A higher genetic risk for depression is a greater risk factor for cardiovascular disease in women
=================================================================================================

* Jiayue-Clara Jiang
* Kritika Singh
* Lea K. Davis
* Naomi R. Wray
* Sonia Shah

## Abstract

**Background** Epidemiological studies show increased cardiovascular disease (CVD) risks amongst individuals with psychiatric disorders, but to what degree this comorbidity is mediated by genetic and non-genetic (including medications) factors remains unclear. This study examined the genetic contribution to this comorbidity, and the sex-specific causal association between higher genetic risk of three psychiatric disorders [major depression (MD), schizophrenia and bipolar disorder] and three CVDs [coronary artery disease (CAD), heart failure (HF) and atrial fibrillation].

**Methods** Genetic correlation and polygenic overlap between psychiatric disorders and CVDs were estimated using summary statistics from genome-wide association studies. Association between polygenic scores (PGS) for psychiatric disorders and CVD outcomes was assessed in 345,340 European-ancestry individuals (UK Biobank) using logistic regression.

**Results** Though substantial genetic overlap was identified between psychiatric disorders and CVDs, a significant genetic correlation (rg ranged between 0.11 and 0.25, p < 7.8E-06) was only observed between MD and all three CVDs. A 1-standard-deviation increase in PGSMD was associated with a 1.11- (95% CI: 1.06 - 1.16; p = 4.9E-06) and 1.09-fold (95% CI: 1.04 - 1.14; p = 0.00041) higher risk of incident CAD and HF in females, but only with a 1.04-fold (95% CI: 1.02 - 1.07; p = 0.0019) increased risk of incident CAD in males. The PGSMD was associated with CAD and HF even in the absence of any psychiatric disorder diagnosis.

**Conclusions** Our results suggest that the increased CVD risk in individuals with schizophrenia and bipolar disorder is likely a consequence of non-genetic factors (such as lifetime hardships and medications). Notably, shared genetic risk is an important contributor to CVD comorbidity in individuals with a diagnosis or a higher risk of MD, and more so in females than in males.

Keywords
*   Psychiatric disorder
*   cardiovascular disease comorbidity
*   depression
*   polygenic risk score
*   sex-specific risks

## Introduction

The idea that mental health is linked to cardiovascular health has long been recognised: William Harvey wrote in 1628 “every affection of the mind that is attended with either pain or pleasure, hope or fear, is the cause of an agitation whose influence extends to the heart” 1. Modern epidemiological studies find that collectively, individuals with psychiatric disorders, namely major depressive disorder (MDD), schizophrenia (SCZ) and bipolar disorder (BD), have a ∼50% higher odds of developing cardiovascular diseases (CVD) 2. In particular, there is a doubling of the odds of developing coronary heart disease amongst individuals with depression, and those with SCZ are at a higher risk of coronary heart disease and congestive heart failure (HF) 2. Furthermore, prospective studies have reported that depression status predicts incidence of cardiovascular heart disease, and cardiac mortality in cohorts with and without clinically diagnosed cardiovascular heart disease at enrolment 3.

There is marked sex difference in psychiatric disorders and CVDs, in both prevalence and clinical presentation. Depression is twice as prevalent in females than males (lifetime prevalence of 26.1% versus 14.7%) 4, 5, with women showing augmented symptom severity 6. While the prevalence of SCZ does not differ between genders, men with SCZ show earlier age of onset and worse premorbid functioning 7. Men also show an earlier onset for BD, although the prevalence is reported to be equal between sexes for bipolar I disorder but higher in women for bipolar II disorder, which is characterised by depressive episodes 8. Furthermore, although it is traditionally considered to be a male disease (lifetime risk of 60.3% in men versus 55.6% in women at an index age of 45 years) 9, CVD is a leading cause of female deaths, and women have higher CVD-related mortality rates relative to men (hazard ratio of 1.07 for death amongst women aged ≤ 55 years, but no significant association amongst men aged ≤ 55 years) 10, 11. Furthermore, although CVD mortality has declined in recent years, the decline has stagnated in young women, a time period that coincides with increased rates of depression 10. Such evidence suggests that the comorbidity of depression and CVDs is likely to have more pronounced effects in females.

The increased CVD risk amongst individuals with psychiatric disorders may be attributed to a combination of non-genetic and genetic factors. Psychiatric disorders are associated with stress (both as cause and consequence) which can lead to diverse biological alterations that predispose individuals to an increased CVD risk, such as changes in immune activities, lipid profiles, heart rate, and blood pressure 12–14. Behaviourally, psychiatric disorders and chronic stress may trigger adoption of unhealthy habits, such as smoking and social isolation, and further exacerbate the cardiac risk-increasing biological impacts 12, 13. Moreover, the use of some commonly prescribed psychiatric medications is associated with increased blood pressure, obesity, stroke, and atrial fibrillation (AF) 15, 16. To what extent comorbidity between different psychiatric disorders and different CVDs is attributable to shared non-genetic or genetic factors is an outstanding question, and one that will better inform CVD prevention strategies.

Large genome-wide association studies (GWAS) have been very successful at identifying genetic loci associated with disease risk. Novel statistical methods applied to GWAS data are facilitating our understanding of shared biology between two diseases, as well as in making causal inferences, overcoming some of the caveats of observational studies of comorbidity. Previous genetic analyses, for example, have identified a causal effect of depression on coronary artery disease (CAD) 17, 18. However, analyses across different psychiatric disorders and CVDs, and investigation of sex differences using such genetic approaches have not yet been done. In this study, we used GWAS summary data to investigate the degree of genetic overlap between three psychiatric disorders (BD, major depression (MD) and SCZ) and three CVDs (AF, CAD and HF). We also used polygenic scores (PGS) derived from GWAS summary statistics (which estimate an individual’s genetic liability to a disease) to examine the causal association in an independent sample, between higher genetic risk of psychiatric disorders and incident CVD in males and females separately.

## Methods

Publicly available datasets analysed in this study are referenced in Methods or Supplementary Materials. For more information on data analysed and scripts please see the Major Resources Table in the Supplemental Materials.

### UK Biobank cohort summary

The UK Biobank cohort consists of over 500,000 participants (aged 40-69 years at baseline) from the UK, who were recruited between 2006 and 2010 19. Diverse phenotypic and genomic data were collected from participants, such as clinical diagnosis, physical and biochemical measurements, and answers from questionnaires. Consents were collected from all participants. Current study was conducted using the UK Biobank Resource under Application Number 12505. This research is covered by the University of Queensland Human Research Ethics Committee approval (HREC number 2020/HE002938). After quality control (details described in Supplementary Methods), a total of 345,340 unrelated (genetic relatedness < 0.05) individuals of European ancestry were subjected to analysis in this study (ancestry calling is described in Yengo et al. 20).

### BioVU cohort summary

Vanderbilt University Medical Center (VUMC) is a tertiary care centre that provides inpatient and outpatient care in middle Tennessee, US. The VUMC electronic health record (EHR) system, established in 1990, includes data on billing codes from the International Classification of Diseases, 9th and 10th editions (ICD-9 and ICD-10), Current Procedural Terminology (CPT) codes, laboratory values, reports, and clinical documentation. In 2007, VUMC launched a biobank, BioVU, which links a patient’s DNA sample to their EHR. The BioVU consent form is provided to patients in the outpatient clinic environments at VUMC. The VUMC Institutional Review Board oversees BioVU and approved this project (IRB#172020). At VUMC, all medical record data were extracted from the EHRs of 72,634 individuals of European ancestry.

### GWAS summary statistics

Association summary statistics derived from European-ancestry or predominantly European-ancestry cohorts were retrieved for three CVDs (AF, CAD and HF), and three psychiatric disorders (BD, MD and SCZ). Summary statistics used for genetic correlation estimation are summarised in Supplementary Table 1. Summary statistics used for PGS single nucleotide polymorphism (SNP) weight generation for prediction in the UK Biobank and BioVU cohorts are summarised in Supplementary Table 2 and Supplementary Table 3 respectively.

### Genetic correlation

Using GWAS summary statistics derived from European-ancestry cohorts (Supplementary Table 1) and linkage disequilibrium (LD) scores from the European 1000 Genomes cohort, pairwise genetic correlations between CVDs and psychiatric disorders were estimated with LD score regression (LDSC) (version 1.0.1) 21. A p < 0.0056 indicated statistical significance (after multiple testing correction for three CVDs and three psychiatric disorders).

### Polygenic overlap analysis

MiXeR is used to estimate polygenic overlap, reflected by fraction of shared genetic variants between two traits, irrespective of genetic correlation 22. Detailed information on MiXeR, and polygenic overlap with and without genetic correlation, is presented in Supplementary Methods and Supplementary Figure 1. We performed MiXeR (version 1.2.0) analysis to estimate the degree of polygenic overlap between CVDs and psychiatric disorders under default settings, using LD references derived from 1000 Genomes Phase 3 22. We excluded the major histocompatibility complex (MHC) (chromosome 6:26000000-34000000) region from analysis due to its complex LD structure.

### UK Biobank phenotype ascertainment

The inclusion and exclusion criteria for the CVDs, psychiatric disorders, and individuals with no psychiatric disorder diagnosis or psychiatric medications are shown in Supplementary Table 4, Supplementary Table 5 and Supplementary Table 6 respectively. Smoking status, defined as ever having smoked, was collated from UK Biobank data-field 20116 (“Smoking status”). The average body mass index (BMI) of individuals was calculated (data-field 21001). The recruitment ages (in years) of individuals were calculated using “Date of attending assessment centre” (data-field 53) and “Year of birth” (data-field 34). Summary of age, smoking status and BMI are presented in Supplementary Table 7 and Supplementary Table 8.

### UK Biobank incident CVD ascertainment

For each CVD, incident cases were defined as individuals who received all disease-associated ICD-10 diagnosis after their recruitment into the UK Biobank (data-field 41280), individuals who self-reported disease-related illness (data-field 20002) after their first assessment, or had a disease-associated ICD-10 code reported as primary or secondary death causes.

### BioVU phenotype ascertainment

For the BioVU cohort, the inclusion and exclusion criteria for CVDs and psychiatric disorders are shown in Supplementary Table 9. Smoking status was ascertained by “tobacco use disorder” (Phecode 318), which is a strong proxy for smoking. The median BMI and median age of individuals were selected from assessments taken at chronological visits, available in the medical records at VUMC. Summary of phenotypes is presented in Supplementary Table 10.

### Polygenic score generation

PGSs are calculated to estimate the lifetime genetic liability of an individual to a trait. Briefly, we define a PGS of an individual, *j*, as a weighted sum of SNP allele counts: ![Graphic][1]</img>, where *m* is the number of SNPs included in the predictor, *bi* is the per-allele weight (based on the estimated effect of the allele on trait value) for the SNP, *xij* is a count of the number (0, 1, or 2) of trait-associated alleles of SNP *i* in individual *j*.

For prediction into the UK Biobank cohort, we estimated the per-allele weights for SNPs with SBayesR using default settings unless otherwise stated (--exclude-mhc; --chain-length 50000; -- burn-in 10000; --no-mcmc-bin) 23. SBayesR implements a Bayesian approach, uses an LD reference (banded LD matrix retrieved from the SBayesR website) to select variants and estimate per-allele weights, does not require a training dataset for identifying trait-associated SNPs and estimating per-allele weights for PGS calculation, and is computationally efficient 23, 24.

GWAS summary statistics used for per-allele weight prediction for the UK Biobank cohort were derived from cohorts that did not include any UK Biobank data (Supplementary Table 2). For each disease, we applied PLINK1.9 25 to generate PGS for each unrelated individual of European ancestry, to avoid biases introduced by population stratification and cryptic relatedness.

PGSs in BioVU were generated using Polygenic Risk Score–Continuous Shrinkage (PRS-CS) 26 (which has been shown to have similar performance to SBayesR 24), with GWAS summary statistics derived from cohorts that did not include the BioVU cohort (Supplementary Table 3). The LD reference panel was constructed from 503 European samples in the 1000 Genomes Project Phase 3.41.

The PGS distribution was scaled to a mean of zero and standard deviation (SD) of 1; therefore, effect size estimates in subsequent logistic regressions were interpreted as changes in outcome per 1-SD increase in PGS. To validate the accuracy of PGS, logistic regression was used to examine the association with disease prevalent cases (hereby defined as incident cases and existing cases at first assessment) in the prediction cohort (UK Biobank or BioVU). PGS performance was evaluated using the Nagelkerke pseudo R2 [R package rcompanion 27 (version 2.4.13)] and area under the receiver operator characteristic curve (AUC) [R package pROC 28 (version 1.18.0)]. The Nagelkerke pseudo R2 compares the null model (containing only the covariates as predictors) and the full model (with PGS added as a predictor), and is a relative measure of how well the full model explains the data compared to the null model. The AUC can be interpreted as a probability of a case having a higher PGS than a control.

### Logistic regression of PGS and incident CVD in the UK Biobank cohort

For each pair of CVD and psychiatric disorder, logistic regression was used to estimate the association between incident CVD and psychiatric disorder PGS, adjusting for sex, genotyping array, age, BMI, smoking status and 20 genetic principal components (PC). For the association analysis in the overall cohort, a p < 0.0056 indicated statistical significance (after multiple testing correction for three psychiatric disorders and three CVDs).

In sex-stratified analysis, logistic regression in males and females were conducted separately, and the logistic regression coefficients [log(OR)] for the PGS were compared between sexes using a Wald test (two-sided p < 0.05 indicated statistical significance). Sensitivity analysis was performed with “death cases” (individuals with CVD as underlying cause of death but no record of CVD from primary care, hospital records or self-reported questionnaires) removed. Due to the large polygenic overlap between CVD and psychiatric disorders, to assess if the psychiatric PGSs were associated with CVD risks independently of the CVD PGS, we performed additional sensitivity analysis where the PGS for the relevant CVD was fitted as a covariate. A p < 0.0083 indicated statistical significance (after multiple testing correction for three CVDs and two sexes).

Similarly, the overall and sex-stratified associations were estimated for CVD risks in UK Biobank participants who did not meet the criteria for the three psychiatric disorders, had no ICD diagnosis of multiple personality disorders or neurosis, had no self-reported history of mood disorders, and were not on any antidepressants or anti-psychotic medications (detailed phenotype information in Supplementary Table 6). Due to the reduced sample size, we investigated the associations for both incident and prevalent CVDs in this cohort. A p < 0.0083 indicated statistical significance.

### Logistic regression in BioVU

Findings from UK Biobank were independently replicated in the BioVU cohort at VUMC. Logistic regression was used to estimate the overall and sex-stratified association between prevalent CVD and psychiatric disorder PGS, adjusting for sex, median age, median BMI, smoking status and 10 genetic PCs. The overall and sex-stratified associations were also estimated for MD genetic risk amongst individuals that did not meet the criteria for the three psychiatric disorders, and had no ICD diagnosis of neurosis and psychosis.

### Comparing incident CVD risks in top and bottom MD risk groups

Using the UK Biobank cohort, we performed logistic regression to compare the risk of incident CVD in individuals who were in the top and bottom quartiles (25%) or deciles (10%) of PGSMD, adjusting for sex, genotyping array, age, BMI, smoking status and 20 genetic PCs. Sex-stratified models were also evaluated for the top and bottom MD risk groups. A p < 0.0083 indicated statistical significance.

## Results

### MD exhibits significant genetic correlation and polygenic overlap with CVDs

Two traits are said to have polygenic overlap if they share genetic variants that confer non-zero effects on both traits. A non-zero genetic correlation reflects specific scenarios of polygenic overlap, where the shared genetic variants confer overall concordant (positive genetic correlation) or discordant (negative genetic correlation) effect directions on the two traits. Genetic correlation ranges between -1 and 1, where 1 means all genetic variants that increase the value of one trait, also increase the value of the second trait, while a genetic correlation of -1 indicates that all genetic variants that increase the value of one trait, decrease the value of the second trait. A lack of genetic correlation may be reported even in the presence of polygenic overlap if two traits share genetic variants, but the genetic variants exhibit a mixture of effect directions and thus showing minimal genetic correlation (further detail in Supplementary Methods and Supplementary Figure 1).

All six psychiatric disorders and CVDs were shown to be highly polygenic by MiXeR analysis, with hundreds of trait-associated variants estimated for each disease (Figure 1). After multiple testing correction, there was significant positive genetic correlation between MD and all three CVDs, with the strongest correlation observed for HF [rg = 0.25 (95% confidence interval (CI): 0.18 - 0.32; p = 1.3E-13)] and CAD [rg = 0.21 (95% CI: 0.17 - 0.25; p = 3E-23)], and lower but significant correlation with AF [rg = 0.11 (95% CI: 0.063 - 0.16; p = 7.8E-6)], which agrees with previous findings 17, 29 (Figure 1A). Corroborating the genetic correlation results, MiXeR analysis demonstrated that a large fraction of CVD variants was shared with MD (Figure 1B). Up to 91% of the genetic variants for HF were also associated with MD, and 76.8% of these shared variants conferred concordant effect directions on both traits, explaining the high genetic correlation. 68.8% and 60% of CAD- and AF-associated variants, respectively, were shared with MD, and up to 83% and 92% of shared variants conferred concordant effect directions. After multiple testing correction, no significant genetic correlation with CVDs was observed for SCZ and BD (Figure 1A). Though some polygenic overlap was detected for SCZ and BD with all three CVDs, the shared genetic variants comprised of ∼50% with concordant effect directions on both diseases and ∼50% with discordant effect directions (Figure 1B), which is consistent with the lack of genetic correlation.

![Figure 1.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F1.medium.gif)

[Figure 1.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F1)

Figure 1. 
MD exhibits significant genetic correlation and polygenic overlap with CVDs. (**A**) Genetic correlations (rg) between the three psychiatric disorders and three CVDs analysed in this study. The dotted line indicates a lack of genetic correlation (rg = 0). Error bars indicate 95% confidence interval for the rg estimates, and the corresponding p-values are shown. (**B**) Polygenic overlaps between psychiatric disorders and CVDs. The Venn diagrams show the number (thousands) of trait-associated variants specific to each psychiatric disorder (blue), specific to each CVD (yellow), and shared by a pair of psychiatric disorder and CVD (overlap), as estimated by MiXeR. The pie charts show the percentages of shared variants (corresponding to the overlapped region in the relevant Venn diagram) that confer concordant (grey) and discordant (black) directions of effects on both traits. AF, atrial fibrillation; BD, bipolar disorder; CAD, coronary artery disease; CVD, cardiovascular disease; HF, heart failure; rg, genetic correlation; MD, major depression; SCZ, schizophrenia.

### PGS validation in UK Biobank cohort

We generated trait-specific PGS for 345,340 unrelated, European-ancestry individuals from the UK Biobank, as a proxy for their genetically estimated risk of developing MD, SCZ or BD. Supporting the validity of PGSs as predictors of disease risk, all PGSs showed a significant odds ratio (OR) with the corresponding prevalent disease cases (Supplementary Table 11). The ORMD was 1.3 (95% CI: 1.27 - 1.32; p < 2E-16) per SD increase in PGSMD, and the corresponding ORs for PGSBD and PGSSCZ were 2.04 (95% CI: 1.92 - 2.18; p < 2E-16) and 2.00 (95% CI: 1.86 - 2.15; p < 2E-16) respectively. Similarly, ORs for PGS of CVDs ranged between 1.19 and 1.63 (p < 2E-16). We used AUC, which can be interpreted as probability that that a case has a higher PGS than a control, to measure the prediction accuracy of our PGS models. The AUC for the models containing PGSs and covariates ranged between 0.65 and 0.8, indicating agreement between the predicted disease liability and observed prevalent cases (Supplementary Table 11). Overall, the addition of PGSs improved the prediction performance of the logistic regression models, as indicated by positive Nagelkerke R2 values (Supplementary Table 11). For the psychiatric disorders PGSs generated for the BioVU cohort, ORs ranged between 1.14 and 1.55 (Supplementary Table 12).

### Genetic risk of MD is correlated with increased incident CVD

We estimated the association between genetic liability to psychiatric disorders and incident CVD (Figure 2 and Supplementary Table 13). Using a multiple testing-corrected significance threshold (p < 0.0056), PGSMD was significantly associated with all three CVDs [ORAF = 1.03 (95% CI: 1.01 - 1.05; p = 0.0034); ORCAD = 1.06 (95% CI: 1.04 - 1.09; p = 3.7E-07); ORHF = 1.04 (95% CI: 1.01 - 1.07; p = 0.0034)]. No significant association with CVD risks was observed for SCZ and BD. Similar to the UK Biobank cohort, we observed a positive association between PGSMD and prevalent CAD in the BioVU cohort, supported by nominal significance (Supplementary Figure 2). However, we also observed a nominally significant, negative association between PGSBD and prevalent CAD in the BioVU cohort. Similar trends were observed in the UK Biobank cohort with prevalent cases (Supplementary Figure 2), and was thus likely due to survival bias, as this association attenuated completely with incident cases in UK Biobank (Figure 2).

![Figure 2.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F2.medium.gif)

[Figure 2.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F2)

Figure 2. 
Association between psychiatric disorder PGSs and incident CVDs in the UK Biobank cohort. Associations were estimated with logistic regressions. X-axis shows the OR per SD increase in PGS, with p-values labelled, and error bars indicate 95% confidence intervals. Dark grey line indicates OR of 1. AF, atrial fibrillation; BD, bipolar disorder; CAD, coronary artery disease; HF, heart failure; MD, major depression; OR, odds ratio; SCZ, schizophrenia.

We thus focussed our analysis on PGSMD and performed sex-stratified logistic regression with incident CVD cases (Figure 3A and Supplementary Table 14). After multiple testing correction, we found that PGSMD was significantly associated with increased CAD risk in both sexes [ORCAD-female = 1.11 (95% CI: 1.06 - 1.16; p = 4.9E-06) and ORCAD-male = 1.04 (95% CI: 1.02 - 1.07; p = 0.0019)], with risk significantly higher in females compared to males (Wald test p < 0.05). PGSMD was significantly associated with HF risk in females only [ORHF-female = 1.09 (95% CI: 1.04 - 1.14; p = 0.00041)] (female versus male Wald test p < 0.05). We did not observe statistically significant association between PGSMD and risk of AF in the sex-stratified analysis. Similar results were observed in sensitivity analyses, where “death cases” were removed, or where relevant CVD PGS was added to the model as a covariate, indicating that the significant association between psychiatric disorder PGS and incident CVDs was independent of the CVD genetic risk (Supplementary Figure 3). We observed similar trends between PGSMD and CAD risk in the overall and female-specific BioVU cohort, supported by nominal significance, but we did not find a positive association with HF risk in the BioVU cohort, possibly due to the reduced sample sizes (Figure 3B and Supplementary Table 15).

![Figure 3.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F3.medium.gif)

[Figure 3.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F3)

Figure 3. 
Sex-stratified association between MD PGS and incident CVDs. (**A**) Sex-stratified associations between PGSMD and incident CVDs (red - female; blue - male) in the UK Biobank cohort. Asterisk (*) indicates a statistically significant difference (two-sided Wald test p < 0.05) in the log(OR) values between females and males. (**B**) Overall and sex-stratified associations between PGSMD and incident CVDs in the BioVU cohort. Associations were estimated with logistic regressions. X-axis shows the OR per SD increase in PGS, with p-values labelled, and error bars indicate 95% confidence intervals. Dark grey line indicates OR of 1. AF, atrial fibrillation; CAD, coronary artery disease; HF, heart failure; OR, odds ratio.

### Individuals in top and bottom MD genetic risk groups differ in CVD risks

Using a logistic regression model, we compared the risk for developing CVDs between individuals who were in the top and bottom quartile (25%) of PGSMD, which represented the groups with the highest and lowest genetic risk of developing MD respectively (Figure 4 and Supplementary Table 16). Compared to individuals in the bottom PGSMD quartile, sex-stratified analysis reported a 32% higher CAD risk in females [ORCAD-female = 1.32 (95% CI: 1.16 - 1.49; p = 1.6E-05)], which was significantly higher (Wald test p < 0.05) than the 12% increased risk in males [ORCAD-male = 1.12 (95% CI: 1.03 - 1.21; p = 0.0049)]. Similarly, compared to the bottom quartile, females in the top 25% PGSMD risk group also had a significantly higher risk of developing HF [ORHF-female = 1.22 (95% CI: 1.07 - 1.39; p = 0.0034)], but not AF. On the contrary, the risk of incident HF amongst males in the high MD risk group did not differ from the low-risk group, while the risk of incident AF was higher in the high MD risk group, supported by nominal significance [ORAF-male = 1.08 (95% CI: 1.01 - 1.15; p = 0.032)]. Similar results were observed for individuals with extremely high genetic risks of MD (the top 10% of PGSMD) (Supplementary Table 16).

![Figure 4.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F4.medium.gif)

[Figure 4.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F4)

Figure 4. 
Association with incident CVDs for individuals in the top PGSMD quartile (25%), compared to the bottom quartile. The ORs of CVDs are shown for the overall (black) and sex-stratified cohorts (red - female; blue - male) of UK Biobank. Associations were estimated with logistic regressions. X-axis shows the OR in the high-risk group versus the low-risk group, with p-values labelled, and error bars indicate 95% confidence intervals. Asterisk (*) indicates a statistically significant difference (two-sided Wald test p < 0.05) in the log(OR) values between females and males. Dark grey line indicates OR of 1. AF, atrial fibrillation; CAD, coronary artery disease; HF, heart failure; OR, odds ratio.

### Genetic risk of MD is associated with increased CVD risk in individuals with no psychiatric diagnosis and medications

We also investigated the association between PGSMD and CVD statuses (both prevalent and incident) in individuals who had no psychiatric disorder diagnosis and were not on any psychiatric medication. We observed a positive association between PGSMD and prevalent CAD in both sexes [ORCAD-female = 1.06 (95% CI: 1.03 - 1.1; p = 0.00099) and ORCAD-male = 1.04 (95% CI: 1.02 - 1.06; p = 1.8E-05)] (Figure 5 and Supplementary Table 17). Similar results were observed in the sensitivity analysis where the PGSCAD was added as a covariate, indicating that the observed associations were independent of the genetic overlap between MD and CAD (Supplementary Figure 4). We did not find a significant association between PGSMD and prevalent AF in the sex-stratified cohorts, except for a nominally significant positive association in the male cohort [ORAF-male = 1.03 (95% CI: 1.01 - 1.05; p = 0.011)] (Figure 5 and Supplementary Table 17). No significant associations were observed of PGSMD with HF risk, except for a nominally significant positive association in the female cohort [ORHF-female = 1.06 (95% CI: 1.01 - 1.12; p = 0.021)]. To avoid survival bias, we repeated the association analysis for incident CVD cases amongst individuals without psychiatric diagnoses and medications. Similar trends to the analysis of prevalent CVDs were observed between PGSMD and incident CVDs in this group, but no sex-stratified association reached statistical significance, possibly due to reduced sample sizes (as indicated by similar effect estimates but larger standard errors) (Figure 5 and Supplementary Table 17). Amongst BioVU participants with no psychiatric disorder diagnosis, we did not observe any significant association of PGSMD with prevalent CAD or HF, although we observed a negative association with prevalent AF for both sexes (Supplementary Figure 5). However, analysis of the BioVU cohort relied only on prevalent data, which could potentially have led to survival bias that confounded the interpretation of results.

![Figure 5.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F5.medium.gif)

[Figure 5.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F5)

Figure 5. 
Association between PGSMD and CVDs amongst UK Biobank individuals with no diagnosis of psychiatric disorders and psychiatric medications. The associations of PGSMD with incident (left) or prevalent (right) CVDs are shown for the overall (black) and sex-stratified cohorts (red - female; blue - male). Associations were estimated with logistic regressions. X-axis shows the OR per SD increase in PGS, with p-values labelled, and error bars indicate 95% confidence intervals. Dark grey line indicates OR of 1. AF, atrial fibrillation; CAD, coronary artery disease; HF, heart failure; OR, odds ratio.

## Discussion

In this study, we find a significant genetic correlation between all three CVDs with MD, but not with SCZ or BD, suggesting that cardiovascular comorbidity within the latter is most likely a result of non-genetic factors (such as lack of physical activity, adverse dietary habits, substance use and psychiatric medications). However, even in the absence of genetic correlation, we find substantial overlap in the genetic variants associated with each psychiatric disorder and CVDs. Knowledge of these shared variants may be important for informing treatment in comorbid individuals, especially when pharmacological targeting of a shared risk locus has opposing effects on comorbid traits. For example, genetic variants in the *ACE* gene have shown to have opposing effects on CVD risk and SCZ risk 30; therefore, inhibition of ACE for lowering blood pressure could potentially have adverse effects on individuals with SCZ 31.

After accounting for age and conventional risk factors, including BMI and smoking, we showed that individuals with a higher genetic risk for MD had significantly higher risk for AF, CAD and HF. This association was independent of genetic risk of CVD, and the association with CAD and HF was of higher magnitude in females relative to males, particularly for HF, where we saw increased risk in females but no significant association in males. Interestingly, our findings showed that this association was observed even in the absence of a diagnosis of any psychiatric disorder and psychiatric medication. This warrants further studies that evaluate the importance of an individual’s risk of depression in addition to any diagnosis in cardiovascular risk assessment.

Epidemiological studies have reported notable sex differences in the prevalence and clinical manifestation of psychiatric disorders and CVDs, with women particularly prone to depression and CVD-related mortality. Similarly, previous evidence has found depression to be a risk factor for HF amongst women but not in men 32. However, observational studies do not provide evidence for a causal association. Previous genetic analyses that utilise GWAS summary data to assess causality have shown that an increased genetic risk for MD is associated with an increased CAD and HF risk 18. However, these studies have not investigated sex-specific associations, except for a previously reported genetic association between MD PGS and CAD observed in females but not males in a (smaller) BioVU cohort 33. Corroborating previous findings, we found that the genetic predisposition to MD was associated with a greater increase in CAD and HF risk amongst females. It has been suggested that MD and CVDs may represent manifestation of a common pathophysiological dysfunction 34. The pathophysiology underlying the female preponderance remains unclear, and possible mechanisms include hormonal dysregulation and pro-inflammatory response. Specific subtypes of depression, such as postpartum depression and postmenopausal depression, suggests the involvement of hormone fluctuations in depression amongst females 35. Interestingly, Takotsubo cardiomyopathy, often triggered by emotional or physical stress, is the most common in postmenopausal women (90% of cases) 36. At the same time, elevated inflammatory biomarkers, such as interleukin-6 and tumour necrosis factor alpha, are reportedly higher in depressed women relative to depressed men, suggesting a sex-differential inflammatory response to MD 35. These pro-inflammatory cytokines are linked to incident CVD and poorer cardiovascular outcomes 37, 38. Further studies are required to understand the heightened risk of CVDs amongst females with depression.

CVD risk calculators failing to incorporate psychiatric disorders as a predictor are estimated to underestimate CVD risk by 30% and 60% in men and women respectively 39. Currently, mental illnesses are included in the QRISK3 prediction algorithm (UK), and are recognised as a risk factor in New Zealand, where individuals with severe mental illness are advised to undergo earlier and more frequent risk assessments 40, 41. Furthermore, the American Heart Association (AHA) has recommended incorporation of depression as a risk factor of adverse prognosis amongst patients with acute coronary syndrome 42; however, our findings highlight the importance of investigating the predictive value of depression diagnosis and depression risk (such as family history or polygenic risk) in the preventative assessment of CVD risk, particularly amongst females.

Our study has several strengths. Our PGSs were derived from GWAS summary statistics with the largest sample sizes where possible, enhancing the statistical power in detecting genetic variants with small but significant effects. Previous Mendelian randomisation (MR) studies have used GWAS summary statistics derived from cohorts with overlapping samples 18, 29, 43, which could lead to biased estimates of genetic association. To avoid any bias, our psychiatric PGSs were derived from discovery cohorts with no sample overlap with the prediction cohort (UK Biobank or BioVU). Furthermore, most of these genetic studies have used summary data from GWAS conducted in a combined sample of both males and females, and therefore were unable to investigate sex-specific association across different diseases. These studies are also unable to dissociate effects of behavioural or medication use as a consequence of depression diagnosis, which our study was able to show through analysis of individuals with no diagnosis of psychiatric disorders and psychiatric medications.

However, limitations of our study need to be acknowledged. We derived PGSs from sex-combined summary statistics of GWAS, which might fail to capture genetic variants that confer sex-differential associations. The between-sex genetic correlations for BD, MD and SCZ are reported to range between 0.86 and 1, indicating moderate to subtle sex differences in the genetic architecture of these psychiatric disorders 44. Nevertheless, PGSs derived from sex-stratified GWAS studies, which become increasingly available for neuropsychiatric and cardiovascular traits 44, 45, will provide additional information on the sex-differential genetic factors in diseases. Moreover, due to the lower number of CVD cases amongst females in UK Biobank, our analysis in the female cohort had lower power. Lastly, findings from the BioVU cohort may be confounded by medication status of participants, and potential survival bias as incident data are not available.

In conclusion, we analysed 345,340 unrelated, European-ancestry individuals from the UK Biobank cohort, and showed that a higher genetic risk of MD increases the risk of CVDs, with higher risks observed for females. To our best knowledge, this is the first study that explored the association between genetic risk of three different psychiatric disorders and risks of CAD, HF and AF, with a particular focus on sex-specific differences. Given that these CVDs have both shared and unique underlying disease biology, and that their polygenic overlap with MD varies, with the highest overlap observed for HF, future research into understanding whether an increased risk of depression increases risk of CAD, AF and HF through the same or unique biological pathways may be important for informing prevention strategies. Our study highlights the necessity of investigating the predictive value of depression diagnosis and risk, in addition to conventional cardiovascular risk factors, in preventative CVD risk assessments, particularly for females. In addition, individuals with depression diagnosis or family histories constitute a high-risk cohort for CVD and merit close monitoring on their risk of CVDs.

## Novelty and Significance

### What is known?

*   Epidemiological evidence reports that individuals with psychiatric disorders are at a highest risk of developing CVDs, but to what degree this cardiovascular comorbidity is attributable to environmental or genetic factors remains unclear

*   There is marked sex differences in the prevalence and clinical presentation of psychiatric disorders, with women reported to show a highest risk of adverse cardiovascular outcomes after a diagnosis of depression.

### What new information does this article contribute?

*   Though we identify substantial overlap in the genetic loci associated with both psychiatric disorders and CVDs, significant genetic correlation between CVDs and depression suggests shared pathophysiology, while the lack of genetic correlation with schizophrenia and bipolar disorder suggests that non-genetic factors (e.g. lifetime hardships that may lead to an adverse lifestyle, as well as psychiatric medication) may play a larger role in CVD comorbidity for these disorders.

*   A higher risk of CVD outcomes was present in individuals with a high genetic risk of depression even in the absence of a depression diagnosis, and this risk was higher in females.

*   Notably, a higher genetic risk of depression was associated with an increased risk of CAD and HF in females, and increased risk of CAD but not HF in males.

Our study uses polygenic scores, an estimate of the genetic risk of psychiatric disorders, to investigate the causal association between psychiatric disorders and risk of CVDs, with a particular focus on sex differences. Identification of the shared genetic loci between CVDs and psychiatric disorders will be important for informing treatment, especially where a pharmacological agent targeting a shared factor may have opposing effects on the comorbid diseases. QRISK3 is the only tool that incorporates diagnosis of severe mental illnesses (including major depression) in primary prevention of CVD; however, our study shows that even in the absence of any psychiatric disorder diagnosis, a high genetic risk of depression may be an important consideration, especially in women. Our study highlights the need for further evaluating both depression risk and diagnosis for primary prevention of CVD.

## Data Availability

All data produced in the present work are contained in the manuscript. Publicly available GWAS summary statistics analysed in this study are properly referenced in the manuscript. The full GWAS summary statistics for the 23andMe discovery dataset will be made available through 23andMe to qualified researchers under an agreement with 23andMe that protects the privacy of the 23andMe participants. Please visit [https://research.23andme.com/collaborate/#dataset-access/](https://research.23andme.com/collaborate/#dataset-access/) for more information and to apply to access the data.

[https://github.com/CNSGenomics/CVD\_psych\_PRS\_Jiang](https://github.com/CNSGenomics/CVD_psych_PRS_Jiang) 

## Sources of Funding

JCJ is supported by a National Health and Medical Research Council (NHMRC) IDEAS grant (2000637). KS is supported by the American Heart Association Predoctoral Fellowship (AHA827137). LKD is supported by funding from the National Institute of Mental Health (NIMH) (R56MH120736); and the National Human Genome Research Institute (NHGRI) and the Office of Research on Women’s Health (R01HG011405). NRW is supported by NHMRC Program Grant (1113400); and NHMRC Investigator Grant (1173790). SS is supported by funding from the NHMRC Program Grant (1113400); NHMRC Early Career Fellowship (APP1142495); and National Heart Foundation Future Leader Fellowship (105638).

## Disclosures

None.

## Supplementary Material

Supplementary Methods

Supplementary Figure 1

Supplementary Figure 2

Supplementary Figure 3

Supplementary Figure 4

Supplementary Figure 5

Supplementary Table 1

Supplementary Table 2

Supplementary Table 3

Supplementary Table 4

Supplementary Table 5

Supplementary Table 6

Supplementary Table 7

Supplementary Table 8

Supplementary Table 9

Supplementary Table 10

Supplementary Table 11

Supplementary Table 12

Supplementary Table 13

Supplementary Table 14

Supplementary Table 15

Supplementary Table 16

Supplementary Table 17

Major Resources Table

## Supplementary Methods

### UK Biobank cohort quality control

Quality control was performed on the UK Biobank cohort to remove individuals who were outliers for heterozygosity and missing rate, were excluded from kinship inference, showed evidence of putative sex chromosome aneuploidy, showed mismatching self-reported and genetically inferred sex, or have withdrawn from the study (definitions of exclusion criteria are described in Bycroft et al. 19). Individuals who have missing phenotypes [BMI or smoking status] were also excluded from the analysis. A total of 345,340 individuals from the UK Biobank cohort were included in the final analysis.

### Genetic correlation and polygenic overlap

Two traits are said to have polygenic overlap if they share genetic variants that have non-zero effects on both traits. The genetic correlation between two traits, estimated using LDSC 21, indicates whether the shared genetic variants conferred predominantly the same or opposite effect directions with both traits. For example, a positive genetic correlation between trait 1 and trait 2 indicates that genetic variants that increase the value of trait 1 predominantly also increases the value of trait 2 (i.e. concordant effect directions) (Scenario B in Supplementary Figure 1). In contrast, a negative genetic correlation indicates that genetic variants that increase the value of trait 1 predominantly decrease the value of trait 2 (i.e. discordant effect directions) (Scenario C in Supplementary Figure 1).

The lack of genetic correlation between two traits can be attributed to two scenarios, where the first scenario is that the two traits share no genetic variants (i.e. no polygenic overlap) (Scenario A in Supplementary Figure 1). In this scenario, genetic variants that increase the value of trait 1 does not influence the value of trait 2, and vice versa. The other scenario is that the two traits do share some genetic variants (polygenic overlap); however, these shared genetic variants conferred a mixture of effect directions on the two traits (Scenario D in Supplementary Figure 1). In this scenario, there is a balance of shared genetic variants that confer concordant effect directions on both traits, and variants with discordant effect directions. As a result of the mixture of effect directions cancelling each other out, the two traits exhibit minimal genetic correlation. For example, while schizophrenia and educational attainment show minimal genetic correlation, genetic studies have identified many genetic variants to be jointly associated with both traits, providing an example of two traits with polygenic overlap yet no genetic correlation 22.

### MiXeR

MiXeR is used to estimate the fraction of shared genetic variants, or the degree of polygenic overlap between a pair of traits 22. Rather than identifying the exact locations of shared genetic variants, MiXeR partitions genetic variants into four categories based on their effects on the trait pair, specifically

1.  Genetic variants with no effect on either trait

2.  Genetic variants with effects on trait 1 only, but no effects on trait 2

3.  Genetic variants with effects on trait 2 only, but no effects on trait 1

4.  Genetic variants with effects on both traits (shared genetic variants)

MiXeR fits a bivariate Gaussian mixture model to estimate the fraction of genetic variants that follow each of the four components above. The fraction of shared genetic variants is then multiplied by the total number of genetic variants in the reference panel to give the estimated number of shared genetic variants. In addition, as part of the model, MiXeR also estimates the proportion of shared genetic variants that confer concordant and discordant effect directions on the two traits under consideration.

In the original MiXeR paper by Frei et al. 22, genetic variants are referred to as “causal” for a trait if they confer a non-zero effect on the trait, regardless of the magnitude of the effect sizes. However, as acknowledged by Frei et al. 22, it is preferred that only genetic variants with sufficiently large effect sizes to be referred to as “causal variants”. Therefore, to avoid confusion with genetic variants that reach genome-wide significance, in this paper we refer to genetic variants that are identified by MiXeR to have non-zero effects as “trait-associated variants”.

## Supplementary Figures

![Supplementary Figure 1.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F6.medium.gif)

[Supplementary Figure 1.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F6)

Supplementary Figure 1. 
Schematic illustration of polygenic overlap with and without genetic correlation (rg). Four different scenarios with and without polygenic overlap are illustrated using synthetic data. For a pair of traits under consideration (trait 1 and trait 2), the first row illustrates the direction of effects of 10 example genetic variants (V), where + and - indicates that the variant increases or decreases the value of each trait respectively. ‘NA’ indicates that the variant has zero effect (not trait-associated) on the trait. The second row shows the effect of variants (β) on trait 2 (y-axis) plotted against their effects on trait 1 (x-axis), and each red dot represents a variant from the synthetic data.

![Supplementary Figure 2.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F7.medium.gif)

[Supplementary Figure 2.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F7)

Supplementary Figure 2. 
Association between psychiatric disorder PGSs and prevalent CVDs in the (**A**) BioVU cohort and (**B**) UK Biobank. Associations were estimated with logistic regressions. X-axis shows the OR per SD increase in PGS, with p-values labelled, and error bars indicate 95% confidence intervals. Dark grey line indicates OR of 1.

![Supplementary Figure 3.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F8.medium.gif)

[Supplementary Figure 3.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F8)

Supplementary Figure 3. 
Sensitivity analyses on association between PGSMD and incident CVDs, performed on the UK Biobank cohort. The base analysis (•), with death cases included as incident cases and no CVD PGS included as covariate, is shown for the purpose of comparison. Sensitivity analyses where the death cases were removed from analysis (♦), or where the corresponding CVD PGSs were added as a covariate to the model (▴), were performed for the entire cohort (black) and for sex-stratified cohorts (red - female; blue - male). Associations were estimated with logistic regressions. X-axis shows the OR per SD increase in PGS, with p-values labelled, and error bars indicate 95% confidence intervals. Asterisk (*) indicates a statistically significant difference in the log(OR) values between females and males (two-sided Wald test p < 0.05). Dark grey line indicates OR of 1.

![Supplementary Figure 4.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F9.medium.gif)

[Supplementary Figure 4.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F9)

Supplementary Figure 4. 
Sensitivity analyses on the association between PGSMD and CVDs amongst UK Biobank individuals with no diagnosis of psychiatric disorders and psychiatric medications. Associations were estimated with logistic regressions, where the corresponding CVD PGS was fitted as a covariate. The associations of PGSMD with incident (left) or prevalent (right) CVDs are shown for the overall (black) and sex-stratified cohorts (red - female; blue - male). X-axis shows the OR per SD increase in PGS, with p-values labelled, and error bars indicate 95% confidence intervals. Dark grey line indicates OR of 1.

![Supplementary Figure 5.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/10/2022.10.08.22280805/F10.medium.gif)

[Supplementary Figure 5.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/F10)

Supplementary Figure 5. 
Association between PGSMD and prevalent CVDs amongst individuals of the BioVU cohort who had no psychiatric disorder diagnosis. The associations are shown for the overall (black) and sex-stratified cohorts (red - female; blue - male). X-axis shows the OR per SD increase in PGS, with p-values labelled, and error bars indicate 95% confidence intervals. Dark grey line indicates OR of 1.

## Supplementary Tables

View this table:
[Supplementary Table 1.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T1)

Supplementary Table 1. 
Sources of GWAS summary statistics used in genetic correlation and polygenic overlap analysis

View this table:
[Supplementary Table 2.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T2)

Supplementary Table 2. 
Sources of GWAS summary statistics used in PGS SNP weight estimation for UK Biobank

View this table:
[Supplementary Table 3.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T3)

Supplementary Table 3. 
Sources of GWAS summary statistics used in PGS SNP weight estimation for BioVU

View this table:
[Supplementary Table 4.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T4)

Supplementary Table 4. 
UK Biobank CVD phenotype ascertainment criteria

View this table:
[Supplementary Table 5.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T5)

Supplementary Table 5. 
UK Biobank psychiatric disorder phenotype ascertainment criteria

View this table:
[Supplementary Table 6.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T6)

Supplementary Table 6. 
Exclusion criteria used to identify UK Biobank individuals with no diagnosis of psychiatric disorders and psychiatric edications

View this table:
[Supplementary Table 7.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T7)

Supplementary Table 7. 
Summary of CVD phenotypes in the UK Biobank cohort

View this table:
[Supplementary Table 8.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T8)

Supplementary Table 8. 
Summary of psychiatric disorder phenotypes in the UK Biobank cohort

View this table:
[Supplementary Table 9.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T9)

Supplementary Table 9. 
BioVU phenotype ascertainment criteria

View this table:
[Supplementary Table 10.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T10)

Supplementary Table 10. 
Summary of phenotypes in the BioVU cohort

View this table:
[Supplementary Table 11.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T11)

Supplementary Table 11. 
Prediction accuracy of PGSs generated for the UK Biobank cohort

View this table:
[Supplementary Table 12.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T12)

Supplementary Table 12. 
Prediction accuracy of PGSs generated for the BioVU cohort

View this table:
[Supplementary Table 13.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T13)

Supplementary Table 13. 
Association with incident CVDs by psychiatric disorder PGS in the UK Biobank cohort

View this table:
[Supplementary Table 14.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T14)

Supplementary Table 14. 
Sex-stratified association with incident CVDs by PGSMD in the UK Biobank cohort

View this table:
[Supplementary Table 15.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T15)

Supplementary Table 15. 
Association with prevalent CVDs by PGSMD in the BioVU cohort

View this table:
[Supplementary Table 16.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T16)

Supplementary Table 16. 
Association with incident CVDs in the high-risk PGSMD group relative to the low-risk PGSMD group, estimated using the UK Biobank cohort

View this table:
[Supplementary Table 17.](http://medrxiv.org/content/early/2022/10/10/2022.10.08.22280805/T17)

Supplementary Table 17. 
Overall and sex-stratified associations with incident and prevalent CVDs by PGSMD amongst UK Biobank individuals with no diagnosis of psychiatric disorders and psychiatric medications

## Acknowledgments

We would like to thank the research participants and employees of 23andMe, inc. for making this work possible. The full GWAS summary statistics for the 23andMe discovery dataset will be made available through 23andMe to qualified researchers under an agreement with 23andMe that protects the privacy of the 23andMe participants. Please visit [https://research.23andme.com/collaborate/#dataset-access/](https://research.23andme.com/collaborate/#dataset-access/) for more information and to apply to access the data.

The dataset(s) used for the BioVU analyses described were obtained from Vanderbilt University Medical Center’s BioVU which is supported by numerous sources: institutional funding, private agencies, and federal grants. These include the NIH funded Shared Instrumentation Grant S10RR025141; and CTSA grants UL1TR002243, UL1TR000445, and UL1RR024975. Genomic data are also supported by investigator-led projects that include U01HG004798, R01NS032830, RC2GM092618, P50GM115305, U01HG006378, U19HL065962, R01HD074711; and additional funding sources listed at [https://victr.vumc.org/biovu-funding/](https://victr.vumc.org/biovu-funding/).

This publication (journal article, etc.) was supported by Grant No. R01 HG011405, partially funded by the Office of Research on Women’s Health, Office of the Director, NIH and the National Human Genome Research Institute. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Office of Research on Women’s Health or the National Human Genome Research Institute.

The scripts used in the study are available on: [https://github.com/CNSGenomics/CVD\_psych\_PRS\_Jiang](https://github.com/CNSGenomics/CVD_psych_PRS_Jiang).

## Footnotes

*   * co first authors

## Non-Standard Abbreviations and Acronyms

AF
:   Atrial fibrillation
AHA
:   American Heart Association
AUC
:   Area under the receiver operator characteristic curve
BD
:   Bipolar disorder
BMI
:   Body mass index
CAD
:   Coronary artery disease
CI
:   Confidence interval
CPT
:   Current Procedural Terminology
CVD
:   Cardiovascular disease
EHR
:   Electronic health record
GWAS
:   Genome-wide association study
HF
:   Heart failure
ICD
:   International Classification of Diseases
LD
:   Linkage disequilibrium
LDSC
:   LD score regression
MD
:   Major depression
MDD
:   Major depressive disorder
MHC
:   Major histocompatibility complex
MR
:   Mendelian randomisation
OR
:   Odds ratio
PC
:   Principal component
PGS
:   Polygenic score
PRS-CS
:   Polygenic Risk Score–Continuous Shrinkage
SCZ
:   Schizophrenia
SD
:   Standard deviation
SNP
:   Single nucleotide polymorphism
VUMC
:   Vanderbilt University Medical Center

*   Received October 8, 2022.
*   Revision received October 8, 2022.
*   Accepted October 10, 2022.


*   © 2022, Posted by Cold Spring Harbor Laboratory

The copyright holder for this pre-print is the author. All rights reserved. The material may not be redistributed, re-used or adapted without the author's permission.

## References

1.  1.Graboys TB. Stress and the aching heart. N Engl J Med. 1984;311:594–595.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1056/NEJM198408303110909&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=6749232&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

2.  2.Correll CU, Solmi M, Veronese N, Bortolato B, Rosson S, Santonastaso P, Thapa-Chhetri N, Fornaro M, Gallicchio D, Collantoni E, et al. Prevalence, incidence and mortality from cardiovascular disease in patients with pooled and specific severe mental illness: a large-scale meta-analysis of 3,211,768 patients and 113,383,368 controls. World Psychiatry. 2017;16:163–180.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/wps.20420&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

3.  3.Carney RM, Freedland KE. Depression, mortality, and medical morbidity in patients with coronary heart disease. Biol Psychiatry. 2003;54:241–247.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0006-3223(03)00111-2&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=12893100&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000184451800011&link_type=ISI) 

4.  4.Bromet E, Andrade LH, Hwang I, Sampson NA, Alonso J, de Girolamo G, de Graaf R, Demyttenaere K, Hu C, Iwata N, et al. Cross-national epidemiology of DSM-IV major depressive episode. BMC Med. 2011;9:90.
    
    

5.  5.Hasin DS, Sarvet AL, Meyers JL, Saha TD, Ruan WJ, Stohl M, Grant BF. Epidemiology of adult DSM-5 major depressive disorder and its specifiers in the United States. JAMA Psychiatry. 2018;75:336–346.
    
    

6.  6.Marcus SM, Kerber KB, Rush AJ, Wisniewski SR, Nierenberg A, Balasubramani GK, Ritz L, Kornstein S, Young EA, Trivedi MH. Sex differences in depression symptoms in treatment-seeking adults: confirmatory analyses from the Sequenced Treatment Alternatives to Relieve Depression study. Compr Psychiatry. 2008;49:238–246.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.comppsych.2007.06.012&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=18396182&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000255168600002&link_type=ISI) 

7.  7.Ochoa S, Usall J, Cobo J, Labad X, Kulkarni J. Gender differences in schizophrenia and first-episode psychosis: a comprehensive literature review. Schizophr Res Treatment. 2012;2012:916198.
    
    

8.  8.Vega P, Barbeito S, de Azúa SR, Martínez-Cengotitabengoa M, González–Ortega I, Saenz M, González-Pinto A. Bipolar disorder differences between genders: special considerations for women. Womens Health. 2011;7:663–676.
    
    

9.  9.Wilkins JT, Ning H, Berry J, Zhao L, Dyer AR, Lloyd-Jones DM. Lifetime risk and years lived free of total cardiovascular disease. JAMA. 2012;308:1795–1801.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/jama.2012.14312&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23117780&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000310726500026&link_type=ISI) 

10. 10.Garcia M, Mulvagh SL, Merz CNB, Buring JE, Manson JE. Cardiovascular disease in women. Circ Res. 2016;118:1273–1293.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTA6ImNpcmNyZXNhaGEiO3M6NToicmVzaWQiO3M6MTA6IjExOC84LzEyNzMiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMi8xMC8xMC8yMDIyLjEwLjA4LjIyMjgwODA1LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

11. 11.Shah AJ, Ghasemzadeh N, Zaragoza-Macias E, Patel R, Eapen DJ, Neeland IJ, Pimple PM, Zafari AM, Quyyumi AA, Vaccarino V. Sex and age differences in the association of depression with obstructive coronary artery disease and adverse cardiovascular events. J Am Heart Assoc. 2014;3:e000741.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NToiYWhhb2EiO3M6NToicmVzaWQiO3M6MTE6IjMvMy9lMDAwNzQxIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMTAvMTAvMjAyMi4xMC4wOC4yMjI4MDgwNS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

12. 12.Abed MA, Kloub MI, Moser DK. Anxiety and adverse health outcomes among cardiac patients: a biobehavioral model. J Cardiovasc Nurs. 2014;29.
    
    

13. 13.Larsen BA, Christenfeld NJS. Cardiovascular disease and psychiatric comorbidity: the potential role of perseverative cognition. Cardiovasc Psychiatry Neurol. 2009;2009:791017–791017.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=20029626&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

14. 14.Hildrum B, Romild U, Holmen J. Anxiety and depression lowers blood pressure: 22-year follow-up of the population based HUNT study, Norway. BMC Public Health. 2011;11:601.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/1471-2458-11-601&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=21797992&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

15. 15.Licht CMM, de Geus EJC, Seldenrijk A, van Hout HPJ, Zitman FG, van Dyck R, Penninx BWJH. Depression is associated with decreased blood pressure, but antidepressant use increases the risk for hypertension. Hypertension. 2009;53:631–638.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1161/HYPERTENSIONAHA.108.126698&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

16. 16.Juang H-T, Chen P-C, Chien K-L. Using antidepressants and the risk of stroke recurrence: report from a national representative cohort study. BMC Neurol. 2015;15:86.
    
    

17. 17.Hagenaars SP, Coleman JRI, Choi SW, Gaspar H, Adams MJ, Howard DM, Hodgson K, Traylor M, Air TM, Andlauer TFM, et al. Genetic comorbidity between major depression and cardio-metabolic traits, stratified by age at onset of major depression. Am J Med Genet B Neuropsychiatr Genet. 2020;183:309–330.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/ajmg.b.32807&link_type=DOI) 

18. 18.Lu Y, Wang Z, Georgakis MK, Lin H, Zheng L. Genetic liability to depression and risk of coronary artery disease, myocardial infarction, and other cardiovascular outcomes. J Am Heart Assoc. 2021;10:e017986.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1161/JAHA.120.017986&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

19. 19.Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, Motyer A, Vukcevic D, Delaneau O, O’Connell J, et al. The UK Biobank resource with deep phenotyping and genomic data. Nature. 2018;562:203–209.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41586-018-0579-z&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=30305743&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

20. 20.Yengo L, Sidorenko J, Kemper KE, Zheng Z, Wood AR, Weedon MN, Frayling TM, Hirschhorn J, Yang J, Visscher PM, et al. Meta-analysis of genome-wide association studies for height and body mass index in ∼700000 individuals of European ancestry. Hum Mol Genet. 2018;27:3641–3649.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/hmg/ddy271&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=30124842&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

21. 21.Bulik-Sullivan B, Finucane HK, Anttila V, Gusev A, Day FR, Loh P-R, Duncan L, Perry JRB, Patterson N, Robinson EB, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet. 2015;47:1236–1241.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/ng.3406&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26414676&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

22. 22.Frei O, Holland D, Smeland OB, Shadrin AA, Fan CC, Maeland S, O’Connell KS, Wang Y, Djurovic S, Thompson WK, et al. Bivariate causal mixture model quantifies polygenic overlap between complex traits beyond genetic correlation. Nat Commun. 2019;10:2417.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41467-019-10310-0&link_type=DOI) 

23. 23.Lloyd-Jones LR, Zeng J, Sidorenko J, Yengo L, Moser G, Kemper KE, Wang H, Zheng Z, Magi R, Esko T, et al. Improved polygenic prediction by Bayesian multiple regression on summary statistics. Nat Commun. 2019;10:5086.
    
    

24. 24.Ni G, Zeng J, Revez JA, Wang Y, Zheng Z, Ge T, Restuadi R, Kiewa J, Nyholt DR, Coleman JRI, et al. A comparison of ten polygenic score methods for psychiatric disorders applied across multiple cohorts. Biol Psychiatry. 2021;90:611–620.
    
    

25. 25.Chang CC, Chow CC, Tellier LCAM, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. GigaScience. 2015;4:s13742-13015-10047-13748.
    
    

26. 26.Ge T, Chen C-Y, Ni Y, Feng Y-CA, Smoller JW. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nat Commun. 2019;10:1776.
    
    

27. 27.Mangiafico S. rcompanion: Functions to support extension education program evaluation. In; 2022.
    
    

28. 28.Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez J-C, Müller M. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:77.
    
    

29. 29.Zhang F, Cao H, Baranova A. Shared genetic liability and causal associations between major depressive disorder and cardiovascular diseases. Front Cardiovasc Med. 2021;8.
    
    

30. 30.Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, Bryois J, Chen C-Y, Dennison CA, Hall LS, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. 2022;604:502–508.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

31. 31.Chauquet S, Zhu Z, O’Donovan MC, Walters JTR, Wray NR, Shah S. Association of antihypertensive drug target genes with psychiatric disorders: a Mendelian randomization study. JAMA Psychiatry. 2021;78:623–631.
    
    

32. 32.Williams SA, Kasl SV, Heiat A, Abramson JL, Krumholz HM, Vaccarino V. Depression and risk of heart failure among the elderly: a prospective community-based study. Psychosom Med. 2002;64:6–12.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoicHN5Y2htZWQiO3M6NToicmVzaWQiO3M6NjoiNjQvMS82IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMTAvMTAvMjAyMi4xMC4wOC4yMjI4MDgwNS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

33. 33.Dennis J, Sealock J, Levinson RT, Farber-Eger E, Franco J, Fong S, Straub P, Hucks D, Song WL, Linton MF, et al. Genetic risk for major depressive disorder and loneliness in sex-specific associations with coronary artery disease. Mol Psychiatry. 2021;26:4254–4264.
    
    

34. 34.Khandaker GM, Zuber V, Rees JMB, Carvalho L, Mason AM, Foley CN, Gkatzionis A, Jones PB, Burgess S. Shared mechanisms between coronary heart disease and depression: findings from a large UK general population-based cohort. Mol Psychiatry. 2020;25:1477–1486.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

35. 35.Tobaldini E, Carandina A, Toschi-Dias E, Erba L, Furlan L, Sgoifo A, Montano N. Depression and cardiovascular autonomic control: a matter of vagus and sex paradox. Neurosci Biobehav Rev. 2020;116:154–161.
    
    

36. 36.Schneider B, Sechtem U. Influence of age and gender in Takotsubo Syndrome. Heart Fail Clin. 2016;12:521–530.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

37. 37.Ridker PM, Rane M. Interleukin-6 signaling and anti-Interleukin-6 therapeutics in cardiovascular disease. Circ Res. 2021;128:1728–1746.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

38. 38.Kaptoge S, Seshasai SRK, Gao P, Freitag DF, Butterworth AS, Borglykke A, Di Angelantonio E, Gudnason V, Rumley A, Lowe GDO, et al. Inflammatory cytokines and risk of coronary heart disease: new prospective study and updated meta-analysis. Eur Heart J. 2013;35:578–589.
    
    

39. 39.Cunningham R, Poppe K, Peterson D, Every-Palmer S, Soosay I, Jackson R. Prediction of cardiovascular disease risk among people with severe mental illness: A cohort study. PLoS One. 2019;14:e0221521–e0221521.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

40. 40.Hippisley-Cox J, Coupland C, Brindle P. Development and validation of QRISK3 risk prediction algorithms to estimate future risk of cardiovascular disease: prospective cohort study. BMJ. 2017;357:j2099.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiYm1qIjtzOjU6InJlc2lkIjtzOjE3OiIzNTcvbWF5MjNfNS9qMjA5OSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIyLzEwLzEwLzIwMjIuMTAuMDguMjIyODA4MDUuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

41. 41.Ministry of Health New Zealand. Cardiovascular disease risk assessment and management for primary care. 2018.
    
    

42. 42.Lichtman JH, Froelicher ES, Blumenthal JA, Carney RM, Doering LV, Frasure-Smith N, Freedland KE, Jaffe AS, Leifheit-Limson EC, Sheps DS, et al. Depression as a risk factor for poor prognosis among patients with acute coronary syndrome: systematic review and recommendations. Circulation. 2014;129:1350–1369.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTQ6ImNpcmN1bGF0aW9uYWhhIjtzOjU6InJlc2lkIjtzOjExOiIxMjkvMTIvMTM1MCI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIyLzEwLzEwLzIwMjIuMTAuMDguMjIyODA4MDUuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

43. 43.Tang B, Yuan S, Xiong Y, He Q, Larsson SC. Major depressive disorder and cardiometabolic diseases: a bidirectional Mendelian randomisation study. Diabetologia. 2020;63:1305–1311.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s00125-020-05131-6&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=32270255&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

44. 44.Martin J, Khramtsova EA, Goleva SB, Blokland GAM, Traglia M, Walters RK, Hübel C, Coleman JRI, Breen G, Børglum AD, et al. Examining sex-differentiated genetic effects across neuropsychiatric and behavioral traits. Biol Psychiatry. 2021;89:1127–1137.
    
    

45. 45.Deloukas P, Kanoni S, Willenborg C, Farrall M, Assimes TL, Thompson JR, Ingelsson E, Saleheen D, Erdmann J, Goldstein BA, et al. Large-scale association analysis identifies new risk loci for coronary artery disease. Nat Genet. 2013;45:25–33.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/ng.2480&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23202125&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

46. 46.Nielsen JB, Thorolfsdottir RB, Fritsche LG, Zhou W, Skov MW, Graham SE, Herron TJ, McCarthy S, Schmidt EM, Sveinbjornsson G, et al. Biobank-driven genomic discovery yields new insight into atrial fibrillation biology. Nat Genet. 2018;50:1234–1239.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41588-018-0171-3&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=30061737&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

47. 47.Mullins N, Forstner AJ, O’Connell KS, Coombes B, Coleman JRI, Qiao Z, Als TD, Bigdeli TB, Børte S, Bryois J, 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.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41588-021-00857-4&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=34002096&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

48. 48.van der Harst P, Verweij N. Identification of 64 novel genetic loci provides an expanded view on the genetic architecture of coronary artery disease. Circ Res. 2018;122:433–443.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTA6ImNpcmNyZXNhaGEiO3M6NToicmVzaWQiO3M6OToiMTIyLzMvNDMzIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMTAvMTAvMjAyMi4xMC4wOC4yMjI4MDgwNS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

49. 49.Shah S, Henry A, Roselli C, Lin H, Sveinbjörnsson G, Fatemifar G, Hedman ÅK, Wilk JB, Morley MP, Chaffin MD, et al. Genome-wide association and Mendelian randomisation analysis provide insights into the pathogenesis of heart failure. Nat Commun. 2020;11:163.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41467-019-13690-5&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

50. 50.Howard DM, Adams MJ, Clarke TK, Hafferty JD, Gibson J, Shirali M, Coleman JRI, Hagenaars SP, Ward J, Wigmore EM, 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.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

51. 51.Lam M, Chen C-Y, Li Z, Martin AR, Bryois J, Ma X, Gaspar H, Ikeda M, Benyamin B, Brown BC, et al. Comparative genetic architectures of schizophrenia in East Asian and European populations. Nat Genet. 2019;51:1670–1678.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/S41588-019-0512-X&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

52. 52.Christophersen IE, Rienstra M, Roselli C, Yin X, Geelhoed B, Barnard J, Lin H, Arking DE, Smith AV, Albert CM, et al. Large-scale analyses of common and rare variants identify 12 new loci associated with atrial fibrillation. Nat Genet. 2017;49:946–952.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/ng.3843&link_type=DOI) 

53. 53.Stahl EA, Breen G, Forstner AJ, McQuillin A, Ripke S, Trubetskoy V, Mattheisen M, Wang Y, Coleman JRI, Gaspar HA, et al. Genome-wide association study identifies 30 loci associated with bipolar disorder. Nat Genet. 2019;51:793–803.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41588-019-0397-8&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=31043756&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

54. 54.Nikpay M, Goel A, Won H-H, Hall LM, Willenborg C, Kanoni S, Saleheen D, Kyriakou T, Nelson CP, Hopewell JC, et al. A comprehensive 1000 Genomes–based genome-wide association meta-analysis of coronary artery disease. Nat Genet. 2015;47:1121–1130.
    
    

55. 55.Wray NR, Ripke S, Mattheisen M, Trzaskowski M, Byrne EM, Abdellaoui A, Adams MJ, Agerbo E, Air TM, Andlauer TMF, et al. Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet. 2018;50:668–681.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41588-018-0090-3&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29700475&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom) 

56. 56.Pardiñas AF, Holmans P, Pocklington AJ, Escott-Price V, Ripke S, Carrera N, Legge SE, Bishop S, Cameron D, Hamshere ML, et al. Common schizophrenia alleles are enriched in mutation-intolerant genes and in regions under strong background selection. Nat Genet. 2018;50:381–389.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41588-018-0059-2&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29483656&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F10%2F2022.10.08.22280805.atom)

 [1]: /embed/inline-graphic-1.gif