Comorbidity alters the genetic relationship between anxiety disorders and major depression

Markos Tesfaye, Alexey Shadrin, Nadine Parker, Piotr Jaholkowski, Pravesh Parekh, Gleda Kutrolli, Viktoria Birkenæs, Nora R. Bakken, Helga Ask, Oleksandr Frei, Srdjan Djurovic, Anders M. Dale, Olav B. Smeland, Kevin S. O’Connell, Ole A. Andreassen
doi: https://doi.org/10.1101/2024.11.19.24317523

Abstract

Importance There is extensive comorbidity between anxiety disorders (ANX) and major depression (MD). Most studies on the genetics of ANX do not exclude comorbid cases of MD, and vice versa, therefore confounding genetic association analyses. Disorder-specific analysis of genomic data may reveal more precise biological pathways and causal relationships.

Objective To investigate the genetic relationship between disorder-specific ANX and MD compared to samples with comorbidity, including their causal relationship.

Design, Setting, and Participants Data from UK Biobank was used to perform genome-wide association studies (GWAS) of ANX-only and MD-only, and generate disorder-specific polygenic risk scores (PRS). The Norwegian Mother, Father, and Child Cohort (MoBa) was used to test the associations of PRS with diagnosis and symptoms. MD and ANX GWAS data including comorbidities (MD-co and ANX-co) were used as comparators. Genetic correlation was assessed using LDSC, and Mendelian randomization was employed to infer causal relationships.

Main Outcomes and Measures GWAS of ICD-10 diagnoses of ANX, MD, or both. Genetic correlations between pairs of ANX and MD phenotypes. PRS associations with diagnoses of ANX, MD, and their comorbid states, and anxiety or depressive symptoms.

Results The GWAS of ANX-only (9,980 cases and 179,442 controls) and MD-only (15,301 cases and 179,038 controls) showed a lower genetic correlation (0.53) than the one between ANX-co and MD-co (0.90). ANX-only showed a causal relationship with MD-only (PFDR=1.5e-02), but not vice versa, while comorbid cases showed a significant bidirectional causal relationship (PFDR=2.9e-12, PFDR =9.3e-06). The PRS-MD-only were differentially associated with MD-only compared to ANX-only cases (β= -0.08; 95%CI: -0.11, -0.03); however, this differential association was not observed for the PRS-MD-co. A similar pattern of differential association with anxiety and depressive symptoms was observed for PRS-ANX-only, but not for PRS-MD-co.

Conclusions and Relevance The genetics and underlying biology of ANX and MD are more distinct when comorbid cases are excluded from analyses and reveals that ANX may be causal for MD. This confounding of genetic relationships as a result of comorbidity is likely to apply to other psychiatric disorders. Disorder-specific genetic studies may help uncover more relevant biological mechanisms and guide more targeted clinical interventions.

Question Is the high genetic correlation between anxiety disorders (ANX) and major depression (MD) due to phenotypic overlap (comorbidity) and will disorder-specific (non-comorbid) analysis provide insight into their underlying causal factors?

Findings Applying statistical genetics tools in two large population samples (UK Biobank, n=500,000; and MoBa, n= 130,000), we found a smaller genetic correlation between MD and ANX when comorbid cases were excluded. Additionally, after excluding comorbid cases, Mendelian randomization revealed that genetic liability to ANX is causally linked to MD but not vice versa.

Meaning The genetic overlap between ANX and MD may be driven by phenotypic comorbidity, which inflates their genetic correlation. The results support that ANX has a causal link to MD, suggesting that early interventions for ANX may prevent subsequent development of MD. Further, investigating the genetic underpinnings of ANX and MD, while excluding their comorbid states, may help advance precision medicine and uncover novel biological mechanisms.

Introduction

Anxiety disorders (ANX) and major depression (MD) are the most prevalent psychiatric conditions, making a substantial contribution to poor health and disability globally.1 Both ANX and MD are more prevalent among women.1–3 Comorbidity between the two disorders occurs in over half of all cases,2,3 and is associated with greater morbidity, worse treatment response, and higher suicide risk.4–6 Overlapping symptomatology, causal relationships, and shared environmental risk factors7–10 as well as shared genetic risk and neurobiology may explain the high comorbidity.9,11–13

Both ANX and MD exhibit moderate heritability,14,15 and consistently show high genetic correlations.9,13,16 However, the high comorbidity makes these genetic associations difficult to interpret. The largest genome-wide association study (GWAS) for ANX did not exclude cases of MD, and vice versa.17,18 Polygenic risk scores (PRS) derived from these GWASs exhibit strong associations with comorbid ANX and MD cases,19 and the PRS of MD was a stronger predictor of ANX than the PRS of ANX.9,13 While including individuals with comorbidity may improve the power of GWAS to identify genetic loci,20 such approaches fail to distinguish genes specific to ANX from those specific to MD. Consequently, identifying distinct, disorder-specific, biological mechanisms may be hampered.21 Therefore, GWAS of anxiety and depressive disorders with better stratification of case definitions are needed.19,20

Here, we investigated the genetic relationship between ANX and MD using two large population-based genotyped samples with clinical psychiatric diagnoses. We conducted non-comorbid ANX-only and MD-only GWAS of the UK Biobank sample and compared them with GWAS including comorbid cases (ANX-co and MD-co). We investigated the genetic relationship between ANX-only and MD-only compared to ANX-co and MD-co, and how the respective PRS associated with ANX and MD diagnoses and symptoms in an independent sample (Norwegian Mother, Father and Child Study, MoBa). We also explored differences in biological pathways and how comorbidity affected causal relationships between ANX and MD using Mendelian Randomization (MR).22,23

Methods

We used two large genotyped cohorts with data on relevant mental phenotypes in Europeans for training and testing: the UK Biobank (UKB; Project No. 27412) for GWAS (training) (eAppendix 1 in Supplement 1), and the Norwegian Mother, Father, and Child Cohort (MoBa) for independent testing (eAppendix 2 in Supplement 1). These cohorts are linked to clinical diagnoses from national healthcare records. We used publicly available GWAS summary data without the testing sample (MoBa) of ANX-co (cases=31,977, controls=82,114),24 and MD-co (cases=170,756, controls=329,443)25 as comparators. These GWAS did not exclude cases with ANX and MD comorbidity (eAppendix 3 in Supplement 1).

GWAS Dataset (Training)

We used UK Biobank data to perform GWAS for ANX-only (cases = 9,980) and MD-only (cases = 15,301). Individuals with a history of diagnoses of organic mental disorders (F00 – F09), psychotic disorders (F20 – F29), bipolar disorders (F30.0 – F30.9 and F31.0 – F31.9), or mixed anxiety and depressive disorder (F41.2) were excluded from all analyses. The ANX-only cases were diagnosed with agoraphobia (F40.0), social phobia (F40.1), panic disorder (F41.0), generalized anxiety disorder (F41.1), or other anxiety disorders (F41.3 - F41.9), excluding those with ‘MD-only’ diagnoses (see below). MD-only cases had a diagnosis of any depressive episode (F32.0 - F32.9), or recurrent depressive disorder (F33.0 - F33.9), excluding those with ‘ANX-only’ diagnoses (see above). Controls for ANX-only (n=179,442) and MD-only (n=179,038) were randomly selected and did not overlap. Individuals with MD were excluded from the ANX-only GWAS control group, and those with ANX were excluded from the MD-only GWAS control group.

Polygenic risk score dataset (Testing)

We obtained genotype and phenotype data on mothers and fathers (n = 130,992) who participated in MoBa.26,27 All study participants provided written informed consent upon recruitment and those who withdrew during follow-up were excluded from our analyses. The Regional Committee for Medical and Health Research Ethics approved the present study (2016/1226/REK sør-øst C).

Data on previous or current diagnoses of mental disorders were obtained from the Norwegian Patient Registry (2008 - 2022). We excluded individuals who withdrew their consent (n=484), with missing covariates (n=1,590), and related individuals. We excluded individuals with a diagnosis of organic mental disorders (F00 – F09), psychotic disorders (F20 – F29), or bipolar disorders (F30.0 – F30.9 and F31.0 – F31.9) (eFigure 1 in Supplement 1).

Diagnoses

We classified the cases within the MoBa sample into three disorder-specific subgroups: 1. Dx-ANX-only (n=1,992) with the same criteria as in the UKB and excluding individuals with MD. 2. Dx-MD-only (n=7,486) using the same criteria as in the UKB and excluding individuals with ANX. 3. Dx-ANX-MD (n=3,468) comprising individuals with both MD and ANX or a mixed anxiety and depressive disorder (F41.2). The number of excluded individuals and the corresponding numbers of controls are provided in the supplement (eFigure 1 in Supplement 1).

ANX and MD symptoms

Symptoms of ANX and MD were assessed using the validated 8-item Hopkins Symptom Checklist (SCL-8), which correlates highly with the original 25-item HSCL.28 Responses were rated on a Likert scale from 1 to 4 reflecting how much respondents were bothered by specific symptoms in the preceding two weeks. Pregnant women participating in MoBa (n = 67,219) completed the SCL-8 during the 30th week of gestation. After excluding individuals with missing covariates (n=787) and related individuals (n=11,570), the analysis included 54,862 mothers (eFigure 2 in Supplement 1).

Genome-wide association study (GWAS)

We performed GWAS for ANX-only and MD-only using version 3 of the UKB imputed genetic data among individuals of “white British” ancestry classified according to self-declared ethnicity and genetic principal component (PC) analysis (UKB data field 22006). The GWAS was done with REGENIE v3.4.1 including related individuals. Variants with an imputation INFO score < 0.8 and minor allele count < 20 were excluded. Age, sex, and the first 20 genotype PCs were used as covariates.

Polygenic risk score (PRS) calculation

PRS were calculated using ANX-only and MD-only GWAS summary statistics. Additionally, a PRS was calculated from the summary statistics of a case-case GWAS (CC-GWAS) performed on ANX-only and MD-only GWAS. Further PRSs were derived from GWAS that included comorbidity i.e., ANX-co24 and MD-co.25 All PRSs were computed using the PRS-PCA approach, extracting PCs PRSs generated by PRSice (2.3.3) for a set of p-value cut-offs (1.0e-06, 1.0e-05, 1.0e-04, 1.0e-03, 1.0e-02, 5.0e-02, 1.0e-01, 5.0e-01, 1.0).29 In logistic regression models, the first PC was used for the prediction of phenotypes in the MoBa sample.

Statistical analysis

PRS Associations

Logistic regression analyses were performed to test the association between each PRS (i.e., ANX-only, MD-only, MD-co, ANX-co, and CC-ANX-MD), and each diagnostic category (i.e., Dx-ANX-only, Dx-MD-only, and Dx-ANX-MD) compared to controls. Additionally, logistic regression was used to assess whether each PRS is differentially associated with one diagnostic category over another. For depression and anxiety symptoms, responses were dichotomized into ‘1’ for ‘3-Quite bothered’ or ‘4-Very bothered’, and ‘0’ for ‘1-Not bothered’ or ‘2-A little bothered’. Logistic regressions were then used to analyze the association between each PRS and the dichotomized symptom responses. All regression models were adjusted for age, sex, and the first 10 genotype PCs. Only one of the related individuals with a kinship coefficient greater than 0.05 was kept, prioritizing cases. Effect sizes were represented as standardized β values with 95% confidence intervals. Bonferroni correction was applied to p-values to account for multiple comparisons.

Genetic Correlations

Linkage disequilibrium score regression30 was applied to the GWAS summary data to estimate genetic correlations between ANX and MD with or without comorbid cases. The single nucleotide polymorphism (SNP) heritability for each phenotype was also computed.

Mendelian Randomization (MR)

We applied MR analyses using the R package TwoSampleMR31 to examine causal associations between ANX and MD. To enhance result reliability, multiple MR methods were applied: inverse variance weighted (IVW),22 weighted median,32 and MR Egger.33 Instruments were selected at a p-value threshold of 1e-05 to account for the low power of the ANX-only and MD-only GWASs,34 following previous studies’ application of more relaxed p-value threshold for low-power GWASs.35,36 Furthermore, the Causal Analysis Using Summary Effect estimates (CAUSE) method was applied to further assess causality and distinguish horizontal pleiotropy.23

Functional Enrichment of Genes

To explore biological differences between ANX and MD, we employed GSA-MiXeR which can provide more specific genes and pathways relevant to a given phenotype. GSA-MiXeR estimates gene-level heritability to compute fold enrichment from the respective GWAS summary statistics,37 with reliability assessed using Akaike Information Criterion (AIC) values. We compared the enriched genes for ANX-only and MD-only as well as ANX-co and MD-co based on AIC and fold enrichment values.

Results

Sample characteristics

UKB data: ANX-only GWAS consisted of 189,422 (54.0 % females) with a mean age(SD) of 56.8 (8.0) years. MD-only GWAS comprised 194,339 participants (54.1 % females) with a mean age (SD) of 56.7 (8.0) years. MoBa data: A total of 128,918 individuals (58.9% females) with a mean age (SD) of 49.0 (5.4) years had genetic and phenotype data for analysis. For the prediction of symptoms of anxiety and depression, the sample included 54,862 mothers with a mean age (SD) of 48.1 (5.0) years.

PRS associations with diagnoses vs controls in MoBa

Almost all PRSs were positively associated with all three diagnostic categories (i.e. Dx-ANX-only, Dx-MD-only, and Dx-ANX-MD). However, CC-ANX-MD PRS exhibited a negative association with Dx-MD-only and no significant association with Dx-ANX-MD. MD-only PRS had stronger associations with Dx-MD-only (β=0.13; 95%CI: 0.10, 0.15) and Dx-ANX-MD (β=0.15; 95%CI: 0.12, 0.18) than with Dx-ANX-only (β=0.05; 95%CI: 0.01, 0.08). ANX-co PRS and MD-co PRS showed stronger associations than the ANX-only PRS and MD-only PRS when comparing Dx-ANX-MD or Dx-ANX-only cases to controls (Figure 1, Panel A).

Figure 1.
Figure 1.

The association of PRS of differently defined anxiety and depressive disorders with anxiety disorder only (Dx-ANX-only), major depression only (Dx-MD-only), or comorbid anxiety and depression (Dx-ANX-MD) compared to controls (CTRL) (Panel A) or compared to each of the other diagnostic categories (Panel B). PRS: polygenic risk score; CI: confidence interval, ANX-co: anxiety disorder including comorbid depression, MD-co: major depression including comorbid anxiety, CC-ANX-MD: case-case GWAS of ANX-only and MD-only. Phenotypes are displayed on the x-axis and PRS on the y-axis.

PRS associations between diagnoses in MoBa

We assessed the association of PRSs with each diagnostic group, using another diagnostic group as a control. As expected, MD-only PRS and MD-co PRS were associated with Dx-ANX-MD compared to Dx-ANX-only. Notably, MD-co PRS was associated with Dx-ANX-MD compared to Dx-MD-only. Unlike MD-co PRS, MD-only PRS showed a positive association with Dx-MD-only compared to Dx-ANX-only. The association of ANX-only PRS with Dx-ANX-only compared to Dx-MD-only appeared consistent but was not statistically significant after correction for multiple testing. CC-ANX-MD PRS was associated with Dx-ANX-only compared to Dx-MD-only (Figure 1, Panel B).

PRS associations with symptoms of anxiety and depression in MoBa

MD-co PRS was associated with all eight symptoms on SCL-8 consistent with the large GWAS sample size and inclusion of comorbid ANX. ANX-only PRS was associated with all anxiety items but only two depression symptoms (worry and hopelessness). The item ‘suddenly scared for no reason’ had the lowest endorsement (1%) which may have contributed to the low precision and non-significant associations with ANX-co PRS and MD-only PRS (Figure 2). Overall, all PRSs generated from the GWAS of a range of diagnostic categories exhibited a similar pattern of prediction of anxiety and depressive symptoms. In contrast, CC-ANX-MD PRS was not associated with any SCL-8 items, highlighting that the symptoms span the ANX and MD continuum.

Figure 2.
Figure 2.

The association of polygenic scores (PRS) of differently defined anxiety and depressive disorders with symptoms of depression (Blue, Effort, Hopeless, Worry) and symptoms of anxiety (Nervous, Fearful, Scared, Tension) reported by mothers from MoBa. ANX-co: anxiety disorder including comorbid depression, MD-co: major depression including comorbid anxiety, ANX-only: anxiety disorder only, MD-only: major depression only, CC-ANX-MD: case-case GWAS of ANX-only and MD-only.

Genetic Correlation and Mendelian Randomization (MR)

The genetic correlation between MD-co and ANX-co was higher 0.90 (p = 5.4e-196) than that between MD-only and ANX-only, 0.53 (p = 2.7e-06). SNP heritability was 8.4% for MD-co, 9.5% for ANX-co, 10.5% for MD-only, and 5.4% for ANX-only.

At least two MR methods showed evidence of a bidirectional causal association between MD-co and ANX-co. In contrast, only ANX-only had a significant causal association with MD-only (Table 1). The CAUSE-MR method, which accounts for potential horizontal pleiotropy suggested causal associations of ANX-co and MD-co, however, these results were not statistically significant after correction for multiple testing (Table 1).

View this table:
Table 1. Bidirectional Mendelian Randomization Analyses of Anxiety Disorders and Major Depression

Enrichment of Genes and Pathways

GSA-MiXeR analyses identified significant fold enrichment for 148 genes in ANX-only and 155 genes in MD-only, with 13 genes enriched in both (ERBB4, ACSL3, TPRG1, MAML3, KCNQ5, INSIG1, KCNB2, FER1L6, B3GLCT, IRX6, TSHZ3, TOX2, and FAM110A). Enriched pathways for ANX-only included: microvesicle, CCR6 chemokine receptor binding, and cardiac muscle myoblast proliferation (Table S1 in Supplement 2). For MD-only, pathways included negative regulation of myoblast proliferation, interleukin 21 production, protein-O-linked fucosylation, and positive regulation of 3-UTR-mediated mRNA stabilization (Table S2 in Supplement 2). More genes showed enrichment in ANX-co (n=190) and MD-co (n=331) (Tables S3 & S4 in Supplement 2) with 48 genes common to both (Figure 3).

Figure 3.
Figure 3.

Venn diagram – The numbers of genes that showed enrichment in GSA-MiXeR analyses using GWAS data for different diagnoses. ANX-co: anxiety disorder including comorbid depression, MD-co: major depression including comorbid anxiety, ANX-only: anxiety disorder only, MD-only: major depression only.

Discussion

Here, we dissected the genetic relationship between ANX and MD using GWAS with and without comorbid cases. ANX-only and MD-only showed lower genetic correlations compared to ANX-co and MD-co. MD-only PRS was differentially associated with Dx-MD-only compared to Dx-ANX-only, while MD-co PRS was associated with both. Our results suggest ANX-only is causally associated with MD-only but not vice versa, whereas ANX-co and MD-co support a bidirectional causal association. Additionally, ANX-co PRS and MD-co PRS had a stronger association with Dx-ANX-MD and did not differentiate between Dx-ANX-only and Dx-MD-only.

The high comorbidity between ANX and MD, and the inclusion of comorbid cases in major GWAS, can confound the reported genetic overlap between the two disorders.9,13,16,24,38 Our findings show lower genetic correlations when comorbid cases are excluded, aligning with a recent family-based adoption study.39 Since the SNP heritabilities for ANX-only and MD-only were comparable to previous reports,24,25 low power alone is unlikely to explain the diminished genetic correlation.30,40

Our MR results from ANX-only and MD-only suggest a direction of causation aligning with epidemiological data,7,10,41 unlike the bidirectional causal association observed using GWAS including comorbid cases. Thus, investigating the genetic relationships between diagnoses with pervasive comorbidity without removing comorbid cases from the initial GWAS may be misleading. Epidemiological studies show that ANX often precedes MD,10,42,43 and longitudinal clinical data support that ANX causes MD but not vice versa.7 A large proportion of individuals with MD (46 – 75%) have a history of ANX, which is associated with greater severity, chronicity, and impairment.43–46 Furthermore, the importance of anxiety symptoms in MD has also been highlighted with the anxious distress specifier in DSM-5.47,48 Therefore, targeting ANX symptoms in the early phase of MD may improve clinical outcomes.

The MD-co PRS25 showed the strongest association with Dx-ANX-only, Dx-MD-only, and Dx-ANX-MD cases compared to controls. The MD-co PRS also demonstrated the greatest prediction of ANX caseness, probably because of its large sample size and inclusion of comorbid ANX in the GWAS.9 Consistent with our findings, a stronger association of MD-co PRS with Dx-ANX-MD than with Dx-MD-only was reported using electronic health records data from the US.19 This may be attributed to a stronger genetic burden among comorbid cases,49 an overrepresentation of comorbid cases in samples included in the GWAS.25 The difference in the associations of the ANX-co PRS with Dx-ANX-MD and those with Dx-ANX-only were not as large, which may be due to the lower power of the ANX-co GWAS.24 Additionally, diagnostic hierarchy requires the exclusion of MD before making a diagnosis of ANX but not vice versa,48 which may lead to underreporting of ANX among individuals with MD.50

Taken together, our findings indicate that the genetic underpinnings of ANX and MD are more distinct than previously suggested.19 Thus, the disease pathways of ANX-only may be different from MD-only or their comorbid state which is relevant for clinical translation and precision psychiatry. Notably, less than 10% of genes enriched based on the heritabilities of MD-only and ANX-only were common to both. This contrasts with our recent report of many shared genetic loci and extensive overlap (85%) of trait-influencing variants between ANX and MD without excluding comorbid cases.9,13 GSA-MiXeR showed 25% of genes enriched in ANX-co were also enriched in MD-co. GWAS of strictly defined MD diagnosis can help identify disorder-specific biological mechanisms.21 For example, our data suggests that immune pathways, such as CCR6-chemokine receptor binding, are implicated in ANX-only, while interleukin-21 production was enriched in MD-only. Intriguingly, interleukin-21 production was also enriched in ANX-co. MD-only PRS and CC-ANX-MD PRS may facilitate the differentiation of clinical subgroups. Notably, the CC-ANX-MD PRS showed a negative association with Dx-MD-only suggesting that it could be a marker of resilience to MD for individuals with ANX.

The pattern of positive association between symptoms (SCL-8) and the range of PRS highlights the pervasive overlap between these disorders at the symptom level.48 Also, anxiety and depressive symptoms have significant correlations with each other.51 The deviation of associations with the item ‘feeling suddenly scared for no reason’ may be related to its low prevalence and large confidence intervals. The CC-ANX-MD PRS showed no association with any symptom suggesting that all eight symptoms span across the ANX – MD dichotomy.

Current efforts to improve genetic discoveries from GWAS of complex phenotypes such as mental disorders emphasize the need for increasing sample sizes.52 However, the current findings support that more narrowly defined, homogeneous phenotypes are more likely to enable the identification of specific genetic variants with significant functional effects.20 The observed differential association of the MD-only PRS and CC-ANX-MD PRS with Dx-MD-only compared to Dx-ANX-only provide support for disorder-specific genetic risk which may have implications for the prediction of treatment outcome.53,54

Our findings should be interpreted in the light of some limitations. The cases identified for the UKB and MoBa data were defined based on available health registry data and may still include some comorbid cases. However, this would bias the results of disorder-specific GWAS towards those of comorbid ones. The results are based on European populations and may not be generalizable to other ancestries. The limitations of our ANX-only GWAS may be biased by low power. Accordingly, the effect directions were consistent with expectations even though these associations did not stand correction for multiple testing. As better powered GWAS of more deeply phenotyped cases become available, the utility of genetic factors in treatment selection and outcome prediction can be realized. Hence, global efforts for genetic discoveries with better phenotypic characterization are needed for precision psychiatry.

Conclusions

Our findings suggest that the genetic liabilities for ANX and MD are more distinct than has been inferred from GWAS where comorbid cases were included, implicating more separate underlying biological pathways. Further, we show that ANX may have a causal link to MD, suggesting that targeting ANX symptoms in early phase MD can be an effective intervention. Together, genetic factors derived from ANX and MD excluding comorbidities may be useful in developing precision medicine approaches and uncovering novel biological mechanisms for drug targets.

Data Availability

All data produced in the present study are available upon reasonable request to the authors.

Authors’ Contributions

Dr. Tesfaye had full access to all of the data in the study and takes full responsibility for the integrity of the data and the accuracy of data analyses. Concept and design: Tesfaye, Parker, Jaholkowski, O’Connell, Andreassen. Acquisition, analysis, or interpretation of data: Tesfaye, Shadrin, Jaholkowski, Parker, Smeland, Parekh, Kutrolli, Birkenæs, Bakken, Ask, Frei, Dale, O’Connell, Andreassen. Drafting of the manuscript: Tesfaye, Shadrin, Parker, Andreassen. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: Tesfaye, Shadrin, Parker, Jaholkowski. Obtained funding: Andreassen. Administrative, technical, or material support: Kutrolli, Parekh, Bakken, Birkenæs, Ask, O’Connell, Shadrin, Andreassen. Supervision: Frei, O’Connell, Andreassen.

Conflict of Interest Disclosures

Ole A. Andreassen is a consultant for Cortechs.ai and Precision Health and has received speaker’s honoraria from Lundbeck, Janssen, Otsuka, and Sunovion. Srdjan Djurovic has received speaker’s honoraria from Lundbeck. Dr. Anders M. Dale is the Founding Director, holds equity in CorTechs Labs, Inc. (DBA Cortechs.ai), and serves on its Board of Directors and the Scientific Advisory Board. He is an unpaid consultant for Oslo University Hospital. The terms of these arrangements have been reviewed and approved by the University of California, San Diego in accordance with its conflict-of-interest policies. Dr. Frei is a consultant to Precision Health. The other authors have no conflicts of interest to declare.

Funding/Support

The authors received support support from the Research Council of Norway (324252, 324499, 300309, 273291, 223273, 248778, 273446, 326813, 334920, 271555/F21, 324620), NordForsk (grant 156298), the European Economic Area and Norway Grants (EEA-RO-NO-2018-0535, EEA-RO-NO-2018-0573), the South-East Norway Regional Health Authority (2022-087 and 2022-073), The University of Oslo, KG Jebsen Stiftelsen (SKGJ-MED-021), the European Union’s Horizon 2020 research and innovation program (847776, 964874), and Marie Skłodowska-Curie grant (801133), and the US National Institutes of Health (R01MH123724-01, R01MH124839-02, U24DA041123, R01AG076838, U24DA055330, and OT2 HL161847). The Norwegian Mother, Father and Child Cohort Study is supported by the Norwegian Ministry of Health and Care Services and the Ministry of Education and Research. This research is part of the HARVEST collaboration, supported by the Research Council of Norway (grant 229624). The Norwegian Centre for Mental Disorders Research (NORMENT) provided genotype data, funded by the Research Council of Norway (grant 223273), South East Norway Health Authorities, and Stiftelsen Kristian Gerhard Jebsen. The Center for Diabetes Research, the University of Bergen, provided genotype data funded by the ERC AdG project SELECTionPREDISPOSED, Stiftelsen Kristian Gerhard Jebsen, the Trond Mohn Foundation, the Research Council of Norway, the Novo Nordisk Foundation, the University of Bergen, and theWestern Norway Health Authorities.

Role of the Funder/Sponsor

The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions

We are grateful to all individuals who participated in the Norwegian MoBa study. We thank the Norwegian Institute of Public Health for generating high-quality genomic data. We also thank the research participants and employees of the Psychiatric Genomic Consortium (Anxiety Disorders, and Major Depressive Disorder Working Groups), the UK Biobank, and iPSYCH.

Additional Information

This work was performed on the TSD (Tjeneste for Sensitive Data) facilities, owned by the University of Oslo, operated and developed by the TSD service group at the University of Oslo, IT-Department (USIT) (tsd-drift{at}usit.uio.no). This research has been conducted using the UK Biobank Resource under Application Number 27412.

Disclaimer

Data from the Norwegian Patient Registry has been used in this publication. The interpretation and reporting of these data are the sole responsibility of the authors, and no endorsement by the Norwegian Patient Registry is intended nor should be inferred.

References

  1. 1.↵
    Collaborators GBDMD. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry. Feb 2022;9(2):137–150. doi:10.1016/S2215-0366(21)00395-3
    OpenUrlCrossRefPubMed
  2. 2.↵
    Penninx BW, Pine DS, Holmes EA, Reif A. Anxiety disorders. Lancet. Mar 6 2021;397(10277):914–927. doi:10.1016/S0140-6736(21)00359-7
    OpenUrlCrossRefPubMed
  3. 3.↵
    Kessler RC, Berglund P, Demler O, et al. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA. Jun 18 2003;289(23):3095–105. doi:10.1001/jama.289.23.3095
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.↵
    Ter Meulen WG, Draisma S, van Hemert AM, et al. Depressive and anxiety disorders in concert-A synthesis of findings on comorbidity in the NESDA study. J Affect Disord. Apr 1 2021;284:85–97. doi:10.1016/j.jad.2021.02.004
    OpenUrlCrossRefPubMed
  5. 5.
    Zhou Y, Cao Z, Yang M, et al. Comorbid generalized anxiety disorder and its association with quality of life in patients with major depressive disorder. Sci Rep. Jan 18 2017;7:40511. doi:10.1038/srep40511
    OpenUrlCrossRef
  6. 6.↵
    Kalin NH. The Critical Relationship Between Anxiety and Depression. Am J Psychiatry. May 1 2020;177(5):365–367. doi:10.1176/appi.ajp.2020.20030305
    OpenUrlCrossRefPubMed
  7. 7.↵
    Mathew AR, Pettit JW, Lewinsohn PM, Seeley JR, Roberts RE. Co-morbidity between major depressive disorder and anxiety disorders: shared etiology or direct causation? Psychol Med. Oct 2011;41(10):2023–34. doi:10.1017/S0033291711000407
    OpenUrlCrossRefPubMed
  8. 8.
    Thorp JG, Campos AI, Grotzinger AD, et al. Symptom-level modelling unravels the shared genetic architecture of anxiety and depression. Nat Hum Behav. Oct 2021;5(10):1432–1442. doi:10.1038/s41562-021-01094-9
    OpenUrlCrossRef
  9. 9.↵
    Tesfaye M, Jaholkowski P, Shadrin AA, et al. Identification of novel genomic loci for anxiety symptoms and extensive genetic overlap with psychiatric disorders. Psychiatry Clin Neurosci. Sep 20 2024;doi:10.1111/pcn.13742
    OpenUrlCrossRef
  10. 10.↵
    Wittchen HU, Kessler RC, Pfister H, Lieb M. Why do people with anxiety disorders become depressed? A prospective-longitudinal community study. Acta Psychiatr Scand Suppl. 2000;(406):14–23.
    OpenUrlPubMed
  11. 11.↵
    Hettema JM. What is the genetic relationship between anxiety and depression? Am J Med Genet C Semin Med Genet. May 15 2008;148C(2):140–6. doi:10.1002/ajmg.c.30171
    OpenUrlCrossRefPubMed
  12. 12.
    McTeague LM, Rosenberg BM, Lopez JW, et al. Identification of Common Neural Circuit Disruptions in Emotional Processing Across Psychiatric Disorders. Am J Psychiatry. May 1 2020;177(5):411–421. doi:10.1176/appi.ajp.2019.18111271
    OpenUrlCrossRefPubMed
  13. 13.↵
    Tesfaye M, Jaholkowski P, Shadrin AA, et al. Identification of Novel Genomic Loci for Anxiety and Extensive Genetic Overlap with Psychiatric Disorders. medRxiv. 2023:2023.09.01.23294920. doi:10.1101/2023.09.01.23294920
    OpenUrlAbstract/FREE Full Text
  14. 14.↵
    Shimada-Sugimoto M, Otowa T, Hettema JM. Genetics of anxiety disorders: Genetic epidemiological and molecular studies in humans. Psychiatry Clin Neurosci. Jul 2015;69(7):388–401. doi:10.1111/pcn.12291
    OpenUrlCrossRef
  15. 15.↵
    Sullivan PF, Neale MC, Kendler KS. Genetic epidemiology of major depression: review and meta-analysis. Am J Psychiatry. Oct 2000;157(10):1552–62. doi:10.1176/appi.ajp.157.10.1552
    OpenUrlCrossRefPubMedWeb of Science
  16. 16.↵
    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. Mar 1 2020;177(3):223–232. doi:10.1176/appi.ajp.2019.19030256
    OpenUrlCrossRefPubMed
  17. 17.↵
    Strom NI, Verhulst B, Bacanu S-A, et al. Genome-wide association study of major anxiety disorders in 122,341 European-ancestry cases identifies 58 loci and highlights GABAergic signaling. medRxiv. 2024:2024.07.03.24309466. doi:10.1101/2024.07.03.24309466
    OpenUrlAbstract/FREE Full Text
  18. 18.↵
    McIntosh AM, Lewis CM. Genome-wide study of half a million individuals with major depression identifies 697 independent associations, infers causal neuronal subtypes and biological targets for novel pharmacotherapies. medRxiv. 2024:2024.04.29.24306535. doi:10.1101/2024.04.29.24306535
    OpenUrlAbstract/FREE Full Text
  19. 19.↵
    Coombes BJ, Landi I, Choi KW, et al. The genetic contribution to the comorbidity of depression and anxiety: a multi-site electronic health records study of almost 178 000 people. Psychol Med. Nov 2023;53(15):7368–7374. doi:10.1017/S0033291723000983
    OpenUrlCrossRefPubMed
  20. 20.↵
    Crouch DJM, Bodmer WF. Polygenic inheritance, GWAS, polygenic risk scores, and the search for functional variants. Proc Natl Acad Sci U S A. Aug 11 2020;117(32):18924–18933. doi:10.1073/pnas.2005634117
    OpenUrlAbstract/FREE Full Text
  21. 21.↵
    Cai N, Revez JA, Adams MJ, et al. Minimal phenotyping yields genome-wide association signals of low specificity for major depression. Nat Genet. Apr 2020;52(4):437–447. doi:10.1038/s41588-020-0594-5
    OpenUrlCrossRefPubMed
  22. 22.↵
    Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. Nov 2013;37(7):658–65. doi:10.1002/gepi.21758
    OpenUrlCrossRefPubMed
  23. 23.↵
    Morrison J, Knoblauch N, Marcus JH, Stephens M, He X. Mendelian randomization accounting for correlated and uncorrelated pleiotropic effects using genome-wide summary statistics. Nat Genet. Jul 2020;52(7):740–747. doi:10.1038/s41588-020-0631-4
    OpenUrlCrossRefPubMed
  24. 24.↵
    Purves KL, Coleman JRI, Meier SM, et al. A major role for common genetic variation in anxiety disorders. Mol Psychiatry. Dec 2020;25(12):3292–3303. doi:10.1038/s41380-019-0559-1
    OpenUrlCrossRef
  25. 25.↵
    Howard DM, Adams MJ, Clarke TK, et al. Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions. Nat Neurosci. Mar 2019;22(3):343–352. doi:10.1038/s41593-018-0326-7
    OpenUrlCrossRefPubMed
  26. 26.↵
    Magnus P, Birke C, Vejrup K, et al. Cohort Profile Update: The Norwegian Mother and Child Cohort Study (MoBa). Int J Epidemiol. Apr 2016;45(2):382–8. doi:10.1093/ije/dyw029
    OpenUrlCrossRefPubMed
  27. 27.↵
    Paltiel L, Haugen A, Skjerden T, et al. The biobank of the Norwegian Mother and Child Cohort Study – present status. Nor J Epidemiol. 2014;24(1-2):29–35. doi:10.5324/nje.v24i1-2.1755
    OpenUrlCrossRef
  28. 28.↵
    Tamb K, Røysamb E. Selection of questions to short-form versions of original psychometric instruments in MoBa. Norsk Epidemiologi. 2014;24(1-2):195–201. doi:10.5324/nje.v24i1-2.1822
    OpenUrlCrossRef
  29. 29.↵
    Coombes BJ, Ploner A, Bergen SE, Biernacka JM. A principal component approach to improve association testing with polygenic risk scores. Genet Epidemiol. Oct 2020;44(7):676–686. doi:10.1002/gepi.22339
    OpenUrlCrossRefPubMed
  30. 30.↵
    Bulik-Sullivan BK, Loh PR, Finucane HK, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. Mar 2015;47(3):291–5. doi:10.1038/ng.3211
    OpenUrlCrossRefPubMed
  31. 31.↵
    Hemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. May 30 2018;7 doi:10.7554/eLife.34408
    OpenUrlCrossRefPubMed
  32. 32.↵
    Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet Epidemiol. May 2016;40(4):304–14. doi:10.1002/gepi.21965
    OpenUrlCrossRefPubMed
  33. 33.↵
    Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. Apr 2015;44(2):512–25. doi:10.1093/ije/dyv080
    OpenUrlCrossRefPubMed
  34. 34.↵
    Cheng W, Parker N, Karadag N, et al. The relationship between cannabis use, schizophrenia, and bipolar disorder: a genetically informed study. Lancet Psychiatry. Jun 2023;10(6):441–451. doi:10.1016/S2215-0366(23)00143-8
    OpenUrlCrossRefPubMed
  35. 35.↵
    Li F, Tang M, Hao C, Yang M, Pan Y, Lei P. Brain imaging traits and epilepsy: Unraveling causal links via mendelian randomization. Brain Behav. Oct 2024;14(10):e70051. doi:10.1002/brb3.70051
    OpenUrlCrossRefPubMed
  36. 36.↵
    Sanna S, van Zuydam NR, Mahajan A, et al. Causal relationships among the gut microbiome, short-chain fatty acids and metabolic diseases. Nat Genet. Apr 2019;51(4):600–605. doi:10.1038/s41588-019-0350-x
    OpenUrlCrossRefPubMed
  37. 37.↵
    Frei O, Hindley G, Shadrin AA, et al. Improved functional mapping of complex trait heritability with GSA-MiXeR implicates biologically specific gene sets. Nat Genet. Jun 2024;56(6):1310–1318. doi:10.1038/s41588-024-01771-1
    OpenUrlCrossRefPubMed
  38. 38.↵
    Ohi K, Otowa T, Shimada M, Sasaki T, Tanii H. Shared genetic etiology between anxiety disorders and psychiatric and related intermediate phenotypes. Psychol Med. Mar 2020;50(4):692–704. doi:10.1017/S003329171900059X
    OpenUrlCrossRefPubMed
  39. 39.↵
    Kendler KS, Abrahamsson L, Ohlsson H, Sundquist J, Sundquist K. An Extended Swedish Adoption Study of Anxiety Disorder and Its Cross-Generational Familial Relationship With Major Depression. Am J Psychiatry. Sep 2022;179(9):640–649. doi:10.1176/appi.ajp.21111110
    OpenUrlCrossRefPubMed
  40. 40.↵
    van Rheenen W, Peyrot WJ, Schork AJ, Lee SH, Wray NR. Genetic correlations of polygenic disease traits: from theory to practice. Nat Rev Genet. Oct 2019;20(10):567–581. doi:10.1038/s41576-019-0137-z
    OpenUrlCrossRefPubMed
  41. 41.↵
    Wittchen HU, Beesdo K, Bittner A, Goodwin RD. Depressive episodes--evidence for a causal role of primary anxiety disorders? Eur Psychiatry. Dec 2003;18(8):384–93. doi:10.1016/j.eurpsy.2003.10.001
    OpenUrlCrossRefPubMedWeb of Science
  42. 42.↵
    Ohayon MM, Schatzberg AF. Social phobia and depression: prevalence and comorbidity. J Psychosom Res. Mar 2010;68(3):235–43. doi:10.1016/j.jpsychores.2009.07.018
    OpenUrlCrossRefPubMed
  43. 43.↵
    Kessler RC, Sampson NA, Berglund P, et al. Anxious and non-anxious major depressive disorder in the World Health Organization World Mental Health Surveys. Epidemiol Psychiatr Sci. Jun 2015;24(3):210–26. doi:10.1017/S2045796015000189
    OpenUrlCrossRefPubMed
  44. 44.
    Fava M, Rush AJ, Alpert JE, et al. Difference in treatment outcome in outpatients with anxious versus nonanxious depression: a STAR*D report. Am J Psychiatry. Mar 2008;165(3):342–51. doi:10.1176/appi.ajp.2007.06111868
    OpenUrlCrossRefPubMedWeb of Science
  45. 45.
    Wiethoff K, Bauer M, Baghai TC, et al. Prevalence and treatment outcome in anxious versus nonanxious depression: results from the German Algorithm Project. J Clin Psychiatry. Aug 2010;71(8):1047–54. doi:10.4088/JCP.09m05650blu
    OpenUrlCrossRefPubMedWeb of Science
  46. 46.↵
    Lamers F, van Oppen P, Comijs HC, et al. Comorbidity patterns of anxiety and depressive disorders in a large cohort study: the Netherlands Study of Depression and Anxiety (NESDA). J Clin Psychiatry. Mar 2011;72(3):341–8. doi:10.4088/JCP.10m06176blu
    OpenUrlCrossRefPubMed
  47. 47.↵
    Gaspersz R, Lamers F, Kent JM, et al. Longitudinal Predictive Validity of the DSM-5 Anxious Distress Specifier for Clinical Outcomes in a Large Cohort of Patients With Major Depressive Disorder. J Clin Psychiatry. Feb 2017;78(2):207–213. doi:10.4088/JCP.15m10221
    OpenUrlCrossRefPubMed
  48. 48.↵
    APA. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders. Fifth ed. DSM-5. American Psychiatric Association; 2013.
  49. 49.↵
    Tabrizi F, Rosén J, Grönvall H, et al. Heritability and polygenic load for combined anxiety and depression. medRxiv. 2024:2024.01.31.24302045. doi:10.1101/2024.01.31.24302045
    OpenUrlAbstract/FREE Full Text
  50. 50.↵
    Brown TA, Campbell LA, Lehman CL, Grisham JR, Mancill RB. Current and lifetime comorbidity of the DSM-IV anxiety and mood disorders in a large clinical sample. J Abnorm Psychol. Nov 2001;110(4):585–99. doi:10.1037//0021-843x.110.4.585
    OpenUrlCrossRefPubMedWeb of Science
  51. 51.↵
    Vaccarino AL, Evans KR, Sills TL, Kalali AH. Symptoms of anxiety in depression: assessment of item performance of the Hamilton Anxiety Rating Scale in patients with depression. Depress Anxiety. 2008;25(12):1006–13. doi:10.1002/da.20435
    OpenUrlCrossRefPubMed
  52. 52.↵
    Abdellaoui A, Yengo L, Verweij KJH, Visscher PM. 15 years of GWAS discovery: Realizing the promise. Am J Hum Genet. Feb 2 2023;110(2):179–194. doi:10.1016/j.ajhg.2022.12.011
    OpenUrlCrossRef
  53. 53.↵
    Dold M, Bartova L, Souery D, et al. Clinical characteristics and treatment outcomes of patients with major depressive disorder and comorbid anxiety disorders - results from a European multicenter study. J Psychiatr Res. Aug 2017;91:1–13. doi:10.1016/j.jpsychires.2017.02.020
    OpenUrlCrossRef
  54. 54.↵
    Hicks PB, Sevilimedu V, Johnson GR, et al. Factors Affecting Antidepressant Response Trajectories: A Veterans Affairs Augmentation and Switching Treatments for Improving Depression Outcomes Trial Report. Psychiatr Res Clin Pract. Winter 2023;5(4):131–143. doi:10.1176/appi.prcp.20230017
    OpenUrlCrossRefPubMed
Back to top
Posted November 20, 2024.
Download PDF

Supplementary Material

Data/Code
Email
Comorbidity alters the genetic relationship between anxiety disorders and major depression
Markos Tesfaye, Alexey Shadrin, Nadine Parker, Piotr Jaholkowski, Pravesh Parekh, Gleda Kutrolli, Viktoria Birkenæs, Nora R. Bakken, Helga Ask, Oleksandr Frei, Srdjan Djurovic, Anders M. Dale, Olav B. Smeland, Kevin S. O’Connell, Ole A. Andreassen
medRxiv 2024.11.19.24317523; doi: https://doi.org/10.1101/2024.11.19.24317523
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area