Skip to main content
NCBI home page
Biological Psychiatry Global Open Science logoLink to Biological Psychiatry Global Open Science
. 2025 Jun 25;5(5):100558. doi: 10.1016/j.bpsgos.2025.100558

Pathway-Specific Polygenic Scores for Predicting Clinical Lithium Treatment Response in Patients With Bipolar Disorder

Nigussie T Sharew 1,2, Scott R Clark 1, Sergi Papiol 3,4, Urs Heilbronner 3, Franziska Degenhardt 5,6, Janice M Fullerton 7,8, Liping Hou 9, Tatyana Shekhtman 10, Mazda Adli 11, Nirmala Akula 9, Kazufumi Akiyama 12, Raffaella Ardau 13, Bárbara Arias 14, Roland Hasler 15, Hélène Richard-Lepouriel 15, Nader Perroud 15, Lena Backlund 16,17, Abesh Kumar Bhattacharjee 18, Frank Bellivier 19, Antonio Benabarre 20, Susanne Bengesser 21, Joanna M Biernacka 22,23, Armin Birner 21, Cynthia Marie-Claire 19,24, Pablo Cervantes 25, Hsi-Chung Chen 26, Caterina Chillotti 13, Sven Cichon 27,28,29, Cristiana Cruceanu 30, Piotr M Czerski 31, Nina Dalkner 21, Maria Del Zompo 32, J Raymond DePaulo 33, Bruno Étain 19, Stephane Jamain 34, Peter Falkai 4,35, Andreas J Forstner 5,29, Louise Frisen 16,36, Mark A Frye 23, Sébastien Gard 37, Julie S Garnham 38, Fernando S Goes 33, Maria Grigoroiu-Serbanescu 38, Andreas J Fallgatter 39, Sophia Stegmaier 40, Thomas Ethofer 41,42, Silvia Biere 43, Kristiyana Petrova 43, Ceylan Schuster 43, Kristina Adorjan 3,44, Monika Budde 3, Maria Heilbronner 3, Janos L Kalman 3,4, Mojtaba Oraki Kohshour 3,45, Daniela Reich-Erkelenz 3, Sabrina K Schaupp 3, Eva C Schulte 3,4, Fanny Senner 3,4, Thomas Vogl 3, Ion-George Anghelescu 46, Volker Arolt 47, Udo Dannlowski 47, Detlef E Dietrich 48,49, Christian Figge 50, Markus Jäger 51, Fabian U Lang 51, Georg Juckel 52, Carsten Konrad 53, Jens Reimer 54,55, Max Schmauß 56, Andrea Schmitt 4,57, Carsten Spitzer 58, Martin von Hagen 59, Jens Wiltfang 60,61, Jörg Zimmermann 62, Till FM Andlauer 63, Andre Fischer 61, Felix Bermpohl 11, Philipp Ritter 64, Silke Matura 43, Anna Gryaznova 3, Irina Falkenberg 65, Cüneyt Yildiz 65, Tilo Kircher 65, Julia Schmidt 66, Marius Koch 66, Kathrin Gade 60, Sarah Trost 60, Ida S Haussleiter 52, Martin Lambert 54, Anja C Rohenkohl 54, Vivien Kraft 54, Paul Grof 67, Ryota Hashimoto 68, Joanna Hauser 31, Stefan Herms 5,28, Per Hoffmann 5,28, Esther Jiménez 20, Jean-Pierre Kahn 69, Layla Kassem 11, Po-Hsiu Kuo 70, Tadafumi Kato 71, John Kelsoe 18, Sarah Kittel-Schneider 72,73, Ewa Ferensztajn-Rochowiak 74, Barbara König 75, Ichiro Kusumi 76, Gonzalo Laje 11, Mikael Landén 77,78, Catharina Lavebratt 16,17, Marion Leboyer 79, Susan G Leckband 80, Alfonso Tortorella 81, Mirko Manchia 82,83, Lina Martinsson 84, Michael J McCarthy 18,85, Susan McElroy 86, Francesc Colom 87,88, Vincent Millischer 16,17,89, Marina Mitjans 88,90,91,92, Francis M Mondimore 33, Palmiero Monteleone 93,94, Caroline M Nievergelt 18, Markus M Nöthen 5, Tomas Novák 95, Claire O’Donovan 38, Norio Ozaki 96, Andrea Pfennig 64, Claudia Pisanu 32, James B Potash 33, Andreas Reif 72, Eva Reininghaus 21, Guy A Rouleau 97, Janusz K Rybakowski 74, Martin Schalling 16,17, Peter R Schofield 7,8, Barbara W Schweizer 33, Giovanni Severino 32, Paul D Shilling 18, Katzutaka Shimoda 98, Christian Simhandl 99, Claire M Slaney 38, Alessio Squassina 32, Thomas Stamm 11,100, Pavla Stopkova 95, Mario Maj 94, Gustavo Turecki 30, Eduard Vieta 20, Julia Veeh 77, Biju Viswanath 101, Stephanie H Witt 102, Adam Wright 103, Peter P Zandi 104, Philip B Mitchell 103, Michael Bauer 64, Martin Alda 38,95, Marcella Rietschel 102, Francis J McMahon 9, Thomas G Schulze 3,9,33,102,105,106, Bernhard T Baune 107,108,109, Klaus Oliver Schubert 1,110, Azmeraw T Amare 1,∗
PMCID: PMC12357302  PMID: 40821394

Abstract

Background

Polygenic scores (PGSs) hold the potential to identify patients who respond favorably to specific psychiatric treatments. However, their biological interpretation remains unclear. In this study, we developed pathway-specific PGSs (PSPGSs) for lithium response and assessed their association with clinical lithium response in patients with bipolar disorder.

Methods

Using sets of genes involved in pathways affected by lithium, we developed 9 PSPGSs and evaluated their associations with lithium response in the International Consortium on Lithium Genetics (ConLi+Gen) (N = 2367), with validation in combined PsyCourse (Pathomechanisms and Signatures in the Longitudinal Course of Psychosis) (N = 105) and BipoLife (N = 102) cohorts. The association between each PSPGS and lithium response—defined both as a continuous ALDA score and a categorical outcome (good vs. poor responses)—was evaluated using regression models, with adjustment for confounders. The cutoff for a significant association was p < .05 after multiple testing correction.

Results

The PGSs for acetylcholine, GABA (gamma-aminobutyric acid), and mitochondria were associated with response to lithium in both categorical and continuous outcomes. However, the PGSs for calcium channel, circadian rhythm, and GSK (glycogen synthase kinase) were associated only with the continuous outcome. Each score explained 0.29% to 1.91% of the variance in the categorical and 0.30% to 1.54% of the variance in the continuous outcomes. A multivariate model combining PSPGSs that showed significant associations in the univariate analysis (combined PSPGS) increased the percentage of variance explained (R2) to 3.71% and 3.18% for the categorical and continuous outcomes, respectively. Associations for PGSs for GABA and circadian rhythm were replicated. Patients with the highest genetic loading (10th decile) for acetylcholine variants were 3.03 times more likely (95% CI, 1.95 to 4.69) to show a good lithium response (categorical outcome) than patients with the lowest genetic loading (1st decile).

Conclusions

PSPGSs achieved predictive performance comparable to the conventional genome-wide PGSs, with the added advantage of biological interpretability using a smaller list of genetic variants.

Keywords: Bipolar disorder, Lithium, Pharmacogenomics, Polygenic score, Psychiatry

Plain Language Summary

Polygenic scores (PGSs) have the potential to identify patients likely to respond to specific psychiatric treatments, but their biological interpretation remains unclear. In this study, we developed 9 pathway-specific PGSs (PSPGSs) for lithium response by aggregating genetic variants involved in pathways affected by lithium. We assessed their associations with lithium response in the International Consortium on Lithium Genetics (ConLi+Gen) (N = 2367) cohort and validated the findings in the PsyCourse (N = 105) and BipoLife (N = 102) cohorts.

Clinical response to lithium treatment was significantly associated with PSPGSs for acetylcholine, GABA (gamma-aminobutyric acid), calcium channel signaling, mitochondria, circadian rhythm, and GSK pathways, with explained variance (R2) ranging from 0.29% to 1.91%. The combined PSPGS explained up to 3.71% of the variability. Associations for GABA and circadian rhythm PGSs were successfully replicated. In a decile-based analysis, patients with the highest genetic load (10th decile) for acetylcholine pathway variants were 3.03 times more likely to respond well to lithium compared with those in the lowest decile (1st decile).

PSPGSs achieved predictive performance comparable to conventional genome-wide PGSs, with better biological interpretability and a more focused set of genetic variants.

Plain Language Summary

Polygenic scores (PGSs) have the potential to identify patients likely to respond to specific psychiatric treatments, but their biological interpretation remains unclear. In this study, we developed 9 pathway-specific PGSs (PSPGSs) for lithium response by aggregating genetic variants involved in pathways affected by lithium. We assessed their associations with lithium response in the International Consortium on Lithium Genetics (ConLi+Gen) (N = 2367) cohort and validated the findings in the PsyCourse (N = 105) and BipoLife (N = 102) cohorts.

Clinical response to lithium treatment was significantly associated with PSPGSs for acetylcholine, GABA (gamma-aminobutyric acid), calcium channel signaling, mitochondria, circadian rhythm, and GSK pathways, with explained variance (R2) ranging from 0.29% to 1.91%. The combined PSPGS explained up to 3.71% of the variability. Associations for GABA and circadian rhythm PGSs were successfully replicated. In a decile-based analysis, patients with the highest genetic load (10th decile) for acetylcholine pathway variants were 3.03 times more likely to respond well to lithium compared with those in the lowest decile (1st decile).

PSPGSs achieved predictive performance comparable to conventional genome-wide PGSs, with better biological interpretability and a more focused set of genetic variants.

Over the past 15 years, there has been significant progress in the development of polygenic scores (PGSs). Key research areas have been centered around evaluating their potential for disease risk prediction, uncovering the genetic basis of complex diseases, and clinical application for disease screening and drug selection through pharmacogenomics, as well as assessing their cross-population transferability (1, 2, 3, 4, 5).

Since the initial implementation of the polygenic theory for assessing the genetic risk of schizophrenia (SCZ) by the International Schizophrenia Consortium in 2009 (1), hundreds of PGSs have been developed and investigated for their association with the risk of common mental health disorders such as SCZ (6), major depressive disorder (MDD) (7), and bipolar disorder (BD) (8). In recent years, PGSs have emerged as a promising tool for understanding the collective influence of common single nucleotide polymorphisms (SNPs) on patients’ pharmacological treatment outcomes (9,10). For example, in patients with SCZ, a PGS for SCZ (PGSSCZ) that was significantly associated with antipsychotic treatment outcomes explained 3.2% of the interindividual variability in treatment response (11). Other studies have shown that a PGSSCZ explained 2.0% of the variance in treatment-resistant SCZ (12, 13, 14, 15), ∼1% of the variance in antipsychotic-induced weight gain (16,17), nearly 2% of the variance in clozapine-induced myocarditis (18), and 2.7% of the variance in prolonged hospitalization (19). Similarly, in patients with MDD, genetic scores for SCZ, MDD, BD, and neuroticism showed significant associations with antidepressant treatment response (20, 21, 22, 23) and resistance (24,25), although each of these scores explained <2% of the variability. Among patients with BD, lithium response was associated with PGSs for SCZ (26), MDD (27), attention-deficit/hyperactivity disorder (28), and lithium responsiveness (29). Combined analysis of SCZ and MDD PGSs with clinical variables resulted in better prediction, with the model accounting for approximately 14% of the variance in lithium treatment response (30), emphasizing the potential clinical relevance of algorithms that combine PGSs with clinical data. This result exceeds the accuracy of any PGSs that have been analyzed individually or in combination using standard measures, which at best have explained up to 5.6% of the variance in psycho-pharmacotherapeutic outcomes, e.g., in resistance to clozapine (31). While PGSs hold significance for research purposes and offer promising clinical implications for the future, their predictive performance remains limited for direct clinical translation (10). Thus, there is a need to utilize novel methods to develop PGSs with better predictive capabilities and to refine existing scores for increased precision.

In this context, newly proposed approaches such as biology-informed polygenic modeling have been evaluated for various traits (32, 33, 34). This PGS approach leverages genetic variants based on their relationship to molecular pathways that are linked to the phenotype of interest, thereby enhancing their predictive power and relevance to pharmacogenomics or disease screening (35). For example, an insulin receptor–based PGS targeting the striatum and prefrontal cortex predicted impulsivity and cognitive abilities in children, as well as addiction and dementia risk in adults (32). Similarly, a PGS composed of variants associated with nervous system development and neuron differentiation explained 6.9% of the variance in liability to psychosis in a sample of patients with DSM-IV diagnoses of SCZ or psychosis-related disorders determined by a structured clinical interview. This result surpasses the 3.7% of the variance explained by a conventional PGSSCZ using genome-wide variants (34). Thus, restricting PGSs to genetic variants within biological pathways known to be associated with lithium response may reduce noise from variants with spurious associations and increase the power of polygenic models while explicitly building our mechanistic understanding (36). Furthermore, the biology-informed polygenic approach may facilitate the effort to identify new treatment targets (33,37).

Building on this knowledge, our study adopted the biologically informed strategy and developed pathway-specific PGSs (PSPGSs) for lithium response. We hypothesized that these scores would improve the prediction of clinical response to lithium and could identify biological targets for future drug development in patients with BD.

Methods and Materials

Study Sample Characteristics

The target data for this study were obtained from the International Consortium on Lithium Genetics (ConLi+Gen) cohort (http://www.conligen.org/), a global initiative established to investigate the genetic underpinnings of lithium treatment response in patients with BD. The discovery and target sample included only patients of European ancestry (N = 2367) who received lithium and were followed up for at least 6 months (38). The number of participants in each country is described in our previous study (29).

To replicate the findings from ConLi+Gen, we utilized combined data from 2 German cohorts: the PsyCourse (Pathomechanisms and Signatures in the Longitudinal Course of Psychosis) study (N = 105) (39,40) and BipoLife (N = 102) cohorts (41). A detailed sample selection procedure for the replication cohorts is included in Supplemental Methods.

Target Outcome Measure

For both target and replication cohorts, the validated retrospective criteria for long-term treatment response in research subjects, known as the ALDA scale, was used to assess patient’s response to lithium treatment (42,43). This score quantifies the degree of improvement during lithium response expressed as a composite measure of change in frequency and severity of mood symptoms (A score). The ALDA scale is adjusted for 5 potential confounding factors that could affect symptom improvement (B scale). These factors include the number (B1) and the frequency (B2) of disease episodes before/off the treatment, the duration of the treatment (B3), and compliance and use of additional medication during the periods of stability. The total ALDA score for each individual was calculated by subtracting the total B score from the total A score. The target outcome, lithium response, was defined as categorical (good response vs. poor response) and continuous outcomes. For the categorical outcome, patients who had a total score ≥7 were classified as good responders, and patients with a score <7 were classified as poor responders (44). The total ALDA score was used as a continuous lithium response measure after excluding patients with B scores >4 or who had missing data. Negative scores were recalibrated as 0. This algorithm has been used in previous studies (26,27,30,45,46) and described in detail elsewhere (44).

Genotyping, Quality Control, and Imputation Procedures for the ConLi+Gen Sample

DNA was extracted from peripheral blood samples collected at 22 participating sites, and samples were genotyped using either Affymetrix or Illumina SNP arrays (44). Prior to imputation, quality control (QC) procedures were implemented on the genotype data using PLINK version 1.9 (47). SNPs with a poor genotyping rate (<95%), strand ambiguity (A/T and C/G SNPs), and a minor allele frequency (MAF) <10% and SNPs that deviated from Hardy-Weinberg equilibrium (p < 10−6) were removed. Individuals with sex inconsistencies between the documented and genotype-derived sex and genetically related individuals were also excluded. The genotypic and QC details of the ConLi+Gen cohort are available elsewhere (44).

The genotype data that passed QC were imputed in the Michigan server separately for each genotyping platform using the Haplotype Reference Consortium reference panel comprising broadly European haplotypes at 39,235,157 SNPs (48). For each cohort, imputation quality procedures were implemented and excluded SNPs of low frequency (MAF < 1%) and low quality (imputation quality score R2 < 0.6). Then, genotype calls for the filtered SNPs were derived and merged using PLINK from the imputed dosage score (47). The genotyping, QC, and imputation procedures for the replication cohorts are provided in Supplemental Methods.

Steps of Developing Pathway-Specific Polygenic Scores

Step 1: Identify Biological Pathways (Targets) of Lithium

To develop a PSPGS, we first conducted a narrative review to identify biological pathways or processes potentially modulated by lithium. In this review, we identified 9 pathways including acetylcholine, GABA (gamma-aminobutyric acid), glutamate, dopamine, calcium channels, mitochondria, circadian rhythm, GSK (glycogen synthase kinase), and NMDA as potential targets for lithium in BD treatment (Supplement).

Step 2: Map Genes and SNPs for Each Biological Pathway

Using the names of pathways relevant to lithium as a search term, we extracted candidate genes in 3 existing databases, specifically Gene Set Enrichment Analysis (https://www.gsea-msigdb.org/gsea/index.jsp), HUGO Gene Nomenclature Committee (https://www.genenames.org/), and Kyoto Encyclopedia of Genes and Genomes (https://www.genome.jp/kegg/). The extracted lists of genes for each pathway are provided in the Supplement. We used MAGMA software (https://ctg.cncr.nl/software/magma) (49), with –annotate window = 100, 20 (100 kb upstream and 20 kb downstream window), to annotate SNPs to these genes in each pathway. The final list of annotated SNPs that were matched with the target dataset (ConLi+Gen) and included in our analysis were acetylcholine (6247), GABA (2994), glutamate (3840), dopamine (5794), calcium channel (4236), mitochondria (7801), circadian rhythm (6673), GSK (707) and NMDA (641). The lists of genes and SNPs that were included in the final analysis are provided in Supplemental Data. We note that some of the genes/SNPs overlap across pathways.

Step 3: Compute Pathway-Specific Polygenic Scores

To compute PSPGSs for participants in 13 countries involved in the ConLi+Gen, we implemented a widely accepted leave-one-country-out (LOC) procedure (50,51) in which PGSs were calculated for participants of one country at a time (target sample) using genome-wide association study (GWAS) summary statistics from the remaining 12 countries (discovery sample). This iterative procedure was conducted separately for the categorical and continuous measures of lithium response, which resulted in a total of 26 analyses. Each discovery GWAS was performed using PLINK, with regression models adjusted for age, sex, chip type, and the first 4 principal components (PCs). Each of the PSPGSs was computed using the polygenic risk score with continuous shrinkage (PRS-CS) method, which incorporates continuous shrinkage priors on effect sizes and accounts for linkage disequilibrium among SNPs (52). In the replication analysis, a summary GWAS data from the full ConLi+Gen was used as the discovery sample (29) to compute PSPGSs in the combined PsyCourse (39,40) and BipoLife (41) samples. Additional details on the development of PSPGSs are provided in Supplemental Methods.

Step 4: Association Analysis

Finally, the associations between each of the PSPGSs and lithium response were evaluated using linear regression analysis for the continuous outcome and binary logistic regression analysis for the categorical outcome. Each association analysis was adjusted for age, sex, chip type, and the first 4 PCs. The cutoff for a statistically significant association was p < .05 after correction for multiple testing using the Benjamini-Hochberg procedure (53). To evaluate the combined effect of multiple PSPGSs on lithium response, we utilized a multivariate regression model considering only PSPGSs that showed a significant association with lithium response in the univariate model. The performance of this combined PSPGS model was compared with the conventional genome-wide PGS model that uses genome-wide variants of lithium responsiveness. This analysis was conducted using r2redux R package (54). We also performed elastic-net regularization with 5-fold nested cross-validation in the continuous outcome model to mitigate potential overestimation of the performance of combined PSPGSs in the multivariate analysis using ordinary least squares regression. Furthermore, we implemented a stratified analysis by dividing the ConLi+Gen sample into deciles, ranging from the lowest to the highest polygenic loading for each PSPGS. The proportion of phenotypic variance explained (R2) by each PSPGS was calculated as the difference in R2 between the model fit with specific PGSs and covariates and the model with only covariates. For the categorical outcome, McFadden’s pseudo R2 was calculated as a measure of model performance (55). The observed R2 values were subsequently transformed to the liability (56), assuming a lithium responsiveness prevalence of 30% (44,57) and responders to nonresponders ratio within both the target and replication cohorts.

Figure 1 shows examples of potential targets of lithium (pathways) and detailed steps of the data analysis process.

Figure 1.

Figure 1

Open in a new tab

Examples of potential targets of lithium (pathways) and detailed steps of the data analysis process. Ach, acetylcholine; Ca2+, calcium ion; GABA, gamma-aminobutyric acid; GSEA, Gene Set Enrichment Analysis; GSK, glycogen synthase kinase; GWAS, genome-wide association study; HGNC, HUGO Gene Nomenclature Committee; KEGG, Kyoto Encyclopedia of Genes and Genomes; Li+, lithium; LOC, leave-one-country-out; PRS-CS, polygenic risk score with continuous shrinkage; PSPGS, pathway-specific polygenic score; SNP, single nucleotide polymorphism.

Results

Description of Study Participants

Our discovery analysis included data from 2367 patients with BD treated with lithium for at least 6 months. Nearly 60% of the participants were female; the mean age (SD) was 47.53 (13.73) years. Six hundred sixty (27.9%) patients had a good response to lithium treatment (defined as an ALDA score ≥7), and the mean (SD) ALDA total score was 4.12 (3.15). Among the replication cohort participants, 48.0% of the BipoLife participants and 40% of the PsyCourse participants were female. About 28% of patients in the BipoLife cohort and 23% of patients in the PsyCourse cohort had a good response to lithium treatment (Table 1).

Table 1.

Characteristics of the Study Cohorts and Participants

Cohort N Sex, Female Age (SD) ALDA Score, Mean (SD) Good Response to Lithium, N (%)
ConLi+Gen 2367 1369 47.53 (13.73) 4.12 (3.15) 660 (27.88%)
BipoLife 102 49 49.87 (13.62) 4.52 (2.93) 29 (28.43%)
PsyCourse 105 42 46.91 (12.64) 3.80 (2.87) 24 (22.86%)
Open in a new tab

ConLi+Gen, International Consortium on Lithium Genetics; PsyCourse, Pathomechanisms and Signatures in the Longitudinal Course of Psychosis.

Associations of Pathway-Specific Polygenic Scores With Clinical Lithium Treatment Response

BD patients with higher PGSs for acetylcholine genetic variants (AChPGS) were more likely to have a good lithium treatment response than BD patients with lower PGSs; adjusted odds ratio (aOR) = 1.34 (95% CI, 1.22 to 1.49; p = 3.54 × 10−8; pseudo R2 = 1.91%) for the categorical outcome and β = 0.38 (95% CI, 0.26 to 0.51; p = 3.24 × 10−9; R2 = 1.56%) for the continuous outcome (Table 2). In the stratified analysis, patients with the highest genetic loading for ACh variants (10th decile) were 3.03 times more likely (95% CI, 1.95 to 4.69) to have a good lithium response than patients with the lowest genetic loading (1st decile) (Figure 2). Similarly, BD patients with higher PGSs for GABA genetic variants (GABAPGS) were more likely to have a good lithium treatment response than BD patients with lower GABAPGS; aOR = 1.15 (95% CI, 1.05 to 1.27; p = 9.14 × 10−3; pseudo R2 = 0.34%) for the categorical outcome and β = 0.15 (95% CI, 0.03 to 0.28; p = .03; R2 = 0.30%) for the continuous outcome (Table 2). In the stratified analysis, patients with the highest genetic loading for GABA variants (10th decile) were 2.01 times more likely (95% CI, 1.30 to 3.09) to have a good lithium treatment response than patients with the lowest genetic loading (1st decile) (Figure 2). Higher PGSs for calcium channel variants (Ca2+PGS) in BD patients were significantly associated with the continuous lithium response (p = .01, R2 = 0.3%) but not with the categorical measure (p = .11) (Table 2). The PGS for GSK genetic variants (GSKPGS) and the PGS for circadian rhythm genetic variants (CIRPGS) were positively associated with better lithium response—with continuous outcome (p = 1.84 × 10−4; R2 = 0.6%) and (p = .01, R2 = 0.2%), respectively, but not with the categorical outcome. In contrast, the increased genetic variance within mitochondria genes was associated with poorer lithium response—categorical (p = 7.42 × 10−4; pseudo R2 = 1.05%) and the continuous outcomes (p = 1.02 × 10−4; R2 = 1.16%) in the PGS for mitochondria genetic variants (MITOPGS). With a decreasing trend across deciles, BD patients with the highest genetic loadings for mitochondria variants (10th decile) had 52% lower odds of responding to lithium than patients with the lowest genetic loadings (1st decile); aOR = 0.52 (95% CI, 0.33 to 0.80) (Figure 2). The remaining PSPGSs were not significant. The full stratified analysis is presented in the Supplement.

Table 2.

The Association of Pathway-Specific Polygenic Scores With Clinical Lithium Treatment Response in Patients With Bipolar Disorder (N = 2367)

Pathways No. of Genes No. of SNPs Categorical Outcome
 Continuous ALDA Score
aOR (95% CI) p Pseudo R2, % β (95% CI) p R2, %
Acetylcholine 164 6247 1.34 (1.22 to 1.49) 3.54 × 10−8a 1.91% 0.38 (0.26 to 0.51) 3.24 × 10−9a 1.54%
GABA Receptor 76 2994 1.15 (1.05 to 1.27) 9.14 × 10−3a 0.34%a 0.15 (0.03 to 0.28) .03a 0.30%
Calcium Channel 134 4236 1.10 (0.99 to 1.21) .11 0.29% 0.18 (0.06 to 0.31) .01a 0.30%
Mitochondria 163 7801 0.82 (0.74 to 0.91) 7.42 × 10−4a 1.05% −0.19 (−0.30 to −0.09) 1.02 × 10−4a 1.16%
Glutamate 92 3840 0.98 (0.89 to 1.08) .77 0.25% 0.09 (−0.03 to 0.22) .21 0.20%
Circadian Rhythm 129 6673 1.09 (0.99 to 1.21) .10 0.62% 0.16 (0.04 to 0.29) .01 0.20%
Dopamine 155 5794 1.03 (0.94 to 1.14) .59 0.01% 0.12 (−0.01 to 0.23) .11 0.01%
GSK 18 707 1.09 (0.99 to 1.21) .15 0.15% 0.24 (0.13 to 0.37) 1.85 × 10−4a 0.60%
NMDA 11 641 0.93 (0.84 to 1.02) .14 0.01% 0.04 (−0.09 to 0.17) .67 0.01%
Combined PSPGS 942 38,933 NA <.01a 3.71% NA <.01a 3.18%
Genome-Wide PGS 1.14 (1.03 to 1.27) 4.13 × 10−9a 2.87% 0.45 (0.31 to 0.57) 2.41 × 10−7a 2.69%
Open in a new tab

A statistically significant association was determined at p < .05 after correction for Benjamini-Hochberg multiple testing. The observed R2 in the categorical outcome is transformed to a liability scale. The analysis models were adjusted for age, sex, chip type, and the first 4 principal components. R2 represents the variance explained by polygenic scores. The combined PSPGS represents a multivariate analysis of PSPGSs with p values < .05 in the univariate analysis at least either of lithium response, i.e., acetylcholine, GABA, calcium signaling, mitochondria, circadian rhythm, and GSK potential pathways. The genome-wide PGS represents a PGS for genome-wide variants of lithium responsiveness for lithium treatment response in a leave-one-country-out procedure.

aOR, adjusted odds ratio; GABA, gamma-aminobutyric acid; GSK, glycogen synthase kinase; NA, not applicable; PSPGS, pathway-specific polygenic score; SNP, single nucleotide polymorphism.

a

Denotes significant association after correction for multiple tests.

Figure 2.

Figure 2

Open in a new tab

Trends in the odds ratios for favorable lithium treatment response [with the categorical (A) and beta coefficeints in the continuous (B) outcomes in patients with bipolar disorder, comparing patients with a high pathway-specific PGS, deciles (2nd–10th) with patients with the lowest genetic scores (1st decile; n = 2367)] for the pathways that had a significant association after multiple testing. The dot points and error bars represent the odds ratios and 95% CIs for the respective polygenic deciles. The PGS deciles that crossed an odds ratio of 1 on the y-axis in the categorical outcome (A) and a beta coefficient of 0 in the continuous outcome (B) are not statistically significant. GABA, gamma-aminobutyric acid; GSK, glycogen synthase kinase; PGS, polygenic score.

Combined Modeling of Pathway-Specific Polygenic Scores

The multivariate modeling combining PSPGSs that showed a significant association in univariate analysis (combined PSPGS) explained 3.71% of the variance in the categorical and 3.18% in the continuous lithium responses. These results were higher than the variance explained by each of the PSPGSs. To compare the predictive performance of PGSs developed from the pathway-specific approach with the conventional genome-wide method, we developed a PGS for lithium responsiveness (Li+PGS) using a similar LOC procedure. The Li+PGS explained significant variance in the categorical (pseudo R2 = 2.87%) and continuous (pseudo R2 = 2.69%) outcomes (Table 2). When comparing these Li+PGS R2 values with the combined PSPGS, no statistically significant difference was found for either the categorical (R2 difference = 84%; p = .58, non-nested model) or the continuous (R2 difference = 0.49%; p = .88, non-nested model) lithium treatment outcomes. These findings indicate that the combined PSPGS resulted in model performance that was comparable to that of the conventional genome-wide PGS (see Supplement). Furthermore, the elastic-net regularization regression model resulted in 2.98% of the variability in the continuous outcome being explained, which is very similar to the results in the multivariate analysis using ordinary least squares regression (R2 = 3.18%) (Supplement).

Replication Analysis

Using the combined datasets from PsyCourse and BipoLife, we found a statistically significant association between the GABAPGS and lithium treatment response, both for the categorical (p < .01, pseudo R2 = 2.74%) and the continuous (p = .01, R2 = 3.30%) outcomes, replicating the findings from the above analysis. The results showed that patients with BD who had higher GABAPGSs were 1.53 times more likely (95% CI, 1.09 to 2.14) to have a good lithium treatment response than patients with lower GABAPGSs (for the categorical outcome). For the continuous outcome, each 1-unit increase in the GABAPGS was associated with a 0.57-point increase in the ALDA score (95% CI, 0.18 to 0.96). The CIRPGS was also significantly associated with the continuous lithium response (p = .01, R2 = 1.38). Each 1-unit increase in the CIRPGS was associated with a 0.40-point increase in the ALDA score (95% CI, 0.01 to 0.79). The association results of the PGSs for other potential biological pathways were not replicated. The full replication analysis results are available in Table 3.

Table 3.

The Association of Pathway-Specific Polygenic Scores With Clinical Lithium Treatment Response Among Patients With Bipolar Disorder in Replication Cohorts (Combined PsyCourse and BipoLife Cohorts) (N = 207)

Pathways No. of Genes No. of SNPs Categorical Outcome
 Continuous Outcome
aOR (95% CI) p Pseudo R2, % β (95% CI) p R2, %
Acetylcholine 164 6247 1.21 (0.88 to 1.67) .09 0.95% 0.24 (−0.15 to 0.64) .22 0.22%
GABA 76 2994 1.53 (1.09 to 2.14) .01a 2.74% 0.57 (0.18 to 0.96) .01a 3.31%
Calcium Channel 134 4236 0.96 (0.69 to 1.33) .79 0.05% 0.25 (−0.14 to 0.65) .21 0.30%
Mitochondria 163 7801 0.96 (0.70 to 1.33) .83 0.03% −0.13 (−0.52 to 0.27) .51 0.01%
Circadian Rhythm 92 3840 0.99 (0.72 to 1.37) .94 0.02% 0.40 (0.01 to 0.79) .04a 1.38%
Glutamate 129 6673 0.98 (0.71 to 1.35) .88 0.02% 0.29 (−0.10 to 0.68) .15 0.54%
Dopamine 155 5794 1.07 (0.78 to 1.47) .67 0.13% −0.20 (−0.60 to 0.19) .31 0.01%
GSK 18 707 1.13 (0.82 to 1.58) .44 0.33% 0.01 (−0.39 to −0.40) .96 0.01%
NMDA 11 641 1.13 (0.77 to 1.69) .81 0.03% 0.13 (−0.26 to 0.52) .52 0.01%
Open in a new tab

The cutoff for statistical significance was p < .05 after Benjamini-Hochberg procedure correction for multiple testing. The analysis models were adjusted for age, sex, chip type, and the first 4 principal components. The observed R2 in the categorical outcome is transformed to a liability scale. aOR is from the total sample (before decile stratification). R2 indicates variance explained by polygenic scores.

aOR, adjusted odds ratio; GABA, gamma-aminobutyric acid; GSK, glycogen synthase kinase; PsyCourse, Pathomechanisms and Signatures in the Longitudinal Course of Psychosis; SNP, single nucleotide polymorphism.

a

Denotes significant association after correction for multiple tests.

Discussion

For the first time, we developed biologically informative PSPGSs in well-characterized datasets and evaluated their association with lithium treatment response in patients with BD. Building on a previous study (29) that utilized the conventional genome-wide PGS approach, the current analysis targeted genetic variants mapped within acetylcholine, GABA, calcium channel, mitochondria, glutamate, circadian rhythm, dopamine, NMDA, and GSK pathways that are characterized as potential pharmacological targets in the treatment of BD (58,59). We found that BD patients with higher genetic loading of variants within the acetylcholine, GABA, calcium channel, GSK, and circadian rhythm pathways were more likely to respond to lithium treatment. In contrast, individuals with higher loading for genetic variants in the mitochondria pathway were less likely to respond to lithium. Our stratified analysis showed that patients with the highest genetic loading for acetylcholine, GABA, and calcium channel pathway variants (in the 10th decile) had a good lithium treatment response compared with patients with the lowest genetic loading (in the 1st decile), with an increasing trend of lithium treatment responsiveness across the 1st decile to the 10th decile. The trend was reversed in the mitochondria pathway.

Combined modeling of PSPGS explained 3.71% of the phenotypic variance in categorical and 3.18% in the continuous lithium response, comparable to the predictive power of the conventional polygenic model developed using genome-wide variants (29). While these approaches appear comparable in predictive performance, the PSPGS has advantages over the traditional whole genome polygenic approach in that it uses biological information to optimize the number of SNPs in each pathway or biological phenotype of interest (e.g., circadian rhythm, mitochondrial function) (60). The process of modeling individual genetic variation in lithium-related pathways attempts to enrich the selection of biologically significant variants at the pathway level, making PSPGS more biologically interpretable in comparison to conventional genome-wide PGS (60). Moreover, PSPGS could make the process of drug repurposing more efficient through its focus on specific genes in each pathway that are associated with pharmacogenomic outcomes of interest (61). The pathway-specific polygenic approach also prioritizes variants that may contribute to higher heritability estimates and the detection of enriched and functionally relevant GWAS signals by minimizing noise variants and leveraging the genetic variation across multiple potential biological pathways, thereby achieving better clinical utility (62, 63, 64, 65). Combining these scores with clinical data may further increase the variance explained, thereby improving their clinical utility (30).

In terms of the direction of associations with lithium treatment response, AChPGS, GABAPGS, Ca2+PGS, CIRPGS, and GSKPGS were positively associated. The positive associations between AChPGS, GABAPGS, and lithium response are consistent with evidence suggesting that lithium acts to correct deficits in acetylcholine and GABA neurotransmission (58,59,66,67). A similar association between lithium treatment response and Ca2+PGS is consistent with evidence of disruption of cellular calcium concentrations in patients with BD (68). Lithium is known to attenuate calcium release and regulate intracellular calcium levels in hippocampal neurons, thereby reducing excitotoxicity (69). Regarding the circadian rhythm pathway, lithium is widely recognized for its efficacy in improving sleep rhythm by increasing amplitude and slowing rhythm cycles (70,71). Good lithium responders show higher amplitude sleep cycle (72) and lithium has been found to correct rhythm abnormalities in patients with BD (73,74). The positive association between GSKPGS and lithium treatment response aligns with previously reported functional enrichment of the PI3K-Akt signaling pathway, which involves GSK-3β and is associated with response to lithium (75). From the above evidence, a higher genetic loading for lithium response appears to be related to better treatability.

On the other hand, the increased genetic variance within mitochondrial genes was associated with poorer lithium treatment response. Evidence suggests that patients with BD may experience reductions in mitochondrial enzyme levels and overall mitochondrial health, resulting in reduced bioenergetic capacity (76). Mitochondrial gene expression tends to be lower in the postmortem brains of patients with BD and is rescued specifically in lithium responders via a number of potential mechanisms including expression of electron transport chain proteins, second messenger systems such as protein kinase A, protein kinase C and in intracellular potassium and calcium regulation (76, 77, 78). Studies in induced pluripotent stem cell–derived neuron culture suggest that lithium may act to correct hyperexcitability via these mechanisms (76). The significant association between MITOPGS and lithium treatment response underscores the centrality of mitochondrial health in lithium’s mechanism of action. The relationship between our negative association of these genes with lithium response and protein expression networks needs further exploration.

Pathway-specific PGSs have been employed in several studies to enhance risk stratification in psychiatry (79, 80, 81, 82). For example, Grama et al. (79) investigated whether behavior- and neuronal-related gene sets, previously implicated in SCZ, were associated with subcortical volumes. They found that PGS derived from an abnormal behavior gene set was associated with right thalamic volume, and this association was robust across p-value thresholds, unlike the finding from a genome-wide approach (79). Warren et al. (80) also studied the relationship between the genome-wide PGSSCZ and the neurotransmitter PGSs (glutamate, GABA, dopamine, and serotonin) with psychotic disorder presentation. In this study, there was no significant association between individual symptom measures and the PGSSCZ, while glutamate and GABA pathway PGSs were associated with psychosis case status, and a dopamine pathway PGS was significantly associated with poorer global functioning in participants with psychosis (80). Other studies have shown that a PGS for oxidative stress pathway significantly differentiated individuals with early psychosis status from control individuals (82), and a dopamine pathway PGS has been implicated in the pathophysiology of SCZ (81).

Other examples of the development of PSPGSs in cardiovascular medicine have showcased their potential in personalizing treatment approaches. For example, a PGS for calcium signaling pathways together with phosphatidylinositol/inositol phosphate pathways was associated with hypertensive status, suggesting that their regulation could be a target for prevention and treatment of hypertension (83). PGSs tailored to pharmacodynamic pathways of angiotensin-converting enzyme (ACE) inhibitors (84) and β blockers, respectively, have enhanced patient selection for ACE inhibitors and predicted mortality in patients with heart failure (85). Similarly, a pharmacogenomic polygenic response score developed with 31 genes associated with adenosine diphosphate–based platelet reactivity during clopidogrel treatment has been applied to predict major adverse cardiovascular events and cardiovascular death in patients with coronary artery disease treated with clopidogrel (86).

Based on the evidence presented in our study and the literature summarized above, PSPGSs hold promise for the future of precision psychiatry by refining patient treatment stratification and improving treatment efficacy through leveraging genetic information across specific pharmacological pathways. PSPGS-based stratification of patients as likely good and poor responders could reduce delays in delivering effective treatment and associated burden. Serious side effects of long-term lithium therapy, including chronic renal failure, hypothyroidism, and mortality due to toxicity (87), could also be minimized.

Limitations

While our study provides novel and robust support for the use of PSPGS methods, the results should be interpreted in conjunction with some limitations. First, the ConLi+Gen cohort’s retrospective study design introduces challenges in determining associations between a PSPGS and lithium response without a placebo arm. However, the B scale in the ALDA score measures and weights the effect of confounding factors such as treatment duration, number and frequency of mood episodes off treatment, compliance, and use of concomitant psychotropics while on lithium. Second, only participants of European ancestry were included, and these results may not be generalized to other ancestrally diverse populations. Third, while our general strategy for pathway selection was based on the narrative review, we included cholinergic and glutamatergic pathways identified in our previous work with the same sample, highlighting the need for external replication of these pathways (29). The PsyCourse and BipoLife cohorts are genuine replications not used in the previous analysis and therefore mitigate this issue somewhat. Fourth, we used relatively small cohorts for the replication analysis, suggesting that the failure to replicate results for acetylcholine, calcium signaling, and mitochondrial pathways may be subject to type II error. Large and diverse cohorts are needed to further explore and replicate our findings. Fifth, our search was not exhaustive, and we may have excluded important biological pathways implicated in lithium pharmacology.

Conclusions

By focusing on biologically relevant genetic variants, PSPGS for lithium response has shown predictive capabilities comparable to those of the conventional genome-wide PGS, but with the advantage of using fewer SNPs and providing biological interpretability. While the variance in lithium treatment response explained by these models is still small, at best 3.71%, future models may include an expanded range of lithium-specific pathways to improve accuracy, reaching an effect size relevant to stratifying individuals by genetic risk for personalization of lithium treatment. Our study invites further investigation of how proteins including acetylcholine, GABA, calcium signaling, mitochondria, GSK, glutamate, and circadian rhythm pathways interact at the molecular level to define lithium treatment response. Replication in larger cohorts, including cohorts of participants of non-European ancestry, is required to establish the clinical utility of PSPGSs.

Acknowledgments and Disclosures

ATA is supported by the National Health and Medical Research Council (NHMRC) Emerging Leadership Investigator (Grant No. 2021–2008 000). NTS is a recipient of the University of Adelaide Research Scholarship. The primary sources of support were the Deutsche Forschungsgemeinschaft (DFG) (Grant Nos. RI 908/7-1, FOR2107, and RI 908/11-1 [to MR] and 246/10-1 [to MMN]) and Intramural Research Program of the National Institute of Mental Health (Grant No. ZIA-MH00284311). The genotyping was funded in part by the German Federal Ministry of Education and Research (BMBF) through the Integrated Network IntegraMent (Integrated Understanding of Causes and Mechanisms in Mental Disorders) under the auspices of the e:Med Programme (grants awarded to TGS, MR, and MMN). The BipoLife (Improving Recognition and Care in Critical Areas of Bipolar Disorders) study was funded by BMBF (to principal investigators [PIs]: FBer, PR, MBa, AR, SK-S, TGS, JW, GJ, AJF, and MLam). UH was supported by the European Union’s Horizon 2020 Research and Innovation Programme (PSY-PGx; Grant Agreement No. 945151). Some data and biomaterials were collected as part of 11 projects (Study 40) that participated in the National Institute of Mental Health (NIMH) Bipolar Disorder Genetics Initiative. From 2003 to 2007, the PIs and co-investigators were Indiana University, Indianapolis, Indiana, Grant No. R01 MH59545, John Nurnberger, M.D., Ph.D., Marvin J. Miller, M.D., Elizabeth S. Bowman, M.D., N. Leela Rau, M.D., P. Ryan Moe, M.D., Nalini Samavedy, M.D., Rif El-Mallakh, M.D. (at University of Louisville), Husseini Manji, M.D. (at Johnson & Johnson), Debra A. Glitz, M.D. (at Wayne State University), Eric T. Meyer, Ph.D., M.S. (at Oxford University, United Kingdom), Carrie Smiley, R.N., Tatiana Foroud, Ph.D., Leah Flury, M.S., Danielle M. Dick, Ph.D (at Virginia Commonwealth University), Howard Edenberg, Ph.D.; Washington University, St. Louis, Missouri, Grant No. R01 MH059534, John Rice, Ph.D, Theodore Reich, M.D., Allison Goate, Ph.D., Laura Bierut, M.D. Grant No. K02 DA21237; Johns Hopkins University, Baltimore, M.D., Grant No. R01 MH59533, Melvin McInnis, M.D., JRD, Jr., M.D., Dean F. MacKinnon, M.D., FMM, M.D., JBP, M.D., PPZ, Ph.D., Dimitrios Avramopoulos, and Jennifer Payne; University of Pennsylvania, Pennsylvania, Grant No. R01 MH59553, Wade Berrettini, M.D., Ph.D.; University of California at San Francisco, California, Grant No. R01 MH60068, William Byerley, M.D., and Sophia Vinogradov, M.D.; University of Iowa, Iowa, Grant No. R01 MH059548, William Coryell, M.D., and Raymond Crowe, M.D.; University of Chicago, Illinois, Grant No. R01 MH59535, Elliot Gershon, M.D., Judith Badner, Ph.D., FJM, M.D., Chunyu Liu, Ph.D., Alan Sanders, M.D., Maria Caserta, Steven Dinwiddie, M.D., Tu Nguyen, Donna Harakal; University of California at San Diego, California, Grant No. R01 MH59567, JK, M.D., Rebecca McKinney, B.A.; Rush University, Illinois, Grant No. R01 MH059556, William Scheftner, M.D., Howard M. Kravitz, D.O., M.P.H., Diana Marta, B.S., Annette Vaughn-Brown, M.S.N., R.N., and Laurie Bederow, M.A.; NIMH Intramural Research Program, Bethesda, Maryland, Grant No. 1Z01MH002810-01, FJM, M.D., LK, Psy.D., Sevilla Detera-Wadleigh, Ph.D., Lisa Austin, Ph.D., Dennis L. Murphy, M.D.; Howard University, William B. Lawson, M.D., Ph.D., Evarista Nwulia, M.D., and Maria Hipolito, M.D. This work was supported by the National Cancer Institute (Grant No. P50CA89392) and the National Institute on Drug Abuse (Grant No. 5K02DA021237). The Canadian part of the study was supported by the Canadian Institutes of Health Research (Grant No. 166098), as well as grants from Genome Canada and Research Nova Scotia (to MAl). Collection and phenotyping of the Australian UNSW sample by PBM, PRS, JMF, and AW was funded by Australian NHMRC Program (Grant No. 1037196). The collection of the Barcelona sample was supported by the Centro de Investigación en Red de Salud Mental (CIBERSAM), Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), and the Centres de Recerca de Catalunya Programme/Generalitat de Catalunya (Grant Nos. PI080247, PI1200906, PI12/00018, 2014SGR1636, and 2014SGR398). The Swedish Research Council, Stockholm County Council, Karolinska Institutet, and the Söderström-Königska Foundation supported this research through grants awarded to LB, LF, CL, and MScha. The collection of the Geneva sample was supported by the Swiss National Foundation (Grant Nos. Synapsy 51NF40-158776 and 32003B-125469). The collection of the Romanian sample was supported by an Executive Agency for Higher Education, Research, Development and Innovation Funding (U.E.F.I.S.C.D.I.), Romania grant (to MG-S).

We thank all patients who participated in the study, and we appreciate the contributions of the clinicians, scientists, research assistants, and study staff who helped in the patient recruitment, data collection, and biological sample preparation of the studies. We are also indebted to the members of the ConLi+Gen Scientific Advisory Board (http://www.ConLi+Gen.org/) for critical input over the course of the project. Angel’s pilot work, as part of her honors thesis, provided the foundation for designing this project, and we gratefully acknowledge her contribution. The analysis of this study was carried out using the high-performance computational capabilities of the University of Adelaide’s Phoenix Supercomputer https://www.adelaide.edu.au/phoenix/.

ATA conceived and designed the project and secured a fellowship to lead the study. NTS developed the research proposal, conducted the statistical analysis, interpreted the findings, and drafted the manuscript. ATA, SRC, and KOS provided supervision and critically reviewed the data analysis steps and article draft. All authors contributed genetic and clinical data, provided feedback, and made significant intellectual contributions to the article.

All data used in this analysis are available to ConLi+Gen members. See http://www.conligen.org/ for more information. The following software were used: PRS-CS (see https://github.com/getian107/PRScs), MAGMA (https://ctg.cncr.nl/software/magma), PLINK 2 (https://zzz.bwh.harvard.edu/plink/tutorial.shtml). The custom codes used in this study will be made available upon reasonable request to the corresponding author.

EV has received grants and served as consultant, adviser, or continuing medical education (CME) speaker for the following entities: AB-Biotics, Allergan, Angelini, Astra Zeneca, Bristol-Myers Squibb, Dainippon Sumitomo Pharma, Farmindustria, Ferrer, Forest Research Institute, Gedeon Richter, GlaxoSmith-Kline, Janssen, Lundbeck, Otsuka, Pfizer, Roche, Sanofi-Aventis, Servier, Shire, Sunovion, Takeda, the Brain and Behaviour Foundation, the Spanish Ministry of Science and Innovation (CIBERSAM), and the Stanley Medical Research Institute. MBa has received grants from the DFG and BMBF and served as consultant, adviser, or CME speaker for the following entities: Allergan, Aristo, Janssen, Lilly, Lundbeck, neuraxpharm, Otsuka, Sandoz, Servier, and Sunovion outside the submitted work. SK-S has received grants and served as consultant, adviser, or speaker for the following entities: Medice Arzneimittel Pütter GmbH and Shire/Takeda. BTB has received grants and served as consultant, adviser, or CME speaker for the following entities: Astra Zeneca, Bristol-Myers Squibb, Janssen, Lundbeck, Otsuka, Servier, the National Health and Medical Research Council, the Fay Fuller Foundation, and the James and Diana Ramsay Foundation. SRC has received grants and served as consultant, adviser, or CME speaker for the following entities: Otsuka Australia, Lundbeck Australia, Janssen-Cilag Australia, and Servier Australia. TKa received honoraria for lectures, manuscripts, and/or consultancy from Kyowa Hakko Kirin Co., Ltd.; Eli Lilly Japan K.K.; Otsuka Pharmaceutical Co., Ltd.; GlaxoSmithKline K.K.; Taisho Toyama Pharmaceutical Co., Ltd.; Dainippon Sumitomo Pharma Co., Ltd.; Meiji Seika Pharma Co., Ltd.; Pfizer Japan Inc.; Mochida Pharmaceutical Co., Ltd.; Shionogi & Co., Ltd.; Janssen Pharmaceutical K.K.; Janssen Asia Pacific; Yoshitomiyakuhin; Astellas Pharma Inc.; Wako Pure Chemical Industries, Ltd.; Wiley Publishing Japan; Nippon Boehringer Ingelheim Co., Ltd.; Kanae Foundation for the Promotion of Medical Science; MSD K.K.; Kyowa Pharmaceutical Industry Co., Ltd.; and Takeda Pharmaceutical Co., Ltd. TKa also received a research grant from Takeda Pharmaceutical Co., Ltd. PF has received grants and served as consultant, adviser, or CME speaker for the following entities Abbott, GlaxoSmithKline, Janssen, Essex, Lundbeck, Otsuka, Gedeon Richter, Servier, and Takeda as well as the German Ministry of Science and the German Ministry of Health. ER has received grants and served as consultant, adviser, or CME speaker for the following entities: Janssen and Institut Allergosan. MLan declares that, over the past 36 months, he has received lecture honoraria from Lundbeck and served as a scientific consultant for EPID Research Oy; he declares no other equity ownership, profit-sharing agreements, royalties, or patents. KAk has received consulting honoraria from Taisho Toyama Pharmaceutical Co., Ltd. In 2021, JZ served as an adviser for Biogen concerning Aducanumab (Alzheimer’s disease). All other authors report no biomedical financial interests or potential conflicts of interest.

Footnotes

Supplementary material cited in this article is available online at https://doi.org/10.1016/j.bpsgos.2025.100558.

Supplementary Material

Supplemental Methods, Results, Figure S1, and Tables S1–S7
mmc1.pdf (523.9KB, pdf)
Supplemental Data_GeneSNPLists
mmc2.xlsx (1.9MB, xlsx)
Key Resources Table
mmc3.xlsx (21.2KB, xlsx)

References

  • 1.International Schizophrenia Consortium. Purcell S.M., Wray N.R., Stone J.L., Visscher P.M., O’Donovan M.C., et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature. 2009;460:748–752. doi: 10.1038/nature08185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wray N.R., Kemper K.E., Hayes B.J., Goddard M.E., Visscher P.M. Complex trait prediction from genome data: Contrasting EBV in livestock to PRS in humans: Genomic prediction. Genetics. 2019;211:1131–1141. doi: 10.1534/genetics.119.301859. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Evans D.M., Visscher P.M., Wray N.R. Harnessing the information contained within genome-wide association studies to improve individual prediction of complex disease risk. Hum Mol Genet. 2009;18:3525–3531. doi: 10.1093/hmg/ddp295. [DOI] [PubMed] [Google Scholar]
  • 4.Wray N.R., Lin T., Austin J., McGrath J.J., Hickie I.B., Murray G.K., Visscher P.M. From basic science to clinical application of polygenic risk scores: A primer. JAMA Psychiatry. 2021;78:101–109. doi: 10.1001/jamapsychiatry.2020.3049. [DOI] [PubMed] [Google Scholar]
  • 5.Sharew N.T., Clark S.R., Schubert K.O., Amare A.T. Pharmacogenomic scores in psychiatry: Systematic review of current evidence. Transl Psychiatry. 2024;14:322. doi: 10.1038/s41398-024-02998-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Schizophrenia Working Group of the Psychiatric Genomics Consortium Biological insights from 108 schizophrenia-associated genetic loci. Nature. 2014;511:421–427. doi: 10.1038/nature13595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Howard D.M., Adams M.J., Clarke T.K., Hafferty J.D., Gibson J., Shirali M., et al. Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions. Nat Neurosci. 2019;22:343–352. doi: 10.1038/s41593-018-0326-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Mullins N., Forstner A.J., O’Connell K.S., Coombes B., Coleman J.R.I., Qiao Z., et al. Genome-wide association study of more than 40,000 bipolar disorder cases provides new insights into the underlying biology. Nat Genet. 2021;53:817–829. doi: 10.1038/s41588-021-00857-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Siemens A., Anderson S.J., Rassekh S.R., Ross C.J.D., Carleton B.C. A systematic review of polygenic models for predicting drug outcomes. J Pers Med. 2022;12:1394. doi: 10.3390/jpm12091394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Johnson D., Wilke M.A.P., Lyle S.M., Kowalec K., Jorgensen A., Wright G.E.B., Drögemöller B.I. A systematic review and analysis of the use of polygenic scores in pharmacogenomics. Clin Pharmacol Ther. 2022;111:919–930. doi: 10.1002/cpt.2520. [DOI] [PubMed] [Google Scholar]
  • 11.Zhang J.P., Robinson D., Yu J., Gallego J., Fleischhacker W.W., Kahn R.S., et al. Schizophrenia polygenic risk score as a predictor of antipsychotic efficacy in first-episode psychosis. Am J Psychiatry. 2019;176:21–28. doi: 10.1176/appi.ajp.2018.17121363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Pardiñas A.F., Smart S.E., Willcocks I.R., Holmans P.A., Dennison C.A., Lynham A.J., et al. Interaction testing and polygenic risk scoring to estimate the association of common genetic variants with treatment resistance in schizophrenia. JAMA Psychiatry. 2022;79:260–269. doi: 10.1001/jamapsychiatry.2021.3799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Talarico F., Costa G.O., Ota V.K., Santoro M.L., Noto C., Gadelha A., et al. Systems-level analysis of genetic variants reveals functional and spatiotemporal context in treatment-resistant schizophrenia. Mol Neurobiol. 2022;59:3170–3182. doi: 10.1007/s12035-022-02794-7. [DOI] [PubMed] [Google Scholar]
  • 14.Werner M.C.F., Wirgenes K.V., Haram M., Bettella F., Lunding S.H., Rødevand L., et al. Indicated association between polygenic risk score and treatment-resistance in a naturalistic sample of patients with schizophrenia spectrum disorders. Schizophr Res. 2020;218:55–62. doi: 10.1016/j.schres.2020.03.006. [DOI] [PubMed] [Google Scholar]
  • 15.Gasse C., Wimberley T., Wang Y., Mors O., Børglum A., Als T.D., et al. Schizophrenia polygenic risk scores, urbanicity and treatment-resistant schizophrenia. Schizophr Res. 2019;212:79–85. doi: 10.1016/j.schres.2019.08.008. [DOI] [PubMed] [Google Scholar]
  • 16.Yoshida K., Marshe V.S., Elsheikh S.S.M., Maciukiewicz M., Tiwari A.K., Brandl E.J., et al. Polygenic risk scores analyses of psychiatric and metabolic traits with antipsychotic-induced weight gain in schizophrenia: An exploratory study. Pharmacogenomics J. 2023;23:119–126. doi: 10.1038/s41397-023-00305-y. [DOI] [PubMed] [Google Scholar]
  • 17.Muntané G., Vázquez-Bourgon J., Sada E., Martorell L., Papiol S., Bosch E., et al. Polygenic risk scores enhance prediction of body mass index increase in individuals with a first episode of psychosis. Eur Psychiatry. 2023;66 doi: 10.1192/j.eurpsy.2023.9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lacaze P., Ronaldson K.J., Zhang E.J., Alfirevic A., Shah H., Newman L., et al. Genetic associations with clozapine-induced myocarditis in patients with schizophrenia. Transl Psychiatry. 2020;10:37. doi: 10.1038/s41398-020-0722-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Facal F., Arrojo M., Paz E., Páramo M., Costas J. Association between psychiatric hospitalizations of patients with schizophrenia and polygenic risk scores based on genes with altered expression by antipsychotics. Acta Psychiatr Scand. 2022;146:139–150. doi: 10.1111/acps.13444. [DOI] [PubMed] [Google Scholar]
  • 20.Pain O., Hodgson K., Trubetskoy V., Ripke S., Marshe V.S., Adams M.J., et al. Identifying the common genetic basis of antidepressant response. Biol Psychiatry Glob Open Sci. 2022;2:115–126. doi: 10.1016/j.bpsgos.2021.07.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Fanelli G., Domschke K., Minelli A., Gennarelli M., Martini P., Bortolomasi M., et al. A meta-analysis of polygenic risk scores for mood disorders, neuroticism, and schizophrenia in antidepressant response. Eur Neuropsychopharmacol. 2022;55:86–95. doi: 10.1016/j.euroneuro.2021.11.005. [DOI] [PubMed] [Google Scholar]
  • 22.Fanelli G., Benedetti F., Kasper S., Zohar J., Souery D., Montgomery 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 doi: 10.1016/j.pnpbp.2020.110170. [DOI] [PubMed] [Google Scholar]
  • 23.Amare A.T., Schubert K.O., Tekola-Ayele F., Hsu Y.H., Sangkuhl K., Jenkins G., et al. Association of the polygenic scores for personality traits and response to selective serotonin reuptake inhibitors in patients with major depressive disorder. Front Psychiatry. 2018;9:65. doi: 10.3389/fpsyt.2018.00065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Taylor R.W., Coleman J.R.I., Lawrence A.J., Strawbridge R., Zahn R., Cleare A.J. Predicting clinical outcome to specialist multimodal inpatient treatment in patients with treatment resistant depression. J Affect Disord. 2021;291:188–197. doi: 10.1016/j.jad.2021.04.074. [DOI] [PubMed] [Google Scholar]
  • 25.Wigmore E.M., Hafferty J.D., Hall L.S., Howard D.M., Clarke T.K., Fabbri C., 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;20:329–341. doi: 10.1038/s41397-019-0067-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.International Consortium on Lithium Genetics (ConLi+Gen) Amare A.T., Schubert K.O., Hou L., Clark S.R., Papiol S., et al. Association of polygenic score for schizophrenia and hla antigen and inflammation genes with response to lithium in bipolar affective disorder: A genome-wide association study. JAMA Psychiatry. 2018;75:65–74. doi: 10.1001/jamapsychiatry.2017.3433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Amare A.T., Schubert K.O., Hou L., Clark S.R., Papiol S., Cearns M., et al. Association of polygenic score for major depression with response to lithium in patients with bipolar disorder. Mol Psychiatry. 2021;26:2457–2470. doi: 10.1038/s41380-020-0689-5. [DOI] [PubMed] [Google Scholar]
  • 28.Coombes B.J., Millischer V., Batzler A., Larrabee B., Hou L., Papiol S., et al. Association of attention-deficit/hyperactivity disorder and depression polygenic scores with lithium response: A consortium for lithium genetics study. Complex Psychiatry. 2021;7:80–89. doi: 10.1159/000519707. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Amare A.T., Thalamuthu A., Schubert K.O., Fullerton J.M., Ahmed M., Hartmann S., et al. Association of polygenic score and the involvement of cholinergic and glutamatergic pathways with lithium treatment response in patients with bipolar disorder. Mol Psychiatry. 2023;28:5251–5261. doi: 10.1038/s41380-023-02149-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Cearns M., Amare A.T., Schubert K.O., Thalamuthu A., Frank J., Streit F., et al. Using polygenic scores and clinical data for bipolar disorder patient stratification and lithium response prediction: Machine learning approach—Corrigendum. Br J Psychiatry. 2022;221:494. doi: 10.1192/bjp.2022.55. [DOI] [PubMed] [Google Scholar]
  • 31.O’Connell K.S., Koch E., Lenk H.Ç., Akkouh I.A., Hindley G., Jaholkowski P., et al. Polygenic overlap with body-mass index improves prediction of treatment-resistant schizophrenia. Psychiatry Res. 2023;325 doi: 10.1016/j.psychres.2023.115217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Hari Dass S.A., McCracken K., Pokhvisneva I., Chen L.M., Garg E., Nguyen T.T.T., et al. A biologically-informed polygenic score identifies endophenotypes and clinical conditions associated with the insulin receptor function on specific brain regions. EBiomedicine. 2019;42:188–202. doi: 10.1016/j.ebiom.2019.03.051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Bodea C.A., Macoritto M., Liu Y., Zhang W., Karman J., King E.A., et al. Pathway specific polygenic risk scores identify pathways and patient clusters associated with inflammatory bowel disease risk, severity and treatment response. medRxiv. 2021 doi: 10.1101/2021.11.19.21266549. [DOI] [Google Scholar]
  • 34.Tubbs J.D., Leung P.B.M., Zhong Y., Zhan N., Hui T.C.K., Ho K.K.Y., et al. Pathway-specific polygenic scores improve cross-ancestry prediction of psychosis and clinical outcomes. medRxiv. 2023 doi: 10.1101/2021.11.19.21266549. [DOI] [Google Scholar]
  • 35.Schärfe C.P.I., Tremmel R., Schwab M., Kohlbacher O., Marks D.S. Genetic variation in human drug-related genes. Genome Med. 2017;9:117. doi: 10.1186/s13073-017-0502-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Sinnott-Armstrong N., Naqvi S., Rivas M., Pritchard J.K. GWAS of three molecular traits highlights core genes and pathways alongside a highly polygenic background. eLife. 2021;10 doi: 10.7554/eLife.58615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Bennett D., O’Shea D., Ferguson J., Morris D., Seoighe C. Controlling for background genetic effects using polygenic scores improves the power of genome-wide association studies. Sci Rep. 2021;11 doi: 10.1038/s41598-021-99031-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Schulze T.G., Alda M., Adli M., Akula N., Ardau R., Bui E.T., et al. The International Consortium on Lithium Genetics (ConLiGen): An initiative by the NIMH and IGSLI to study the genetic basis of response to lithium treatment. Neuropsychobiology. 2010;62:72–78. doi: 10.1159/000314708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Dwyer D.B., Kalman J.L., Budde M., Kambeitz J., Ruef A., Antonucci L.A., et al. An investigation of psychosis subgroups with prognostic validation and exploration of genetic underpinnings: The PsyCourse study. JAMA Psychiatry. 2020;77:523–533. doi: 10.1001/jamapsychiatry.2019.4910. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Budde M., Anderson-Schmidt H., Gade K., Reich-Erkelenz D., Adorjan K., Kalman J.L., et al. A longitudinal approach to biological psychiatric research: The PsyCourse study. Am J Med Genet B Neuropsychiatr Genet. 2019;180:89–102. doi: 10.1002/ajmg.b.32639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ritter P.S., Bermpohl F., Gruber O., Hautzinger M., Jansen A., Juckel G., et al. Aims and structure of the German Research Consortium BipoLife for the study of bipolar disorder. Int J Bipolar Disord. 2016;4:26. doi: 10.1186/s40345-016-0066-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Duffy A., Alda M., Milin R., Grof P. A consecutive series of treated affected offspring of parents with bipolar disorder: Is response associated with the clinical profile? Can J Psychiatry. 2007;52:369–376. doi: 10.1177/070674370705200606. [DOI] [PubMed] [Google Scholar]
  • 43.Manchia M., Adli M., Akula N., Ardau R., Aubry J.M., Backlund L., et al. Assessment of response to lithium maintenance treatment in bipolar disorder: A consortium on lithium genetics (ConLiGen) report. PLoS One. 2013;8 doi: 10.1371/journal.pone.0065636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Hou L., Heilbronner U., Degenhardt F., Adli M., Akiyama K., Akula N., et al. Genetic variants associated with response to lithium treatment in bipolar disorder: A genome-wide association study. Lancet. 2016;387:1085–1093. doi: 10.1016/S0140-6736(16)00143-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Schubert K.O., Thalamuthu A., Amare A.T., Frank J., Streit F., Adl M., et al. Combining schizophrenia and depression polygenic risk scores improves the genetic prediction of lithium response in bipolar disorder patients. Transl Psychiatry. 2021;11:606. doi: 10.1038/s41398-021-01702-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Nunes A., Stone W., Ardau R., Berghöfer A., Bocchetta A., Chillotti C., et al. Exemplar scoring identifies genetically separable phenotypes of lithium responsive bipolar disorder. Transl Psychiatry. 2021;11:36. doi: 10.1038/s41398-020-01148-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Purcell S., Neale B., Todd-Brown K., Thomas L., Ferreira M.A.R., Bender D., et al. PLINK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–575. doi: 10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.McCarthy S., Das S., Kretzschmar W., Delaneau O., Wood A.R., Teumer A., et al. A reference panel of 64,976 haplotypes for genotype imputation. Nat Genet. 2016;48:1279–1283. doi: 10.1038/ng.3643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.de Leeuw C.A., Mooij J.M., Heskes T., Posthuma D. MAGMA: Generalized gene-set analysis of GWAS data. PLoS Comput Biol. 2015;11 doi: 10.1371/journal.pcbi.1004219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Lever J., Krzywinski M., Altman N. Model selection and overfitting. Nat Methods. 2016;13:703–704. [Google Scholar]
  • 51.Gareth J., Daniela W., Trevor H., Robert T. Spinger; New York, NY: 2013. An Introduction to Statistical Learning: With applications in R. [Google Scholar]
  • 52.Ge T., Chen C.Y., Ni Y., Feng Y.A., Smoller J.W. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nat Commun. 2019;10:1776. doi: 10.1038/s41467-019-09718-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Benjamini Y., Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc B (Methodol) 1995;57:289–300. [Google Scholar]
  • 54.Momin M.M., Lee S., Wray N.R., Lee S.H. Significance tests for R2 of out-of-sample prediction using polygenic scores. Am J Hum Genet. 2023;110:349–358. doi: 10.1016/j.ajhg.2023.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.McFadden D. University of California, Berkeley; Berkeley, CA: 1972. Conditional Logit Analysis of Qualitative Choice Behavior. [Google Scholar]
  • 56.Lee S.H., Wray N.R., Goddard M.E., Visscher P.M. Estimating missing heritability for disease from genome-wide association studies. Am J Hum Genet. 2011;88:294–305. doi: 10.1016/j.ajhg.2011.02.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Garnham J., Munro A., Slaney C., MacDougall M., Passmore M., Duffy A., et al. Prophylactic treatment response in bipolar disorder: Results of a naturalistic observation study. J Affect Disord. 2007;104:185–190. doi: 10.1016/j.jad.2007.03.003. [DOI] [PubMed] [Google Scholar]
  • 58.Malhi G.S., Outhred T. Therapeutic mechanisms of lithium in bipolar disorder: Recent advances and current understanding. CNS Drugs. 2016;30:931–949. doi: 10.1007/s40263-016-0380-1. [DOI] [PubMed] [Google Scholar]
  • 59.Malhi G.S., Tanious M., Das P., Coulston C.M., Berk M. Potential mechanisms of action of lithium in bipolar disorder. Current understanding. CNS Drugs. 2013;27:135–153. doi: 10.1007/s40263-013-0039-0. [DOI] [PubMed] [Google Scholar]
  • 60.Choi S.W., García-González J., Ruan Y., Wu H.M., Porras C., Johnson J., et al. PRSet: Pathway-based polygenic risk score analyses and software. PLoS Genet. 2023;19 doi: 10.1371/journal.pgen.1010624. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Kraft J., Braun A., Awasthi S., Panagiotaropoulou G., Schipper M., Bell N.Y., et al. Identifying drug targets for schizophrenia through gene prioritization. medRxiv. 2024 doi: 10.1101/2024.05.15.24307423. [DOI] [Google Scholar]
  • 62.Schrodi S.J., Mukherjee S., Shan Y., Tromp G., Sninsky J.J., Callear A.P., et al. Genetic-based prediction of disease traits: Prediction is very difficult, especially about the future. Front Genet. 2014;5:162. doi: 10.3389/fgene.2014.00162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Hu Y., Lu Q., Powles R., Yao X., Yang C., Fang F., et al. Leveraging functional annotations in genetic risk prediction for human complex diseases. PLOS Comput Biol. 2017;13 doi: 10.1371/journal.pcbi.1005589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Hu Y., Lu Q., Liu W., Zhang Y., Li M., Zhao H. Joint modeling of genetically correlated diseases and functional annotations increases accuracy of polygenic risk prediction. PLoS Genet. 2017;13 doi: 10.1371/journal.pgen.1006836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Shi J., Park J.-H., Duan J., Berndt S.T., Moy W., Yu K., et al. Winner’s curse correction and variable thresholding improve performance of polygenic risk modeling based on genome-wide association study summary-level data. PLoS Genet. 2016;12 doi: 10.1371/journal.pgen.1006493. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Rijal S., Jang S.H., Park S.J., Han S.K. Lithium enhances the GABAergic synaptic activities on the hypothalamic preoptic area (hPOA) neurons. Int J Mol Sci. 2021;22:3908. doi: 10.3390/ijms22083908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Wood A.J., Goodwin G.M. A review of the biochemical and neuropharmacological actions of lithium. Psychol Med. 1987;17:579–600. doi: 10.1017/s0033291700025836. [DOI] [PubMed] [Google Scholar]
  • 68.D’Onofrio S., Mahaffey S., Garcia-Rill E. Role of calcium channels in bipolar disorder. Curr Psychopharmacol. 2017;6:122–135. doi: 10.2174/2211556006666171024141949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Sourial-Bassillious N., Rydelius P.A., Aperia A., Aizman O. Glutamate-mediated calcium signaling: A potential target for lithium action. Neuroscience. 2009;161:1126–1134. doi: 10.1016/j.neuroscience.2009.04.013. [DOI] [PubMed] [Google Scholar]
  • 70.Li J., Lu W.-Q., Beesley S., Loudon A.S.I., Meng Q.-J. Lithium impacts on the amplitude and period of the molecular circadian clockwork. PLoS One. 2012;7 doi: 10.1371/journal.pone.0033292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.McCarthy M.J., Wei H., Marnoy Z., Darvish R.M., McPhie D.L., Cohen B.M., Welsh D.K. Genetic and clinical factors predict lithium’s effects on PER2 gene expression rhythms in cells from bipolar disorder patients. Transl Psychiatry. 2013;3 doi: 10.1038/tp.2013.90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.McCarthy M.J., Wei H., Nievergelt C.M., Stautland A., Maihofer A.X., Welsh D.K., et al. Chronotype and cellular circadian rhythms predict the clinical response to lithium maintenance treatment in patients with bipolar disorder. Neuropsychopharmacology. 2019;44:620–628. doi: 10.1038/s41386-018-0273-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Scott J., Hennion V., Meyrel M., Bellivier F., Etain B. An ecological study of objective rest–activity markers of lithium response in bipolar-I-disorder. Psychol Med. 2022;52:2281–2289. doi: 10.1017/S0033291720004171. [DOI] [PubMed] [Google Scholar]
  • 74.Federoff M., McCarthy M.J., Anand A., Berrettini W.H., Bertram H., Bhattacharjee A., et al. Correction of depression-associated circadian rhythm abnormalities is associated with lithium response in bipolar disorder. Bipolar Disord. 2022;24:521–529. doi: 10.1111/bdi.13162. [DOI] [PubMed] [Google Scholar]
  • 75.Ou A.H., Rosenthal S.B., Adli M., Akiyama K., Akula N., Alda M., et al. Lithium response in bipolar disorder is associated with focal adhesion and PI3K-Akt networks: A multi-omics replication study. Transl Psychiatry. 2024;14:109. doi: 10.1038/s41398-024-02811-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Cordeiro R.C., Lima C.N.C., Fries G.R., Zunta-Soares G., Soares J.C., Quevedo J., Scaini G. Mitochondrial health index correlates with plasma circulating cell-free mitochondrial DNA in bipolar disorder. Mol Psychiatry. 2023;11:4622–4631. doi: 10.21203/rs.3.rs-2821492/v1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Stacey D., Schubert K.O., Clark S.R., Amare A.T., Milanesi E., Maj C., et al. A gene co-expression module implicating the mitochondrial electron transport chain is associated with long-term response to lithium treatment in bipolar affective disorder. Transl Psychiatry. 2018;8:183. doi: 10.1038/s41398-018-0237-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78.Sun X., Wang J.-F., Tseng M., Young L.T. Downregulation in components of the mitochondrial electron transport chain in the postmortem frontal cortex of subjects with bipolar disorder. J Psychiatry Neurosci. 2006;31:189–196. [PMC free article] [PubMed] [Google Scholar]
  • 79.Grama S., Willcocks I., Hubert J.J., Pardiñas A.F., Legge S.E., Bracher-Smith M., et al. Polygenic risk for schizophrenia and subcortical brain anatomy in the UK Biobank cohort. Transl Psychiatry. 2020;10:309. doi: 10.1038/s41398-020-00940-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 80.Warren T.L., Tubbs J.D., Lesh T.A., Corona M.B., Pakzad S.S., Albuquerque M.D., et al. Association of neurotransmitter pathway polygenic risk with specific symptom profiles in psychosis. Mol Psychiatry. 2024;29:2389–2398. doi: 10.1038/s41380-024-02457-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 81.Sportelli L., Eisenberg D.P., Passiatore R., D’Ambrosio E., Antonucci L.A., Bettina J.S., et al. Dopamine signaling enriched striatal gene set predicts striatal dopamine synthesis and physiological activity in vivo. Nat Commun. 2024;15:3342. doi: 10.1038/s41467-024-47456-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 82.Pistis G., Vázquez-Bourgon J., Fournier M., Jenni R., Cleusix M., Papiol S., et al. Gene set enrichment analysis of pathophysiological pathways highlights oxidative stress in psychosis. Mol Psychiatry. 2022;27:5135–5143. doi: 10.1038/s41380-022-01779-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83.Maj C., Salvi E., Citterio L., Borisov O., Simonini M., Glorioso V., et al. Dissecting the polygenic basis of primary hypertension: Identification of key pathway-specific components. Front Cardiovasc Med. 2022;9 doi: 10.3389/fcvm.2022.814502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 84.Brugts J.J., Isaacs A., Boersma E., Van Duijn C.M., Uitterlinden A.G., Remme W., et al. Genetic determinants of treatment benefit of the angiotensin-converting enzyme-inhibitor perindopril in patients with stable coronary artery disease. Eur Heart J. 2010;31:1854–1864. doi: 10.1093/eurheartj/ehq169. [DOI] [PubMed] [Google Scholar]
  • 85.Lanfear D.E., Luzum J.A., She R., Gui H., Donahue M.P., O’Connor C.M., et al. Polygenic score for β-blocker survival benefit in European ancestry patients with reduced ejection fraction heart failure. Circ Heart Fail. 2020;13 doi: 10.1161/CIRCHEARTFAILURE.119.007012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 86.Lewis J.P., Backman J.D., Reny J.L., Bergmeijer T.O., Mitchell B.D., Ritchie M.D., et al. Pharmacogenomic polygenic response score predicts ischaemic events and cardiovascular mortality in clopidogrel-treated patients. Eur Heart J Cardiovasc Pharmacother. 2020;6:203–210. doi: 10.1093/ehjcvp/pvz045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 87.Malhi G.S., Tanious M., Das P., Berk M. The science and practice of lithium therapy. Aust N Z J Psychiatry. 2012;46:192–211. doi: 10.1177/0004867412437346. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplemental Methods, Results, Figure S1, and Tables S1–S7
mmc1.pdf (523.9KB, pdf)
Supplemental Data_GeneSNPLists
mmc2.xlsx (1.9MB, xlsx)
Key Resources Table
mmc3.xlsx (21.2KB, xlsx)

Articles from Biological Psychiatry Global Open Science are provided here courtesy of Elsevier

RESOURCES