepressant Treatment Outcomes: A CAN-BIND-1 Study Report: Analyse des résultats du traitement antidépresseur à l’aide des scores de risque polygéniques : Rapport sur l’étude CAN-BIND-1 by Leen Magarbeh, Samar S. M. Elsheikh, Farhana Islam, Victoria S. Marshe, Xiaoyu Men, Emytis Tavakoli, Martin Kronenbuerger, Stefan Kloiber, Benicio N. Frey, Roumen Milev, Claudio N. Soares, Sagar V. Parikh, Franca Placenza, Stefanie Hassel, Valerie H. Taylor, Francesco Leri, Pierre Blier, Rudolf Uher, Faranak Farzan, Raymond W. Lam, Gustavo Turecki, Jane A. Foster, Susan Rotzinger, Sidney H. Kennedy and Daniel J. Müller in The Canadian Journal of Psychiatry

Articles from Canadian Journal of Psychiatry. Revue Canadienne de Psychiatrie are provided here courtesy of SAGE Publications

RESOURCES