Causal effects between negative moods and cardiovascular diseases: a Mendelian randomization study

doi:10.21203/rs.3.rs-3464580/v1

Background.

Previous observational studies have found that negative moods are associated with cardiovascular diseases (CVDs). However, their causal association remains unclear. We aim to evaluate the causal associations of genetically determined negative moods with CVDs including coronary artery disease (CAD), myocardial infarction (MI), stroke, and subtypes of stroke using Mendelian randomization study.

Methods.

Summary statistics for negative moods were extracted from the UK Biobank and a large genome-wide association study including 366,726 subjects. Six negative moods, tiredness, depressed mood, tenseness, unenthusiasm, anxiety, and anger, were included. Genetic associations with CVDs were extracted from the CardiogramplusC4D consortium and MEGASTROKE consortium: CAD (n = 184,305), MI (n = 171,875), stroke (n = 446,696), and subtypes of stroke. Causal effects were inferred between related traits by univariable, reverse and multivariable Mendelian randomization (MR) method.

Results.

Univariable MR analyses found that tiredness was genetically correlated with CAD (OR 1.91; 95% CI 1.23–2.97; P = 0.004) and MI (OR 2.12; 95% CI 1.29–3.47; P = 0.003). Besides, we found that the genetic liability to depressed mood was associated with higher risks of CAD (OR 2.97; 95%CI: 1.25–7.09; P = 0.014) and MI (OR 2.81; 95%CI: 1.06–7.42; P = 0.037). No causal relationship between negative moods and stroke was found. Bidirectional MR analysis failed to find any significant causal effect of CVDs on the risk of negative moods. Multivariable MR analyses revealed compelling evidence of tiredness on the risk of CAD and MI. However, after adjusting for other cardiovascular risk traits, the causal relationship between depressed mood and CVDs disappeared.

Conclusions.

This MR study found that genetic predisposition to tiredness may have a causal effect on the risk of CAD and MI. Early diagnosis and prevention of tiredness may be helpful in the management of CAD and MI.

Cardiovascular disease

genetics

Coronary artery disease

Mendelian randomization

Emotions

fatigue

Cardiovascular diseases (CVDs), including coronary artery disease, myocardial infarction (MI) and stroke, have emerged as a significant global health concern, representing the primary contributors to mortality worldwide [1–3]. Based on the pertinent scholarly literature, the confluences of climate change, air pollution, and the COVID-19 has the potential to affect psychological well-being, encompassing a spectrum of mild negative emotions to extensive psychiatric disorders [4]. Negative moods, encompassing fatigue, depression, stress, anxiety, anger, and pessimism, always exist in patients diagnosed with CVDs [5, 6]. Furthermore, Substantial evidence indicates that individuals exhibiting negative affective states have demonstrated an increased susceptibility to the onset of cardiovascular diseases. A prospective cardiovascular population study of 19,329 patients in Copenhagen City found that fatigue was one of the most critical risk factors for coronary heart disease [7]. A meta-analysis including 893,850 participants performed in Western Europe, North America, and Asia suggested that depression was associated with an increased risk of MI (RR 1.30; 95% CI 1.22–1.40) and incident coronary heart disease (RR 1.30; 95% CI 1.18–1.44) [8].

Nevertheless, it is imperative to note that observational studies are unsuitable for establishing causal relationships due to their susceptibility to confounding variables and reverse causality bias. The findings of a randomized controlled trial investigating the effects of behavioral intervention on patients who underwent angioplasty revealed that no sufficient evidence to establish a cause-and-effect connection between the intervention and the occurrence of new coronary events [9]. Further investigation is warranted to reveal whether this association is causal. Therefore, we chose six negative moods (tiredness, depressed mood, anxiety, anger, tenseness, and unenthusiasm) which were mostly associated with cardiovascular disease as our research objects [10–12].

Mendelian randomization (MR) is extensively employed in the assessment of causal relationship between risk factors and disease outcomes in academic research [13]. Genetic variants exhibit a random distribution pattern during the transmission from parents to offspring, independent of self-adopted lifestyles and environmental confounders [14]. Consequently, the employed approach has the capability to mitigate the issue of reverse causality and circumvent the potential influence of confounding variables [15]. Multivariable Mendelian randomization (MVMR) analyses can be employed to investigate the potential mechanisms that contribute to the relationships observed in univariable analyses [16]. In our study, we employed MR analysis to evaluate the causal impact of various negative moods on cardiovascular diseases using the genome-wide association study.

Study design

We use a two-sample Mendelian randomization study to assess the causal effect between genetically determined negative moods and the risk of CVDs. Before analysis, appropriate SNPs were selected as valid instruments (IVs) for causal inference. In accordance with the previous studies, the MR study we employed must obey the three fundamental assumptions: (1) The IVs are robustly associated with negative moods; (2) The IVs are not related to any possible confounding factors; (3) IVs must affect the risk of the CVDs only through the negative mood traits (Fig. 1).

Data sources

DNA variants with genome-wide significant associations with negative mood-related traits, including tiredness, depressed mood, tenseness, unenthusiasm, and anxiety were obtained from UK Biobank [17]. Genetic signatures for anger were obtained from a large GWAS database [18]. For disease outcomes, we extracted summary-level data for the genetic association with coronary artery disease (CAD) and myocardial infarction (MI) from the Coronary Artery Disease Genome-Wide Replication and Metaanalysis plus the Coronary Artery Disease Genetics (CardiogramplusC4D) consortium, including 184,305 individuals (60,801 CAD cases and 123,504 controls) [19]. Additionally, the causal relationship between stroke and negative moods was investigated as the secondary outcome. The summary-level data are extracted from the MEGASTROKE consortium and meta-analyses of genetic studies for all stroke (40,585 cases and 406,111 controls), subtypes of stroke, intracerebral hemorrhage (3,723 cases and 3,725 controls) and subarachnoid hemorrhage (5,140 cases and 71,952 controls) [20-22].

From corresponding consortiums or GWAS, we also extracted cardiovascular risk factors with genetic associations (serum lipids, body composition, blood pressure, etc.). Table 1 provides detailed information about the data source. The individuals included in our datasets for MR analysis predominantly belonged to European ancestry. As the study relied on publicly available data, there was no need for ethical approval or informed consent.

Selection and validation of instrumental variables (SNPs)

The selection of IVs was conducted in accordance with the three fundamental assumptions of MR studies. We carried out a series of quality control steps to ensure SNPs selected as IV are closely related to exposure. First, to establish a strong correlation between independent variables (IVs) and exposures, single nucleotide polymorphisms (SNPs) linked to six negative moods were chosen from the respective GWAS summary datasets, based on their significance at a genome-wide level ( P < 5 × 10⁻⁸). Second, the LD-Link approach, which is grounded in European populations, was employed to ascertain the independence of the chosen SNPs through pairwise linkage disequilibrium (LD) analysis [29,30]. SNPs that were associated with more SNPs or higher p values dropped off when the r² was greater than 0.001. Third, we evaluated whether the selected SNPs were associated with other traits by using PhenoScanner [31]. We conducted an additional exclusion of SNPs that are linked to confounding factors at the genome-wide significance level, as they have the potential to influence the outcomes of our study. Fourth, in order to ascertain the robustness of the chosen independent variables (IVs), the F statistic was computed for each SNP to determine its validity (F > 10). F statistics exceeding 10 are commonly regarded as sufficiently influential to alleviate any potential bias in estimating the causal IV. Finally, the identified SNPs serving as IVs were subsequently compared to those present in the outcome datasets in order to establish genetic associations. We standardized our statistics for SNP-phenotypes and SNP-outcomes in order to align effect sizes, and excluded any palindromic SNPs (Supplementary Table 9). Supplementary Figure 1 presented the SNP screening process; Supplementary Table 1-8 presented detailed information for identified SNPs.

Statistical analysis

We employed the conventional inverse-variance-weighted (IVW) method as the major approach. Under the assumption that all SNPs were valid, the IVW approach utilized an inverse-variance weighted formula to conduct a meta-analysis of the ratio estimate for each SNP [32]. To assess the potential causal associations between six negative moods and CVDs, several methods including weighted median, MR-Egger, maximum likelihood, weighted mode, and simple mode were employed as complements. In sensitivity analysis, we employed the weighted median approach, MR-Egger approach, and MR Pleiotropy Residual Sum and Outlier (MR-PRESSO) approach [33-35]. Furthermore, the heterogeneity among SNPs in the IVW approach was assessed by using Q statistics and visually examined through funnel plots [36, 37]. In addition, we conducted a leave-one-out analysis to systematically exclude each exposure-associated SNP individually, aiming to detect any potential pleiotropic or outlier SNP. The results are reported in the form of odds ratios (ORs) with 95% confidence intervals (95% CI) and P-values, along with a forest plot. Using the same methods, a reverse MR study explored the causal relationship between negative moods and cardiovascular disease.

We conducted a multivariable MR (MVMR) analysis to evaluate the direct effect of negative moods on CVDs risk [38]. The adjusted factors in MVMR included serum lipids, body composition, blood pressure, HbAlc, smoke, year of schooling, and physical activity. All analyses were performed using the “TwoSampleMR”， “MR-PRESSO”, “Mendelian Randomization”, and "MVMR" packages in R Version 4.2.3. Power calculations were performed on mRnd to determine the minimum effect size necessary to attain a power exceeding 80%, while considering the sample size as the determinant of significance for each outcome (Supplementary Table 11) [39].

Instrumental variables selection

The data sources are displayed in Table 1. The ultimate number of SNPs that were selected as the IVs were 20 (Tiredness), 9 (Depressed mood), 8 (Tenseness), 3 (Unenthusiasm), 38 (Anxiety), and 27 (Anger). The detailed process is shown in the Supplementary Figure 1. The F-statistic of each SNP selected as IVs was greater than 10. As a result, there are no weak instrumental variables (Supplementary Table 1–6).

Univariable MR analysis of the association between negative mood traits and CVDs risk

The MR results for six negative mood traits associated with CVDs risk are presented in Figure 2,3. The IVW analyses revealed that tiredness was associated with an increased risk of CAD and MI, the ORs were 1.91 (95% confidence interval [95% CI]: 1.23 - 2.96; P = 0.004) for CAD and 2.12 (95% CI: 1.29 - 3.47; P = 0.003) for MI (Fig. 2). Also, genetically predicted depressed mood was associated with a raised risk of CAD and MI. The ORs were 2.97(95%CI: 1.25-7.09; P = 0.014) for CAD and 2.81 (95%CI: 1.06-7.42; P = 0.037) for MI. Several methods were employed as complements (Supplementary Table 10).

However, in the study of the effect of other negative moods on cardiovascular disease risk. There was a lack of statistically significant correlation observed between genetically predicted moods of tenseness, unenthusiasm, anxiety, anger, and cardiovascular diseases such as CAD, MI, stroke, and so on.

The results were stable in sensitivity analyses. There was no evidence of heterogeneity and horizontal pleiotropy in CAD and MI using the Q-statistic and the intercept of MR-Egger regression (Supplementary Table 12). Additionally, scatter plots provided further evidence of the absence of directional pleiotropy, while the leave-one-out analysis demonstrated that the causal association between CAD and MI was not affected by any individual SNP. Nevertheless, in the analysis of the association between depressed mood and CAD and MI, the IVW results were not consistent with other methodologies. In addition, unenthusiasm cannot perform sensitivity analysis well due to the small number of SNPs.

Bidirectional MR analysis of the association between negative mood traits and CVDs risk

To analyze the effect of CAD, and MI (exposure) on the risk of six negative moods (outcomes), we obtained 29 and 20 SNPs as IVs for exposure with a genome-wide level (P <5×10^-8) and LD removed (Supplementary Table 7-8). The F-statistic of IVs was greater than 10, which manifested the genetic strength was sufficient. The IVW method indicated that CAD and MI were not associated with the risk of six negative moods. More details of the results are shown in Figure 4.

Multivariable MR analysis of the association of tiredness and depressed mood with CAD/MI risk conditional on cardiovascular risk factors

We utilized multivariable MR analyses to estimate the causal association of tiredness and depressed mood on the likelihood of developing CAD and MI while considering cardiovascular risk factors. After the adjustment for cardiovascular risk factors, a robust correlation between tiredness and the risk of CAD and MI was observed (Table 2). However, when accounting for diastolic blood pressure (DBP) or total cholesterol (TC), the effect of tiredness on the risk of MI became insignificant. Similarly, the effect between tiredness and the risk of CAD was insignificant when adjusted for TC, DBP, or education. Based on the fully adjusted model, the effect of tiredness on the risk of CAD (OR: 1.979; 95% CI: 1.015-3.857; P=0.045) and MI (OR: 2.406; 95% CI:1.217-4.756; P=0.012) was still significant (Table 2). Besides, the causal association of depressed mood with the risk of CAD and MI was not observed after adjusting cardiovascular risk factors (Table 3).

This MR study explored the causal association between negative moods and CVDs. Our study demonstrated a significant positive correlation between genetic liability to tiredness and depressed mood with CAD and MI. However, there was a lack of empirical evidence suggesting genetically predicted tenseness, unenthusiasm, anxiety and, anger were causally associated with CVDs risk. In the reverse MR study, the results indicated no causal association between CAD, MI and negative mood risk. Using MVMR analyses, our study revealed an independent association between tiredness and CAD as well as MI. However, this effect was diminished when adjusting for diastolic blood pressure (DBP) or total cholesterol (TC). Additionally, after adjusting risk traits related to cardiovascular disease, the casual effect between depressed mood and CVDs disappeared. Other cardiovascular risk factors may contribute to the impact of depressed mood on CVDs.

Multiple systematic reviews and prospective analyses have suggested that psychosocial fact independently forecasts the occurrence of coronary heart disease in initially healthy populations, regardless of conventional risk factors [40, 41]. But there is still no consensus on which emotions cause cardiovascular disease. Previous research had yielded inconsistent findings regarding the association between tiredness and cardiovascular diseases. Williams et al. conducted a study involving a diverse group of participants without preexisting heart disease and discovered that both depressed mood and tiredness were strong predictors of developing cardiac events [42]. However, a prospective study focusing on the treatment of exhaustion in Angioplasty patients and its impact on the occurrence of new cardiac events did not confirm the effectiveness of behavioral intervention in reducing the risk (RR: 0.34; 95% CI: 0.11–1.07; P = 0.07) [11]. Various negative moods exhibit varying degrees of predictive ability for incident or recurrent cardiac events. A comprehensive meta-analysis of 30 prospective studies revealed a significant association between depressed mood and both coronary artery disease (RR: 1.30; 95% CI: 1.22–1.40) and myocardial infarction (RR: 1.30; 95% CI:1.18–1.44) risk [43]. Accordingly, as Pedersen and Zimmermann discovered, depressed mood following percutaneous coronary angioplasty predicts adverse cardiac events, fatigue did not demonstrate the same predictive capacity [44, 45]. Appels et al. found that AMI was predicted by fatigue in men without heart disease, while depressed mood was not a significant predictor of AMI when tiredness was taken into account. Above all, these results indicated the potentially pivotal role of negative moods, especially tiredness, depressed mood, tenseness, unenthusiasm, anxiety, and anger, in the pathogenesis of CVD and corresponding major adverse cardiovascular events (MACEs). However, the causal relationship between negative moods and clinical outcomes remained to be unclear by conventional observational studies, limited by their lower. Therefore, we chose six negative moods that may have resulted in CVDs using MR study.

In our study, we initially employed MR analysis to investigate the causal relationship between six negative moods and CVDs. Tiredness, especially mental fatigue, is a psychobiological condition resulting from extended periods of demanding cognitive tasks [46, 47]. This state has significant implications for various facets of everyday life. Numerous comprehensive prospective studies have demonstrated that tiredness is a predictor of the CVDs [48–52]. Our findings indicate that tiredness can serve as an independent predictor for CAD and MI, which aligns with the conclusions drawn by Williams et al. Furthermore, we conducted sensitivity analyses and intervened on potential cardiovascular risk factors to ensure the robustness of our evaluation. However, a prospective study focusing on the treatment of exhaustion in Angioplasty patients did not confirm the existence of a causal relationship [11]. This discrepancy may be attributed to the fact that fatigued patients tend to seek more medical assistance and adhere to secondary prevention measures. The potential mechanism underlying the association between tiredness and CVDs is proposed to involve serological markers of inflammation, such as heightened leukocyte adhesiveness/aggregation and elevated levels of interleukin (IL-6, IL-10), tumor necrosis factor, and C-reactive protein [53]. Additionally, it has been proposed that tiredness plays an important role in the dysregulation of the hypothalamus-pituitary-adrenal (HPA) axis as a result of prolonged exposure to stress, leading to reduced levels of the adrenocorticotropic hormone, decreased cortisol levels, and diminished concentrations of thyroid hormones [54].

We also demonstrated a causal relationship between depressed mood and CAD and MI, which is consistent with previous MR studies. The pathophysiological process of depressed mood is multifaceted, involving higher body composition traits, hypertension, endothelial dysfunction, and increased sympathetic excitability. A meta-analysis corroborated these findings by revealing shared risk factors, such as interleukin-6, C-reactive protein, and serum lipids, between depressed mood and coronary artery disease, aligning with clinical observations [55–61]. However, after adjusting for potential confounders in MVMR, the causal relationship disappeared. This discrepancy may be attributed to selection bias. Firstly, the subjects we studied primarily had depressed mood, and the causal effect on cardiovascular disease may not be as strong. Secondly, emotions are often influenced by numerous confounding factors. Our findings do not contradict previous MR studies. In contrast, our study further established a causal association between depression and CAD/MI. Observational studies have consistently reported an association between mental health and stroke, including both ischemic and hemorrhagic types. However, our study did not find any causal relationship between genetic predisposition to negative moods and the overall risk of stroke and its subtypes. Furthermore, we employed a bidirectional MR study design to mitigate the impact of reverse causality and minimize residual confounding. Consequently, the causal relationship between genetic susceptibility to CAD, MI, and negative moods remains unproven.

Our study encompasses several strengths. Firstly, we explored the correlation between genetically determined negative moods and cardiovascular diseases using MR methods, incorporating multiple sensitivity analyses and intervening on potential cardiovascular risk factors to ensure the robustness of our findings. Secondly, by utilizing genetically determined instrumental variables, we aimed to mitigate the impact of confounding factors and reverse causality commonly encountered in observational studies. The utilization of MR analysis facilitates a more robust causal inference regarding the relationship between tiredness, depressed mood, and the risk of CAD. Lastly, the implementation of MVMR analyses was employed to counteract the biases of confounding factors and reverse causal relationships, thereby aiding in the elucidation of relationships observed in univariable analyses.

However, our research encountered several limitations. Firstly, it was challenging to completely eliminate the possibility of pleiotropy, which was an inherent constraint in the MR analysis. Nonetheless, we mitigated this limitation by adjusting for other traits in our MVMR, and the sensitivity analyses produced consistent and reliable outcomes. Secondly, the statistical power of our investigation fell short of the desired 80% when examining the impact of six negative moods on the risk of stroke and its subtypes (Supplementary Table 11). Insufficient sample size and low variability in exposures may account for this inadequacy. Furthermore, the present study solely examined the correlation between negative moods and the risk of CVDs from a genetic perspective. Lastly, uncertainty remains regarding the applicability of these findings to other populations, considering that study participants were of European ancestry.

In summary, our MR study demonstrated that genetic liability to tiredness had a risk effect on the occurrence of CAD and MI. In individuals at risk for CAD and MI, tiredness mood traits may prove helpful in preclinical diagnosis and intervention.

CVDs = cardiovascular diseases; CAD = coronary artery disease; MI = myocardial infarction; MR = Mendelian randomization; MVMR = Multivariable Mendelian randomization; IVs = valid instruments; OR = odds ratio; 95% CI = 95% confidence intervals; DBP = diastolic blood pressure; TC = total cholesterol

Acknowledgements

The authors thank the CARDIoGRAMplusC4D Consortium, UK Biobank, MEGASTROKE Consortium, ISGC Consortium, GIANT Consortium, ICBP Consortium, GLGC Consortium, SSGAC Consortium, and GSCAN Consortium for making the GWAS data publicly available. We would like to thank the researchers and study participants for their contributions.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82000404 and No.82100421) and a grant from the Natural Science Funds of Zhejiang Province, China (Project No. LQ21H020003 and LQ21H020006).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Joseph P, Leong D, McKee M, Anand SS, Schwalm JD, Teo K, et al. (2017). Reducing the global burden of cardiovascular disease, part 1: the epidemiology and risk factors. Circ Res, 121, 677–694. doi: 10.1161/CIRCRESAHA.117.308903.
Sanchis-Gomar F, Perez-Quilis C, Leischik R, Lucia A. (2016). Epidemiology of coronary heart disease and acute coronary syndrome. Ann Transl Med, 4(13), 256. doi: 10.21037/atm.2016.06.33.
Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R, et al. (2017). American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics-2017 update: a report from the American Heart Association. Circulation, 135(10), e146-e603. doi: 10.1161/CIR.0000000000000485.
Marazziti D, Cianconi P, Mucci F, Foresi L, Chiarantini I, Della Vecchia A. (2021). Climate change, environment pollution, COVID-19 pandemic and mental health. Sci Total Environ, 773, 145182. doi: 10.1016/j.scitotenv.2021.145182.
Tully PJ, Cosh SM, Baumeister H. (2014). The anxious heart in whose mind? A systematic review and meta-regression of factors associated with anxiety disorder diagnosis, treatment and morbidity risk in coronary heart disease. J Psychosom Res, 77(6), 439-448.doi: 10.1016/j.jpsychores.2014.10.001.
Davidson KW, Alcántara C, Miller GE. (2018). Selected psychological comorbidities in coronary heart disease: Challenges and grand opportunities. Am Psychol, 73(8), 1019-1030. doi: 10.1037/amp0000239.
Schnohr P, Marott JL, Kristensen TS, Gyntelberg F, Grønbæk M, Lange P, et al. (2015). Ranking of psychosocial and traditional risk factors by importance for coronary heart disease: the Copenhagen City Heart Study. Eur Heart J, 36(22), 1385-1393. doi: 10.1093/eurheartj/ehv027.
Gan Y, Gong Y, Tong X, Sun H, Cong Y, Dong X, et al. (2014). Depression and the risk of coronary heart disease: a metaanalysis of prospective cohort studies. BMC Psychiatry, 14, 371. doi: 10.1186/s12888-014-0371-z.
Appels A, Bär F, van der Pol G, Erdman R, Assman M, Trijsburg W, et al. (2005). Effects of treating exhaustion in angioplasty patients on new coronary events: results of the randomized Exhaustion Intervention Trial (EXIT). Psychosom Med, 67(2), 217-223. doi: 10.1097/01.psy.0000151485.
Levine GN, Cohen BE, Commodore-Mensah Y, Fleury J, Huffman JC, Khalid U, et al. (2021). Psychological Health, Well-Being, and the Mind-Heart-Body Connection: A Scientific Statement From the American Heart Association. Circulation, 143(10), e763-e783. doi: 10.1161/CIR.0000000000000947.
Appels A, Mulder P. (1988). Excess fatigue as a precursor of myocardial infarction. Eur Heart J, 9(7), 758-764. doi: 10.1093/eurheartj/9.7.758.
Eurelings LS, van Dalen JW, Ter Riet G, Moll van Charante EP, Richard E, van Gool WA, et al. (2018). Apathy and depressive symptoms in older people and incident myocardial infarction, stroke, and mortality: a systematic review and meta-analysis of individual participant data. Clin Epidemiol, 10, 363-379. doi: 10.2147/CLEP.S150915.
Emdin CA, Khera AV, Kathiresan S. (2017). Mendelian Randomization. JAMA, 318(19), 1925-1926. doi: 10.1001/jama.2017.17219.
Davey Smith G, Hemani G. (2014). Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet, 23(R1), R89-98. doi: 10.1093/hmg/ddu328.
Holmes MV, Ala-Korpela M, Smith GD. (2017). Mendelian randomization in cardiometabolic disease: challenges in evaluating causality. Nat Rev Cardiol, 14(10), 577-590. doi: 10.1038/nrcardio.2017.78.
Sanderson E, Davey Smith G, Windmeijer F, Bowden J. (2019). An examination of multivariable mendelian randomization in the single-sample and two-sample summary data settings. Int J Epidemiol, 48(3), 713–727. doi: 10.1093/ije/dyy262.
Backman JD, Li AH, Marcketta A, Sun D, Mbatchou J, Kessler MD, et al. (2021). Exome sequencing and analysis of 454,787 UK Biobank participants. Nature, 599(7886), 628-634. doi: 10.1038/s41586-021-04103-z.
Nagel M, Watanabe K, Stringer S, Posthuma D, van der Sluis S. (2018). Item-level analyses reveal genetic heterogeneity in neuroticism. Nat Commun, 9(1), 905. doi: 10.1038/s41467-018-03242-8.
Nikpay M, Goel A, Won HH, Hall LM, Willenborg C, Kanoni S, et al. (2015). A comprehensive 1,000 Genomes-based genome-wide association meta-analysis of coronary artery disease. Nat Genet, 47, 1121–1130. doi: 10.1038/ng.3396.
Malik R, Chauhan G, Traylor M, Sargurupremraj M, Okada Y, Mishra A, et al. (2018). Multiancestry genome-wide association study of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes. Nat Genet, 50, 524–537. doi: 10.1038/s41588-018-0058-3.
Woo D, Falcone GJ, Devan WJ, Brown WM, Biffi A, Howard TD, et al. (2014). Meta-analysis of genome-wide association studies identifies 1q22 as a susceptibility locus for intracerebral hemorrhage. Am J Hum Genet, 94, 511–521.doi: 10.1016/j.ajhg.2014.02.012.
Bakker MK, van der Spek RA, van Rheenen W, Morel S, Bourcier R, Hostettler IC, et al. (2020). Genome-wide association study of intracranial aneurysms identifies 17 risk loci and genetic overlap with clinical risk factors. Nat Genet, 52, 1303–1313. doi: 10.1038/s41588-020-00725-7.
Yengo L, Sidorenko J, Kemper KE, Zheng Z, Wood AR, Weedon MN, et al. (2018). GIANT Consortium. Meta-analysis of genome-wide association studies for height and body mass index in ∼700000 individuals of European ancestry. Hum Mol Genet, 27(20), 3641-3649. doi: 10.1093/hmg/ddy271.
Evangelou E, Warren HR, Mosen-Ansorena D, Mifsud B, Pazoki R, Gao H, et al. (2018). Million Veteran Program. Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits. Nat Genet, 50(10), 1412-1425. doi: 10.1038/s41588-018-0205-x.
Willer CJ, Schmidt EM, Sengupta S, Peloso GM, Gustafsson S, Kanoni S, et al. (2013). Discovery and refinement of loci associated with lipid levels. Nat Genet, 45(11), 1274-1283. doi: 10.1038/ng.2797.
Lee JJ, Wedow R, Okbay A, Kong E, Maghzian O, Zacher M, et al. (2018). Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nat Genet, 50(8), 1112–1121. doi: 10.1038/s41588-018-0147-3.
Liu M, Jiang Y, Wedow R, Li Y, Brazel DM, Chen F, et al. (2019). Association studies of up to 1.2 million individuals yield new insights into the genetic etiology of tobacco and alcohol use. Nat Genet, 51(2), 237-244. doi: 10.1038/s41588-018-0307-5.
Howe LJ, Nivard MG, Morris TT, Hansen AF, Rasheed H, Cho Y, et al. (2022). Within-sibship genome-wide association analyses decrease bias in estimates of direct genetic effects. Nat Genet, 54(5), 581-592. doi: 10.1038/s41588-022-01062-7.
Machiela MJ, Chanock SJ. (2015). LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics, 31(21), 3555-3557. doi: 10.1093/bioinformatics/btv402.
Myers TA, Chanock SJ, Machiela MJ. (2020). LDlinkR: An R Package for Rapidly Calculating Linkage Disequilibrium Statistics in Diverse Populations. Front Genet, 11, 157. doi: 10.3389/fgene.2020.00157.
Staley JR, Blackshaw J, Kamat MA, Ellis S, Surendran P, Sun BB, et al. (2016). PhenoScanner: a database of human genotype-phenotype associations. Bioinformatics, 32(20), 3207-3209. doi: 10.1093/bioinformatics/btw373.
Burgess S, Butterworth A, Thompson SG. (2013). Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol, 37(7), 658-665. doi: 10.1002/gepi.21758.
Bowden J, Davey Smith G, Haycock PC, Burgess S. (2016). Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol, 40(4), 304–314. doi: 10.1002/gepi.21965.
Bowden J, Davey Smith G, Burgess S. (2015). Mendelian randomization with invalid instruments: Effect estimation and bias detection through Egger regression. Int J Epidemiol, 44(2), 512–525. doi: 10.1093/ije/dyv080.
Verbanck M, Chen CY, Neale B, Do R. (2018). Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet, 50(5), 693–698. doi: 10.1038/s41588-018-0099-7.
Greco M FD, Minelli C, Sheehan NA, Thompson JR. (2015). Detecting pleiotropy in Mendelian randomisation studies with summary data and a continuous outcome. Stat Med, 34(21), 2926-2940. doi: 10.1002/sim.6522.
Sterne JA, Sutton AJ, Ioannidis JP, et al. (2011). Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ, 343, d4002. doi: 10.1136/bmj.d4002.
Burgess S, Thompson SG. (2015). Multivariable Mendelian randomization: the use of pleiotropic genetic variants to estimate causal effects. Am J Epidemiol, 181(4), 251-260. doi: 10.1093/aje/kwu283.
Brion MJ, Shakhbazov K, Visscher PM. (2013). Calculating statistical power in Mendelian randomization studies. Int J Epidemiol, 42, 1497–1501. doi: 10.1093/ije/dyt179.
Everson-Rose SA, Lewis TT. (2005). Psychosocial factors and cardiovascular diseases. Annu Rev Public Health, 26, 469–500. doi: 10.1146/annurev.publhealth.26.021304.144542.
Kuper H, Marmot M, Hemingway H. (2002). Systematic review of prospective cohort studies of psychosocial factors in the etiology and prognosis of coronary heart disease. Semin Vasc Med, 2, 267–314. doi: 10.1055/s-2002-35401.
Williams JE, Mosley TH Jr, Kop WJ, Couper DJ, Welch VL, Rosamond WD. (2010). Vital exhaustion as a risk factor for adverse cardiac events (from the Atherosclerosis Risk In Communities [ARIC] study). Am J Cardiol, 105, 1661Y5. doi: 10.1016/j.amjcard.2010.01.340.
Gan Y, Gong Y, Tong X, Sun H, Cong Y, Dong X, et al. (2014). Depression and the risk of coronary heart disease: a meta-analysis of prospective cohort studies. BMC Psychiatry, 14, 371. doi: 10.1186/s12888-014-0371-z.
Stress Pedersen SS, Denollet J, Daemen J, van de Sande M, de Jaegere PT, Serruys PW, et al. (2007). Fatigue, depressive symptoms, and hopelessness as predictors of adverse clinical events following percutaneous coronary intervention with paclitaxel-eluting stents. J Psychosom Res, 62, 455Y61. doi: 10.1016/j.jpsychores.
Zimmermann-Viehoff F, Wang HX, Kirkeeide R, Schneiderman N, Erdur L, Deter HC, et al. (2013). Women's exhaustion and coronary artery atherosclerosis progression: The Stockholm Female Coronary Angiography Study. Psychosom Med, 75(5), 478-485. doi: 10.1097/PSY.0b013e3182928c28.
Desmond PA, Hancock PA. Active and passive fatigue states. In: Desmond PA, Hancock PA, editors. Stress, workload and fatigue. Mahwah: Lawrence Erlbaum Associates; 200, p. 455–465.
Job RFS, Dalziel J. Defining fatigue as a condition of the organism and distinguishing it from habituation, adaptation and boredom. In: Hancock PA, Desmond PA, editors. Stress, workload and fatigue. Mahwah: Lawrence Erlbaum Associates; 2001, p. 466–475.
Appels A, Falger PR, Schouten EG. (1993). Vital exhaustion as risk indicator for myocardial infarction in women. J Psychosom Res, 37, 881Y90. doi: 10.1016/0022-3999(93)90177-h.
Appels A, Mulder P. (1989). Fatigue and heart disease. The association between ‘vital exhaustion’ and past, present and future coronary heart disease. J Psychosom Res, 33, 727Y38. doi: 10.1016/0022-3999(89)90088-3.
Cole SR, Kawachi I, Sesso HD, Paffenbarger RS, Lee IM. (1999). Sense of exhaustion and coronary heart disease among college alumni. Am J Cardiol, 84, 1401Y5. doi: 10.1016/s0002-9149(99)00585-8.
Prescott E, Holst C, Gronbaek M, Schnohr P, Jensen G, Barefoot J. (2003). Vital exhaustion as a risk factor for ischaemic heart disease and all-cause mortality in a community sample. A prospective study of 4084 men and 5479 women in the Copenhagen City Heart Study. Int J Epidemiol, 32, 990Y7. doi: 10.1093/ije/dyg235.
Schuitemaker GE, Dinant GJ, van der Pol GA, Appels A. (2004). Assessment of vital exhaustion and identification of subjects at increased risk of myocardial infarction in general practice. Psychosomatics, 45, 414Y8. doi: 10.1176/appi.psy.45.5.414.
Osimo EF, Pillinger T, Rodriguez IM, Khandaker GM, Pariante CM, Howes OD. (2020). Inflammatory markers in depression: a meta-analysis of mean differences and variability in 5,166 patients and 5,083 controls. Brain Behav Immun, 87, 901–909. doi: 10.1016/j.bbi.2020.02.010.
Bunevicius A, Gintauskiene V, Podlipskyte A, Zaliunas R, Brozaitiene J, Prange AJ Jr, Bunevicius R. (2012). Fatigue in patients with coronary artery disease: association with thyroid axis hormones and cortisol. Psychosom Med, 74(8), 848-853. doi: 10.1097/PSY.0b013e318268e7f0.
Khandaker GM, Zuber V, Rees JMB, Carvalho L. (2020). Shared mechanisms between coronary heart disease and depression: findings from a large UK general population-based cohort. Mol Psychiatry, 25, 1477–1486. doi: 10.1038/s41380-019-0395-3.
Mulugeta A, Zhou A, Vimaleswaran KS, Dickson C, Hypponen E. (2019). Depression increases the genetic susceptibility to high body mass index: evidence from UK Biobank. Depress Anxiety, 36, 1154 –1162. doi: 10.1002/da.22963.
Speed MS, Jefsen OH. (2019). Investigating the association between body fat and depression via Mendelian randomization. Transl Psychiatry, 9, 184. doi: 10.1038/s41398-019-0516-4.
Meng L, Chen D, Yang Y, Zheng Y, Hui R. (2012). Depression increases the risk of hypertension incidence: a meta-analysis of prospective cohort studies. J Hypertens, 30, 842–851. doi: 10.1097/HJH.0b013e32835080b7.
Veith RC, Lewis N, Linares OA, Barnes RF, Raskind MA, Villacres EC, et al. (1994). Sympathetic nervous system activity in major depression. Basal and desipramine-induced alterations in plasma norepinephrine kinetics. Arch Gen Psychiatry, 51, 411– 422. doi: 10.1001/archpsyc.
Sherwood A, Hinderliter AL, Watkins LL, Waugh RA, Blumenthal JA. (2005). Impaired endothelial function in coronary heart disease patients with depressive symptomatology. J Am Coll Cardiol, 46, 656–659. doi: 10.1016/j.jacc.2005.05.041.
Osimo EF, Pillinger T, Rodriguez IM, Khandaker GM, Pariante CM, Howes OD. (2020). Inflammatory markers in depression: a meta-analysis of mean differences and variability in 5,166 patients and 5,083 controls. Brain Behav Immun, 87, 901–909. doi: 10.1016/j.bbi.2020.02.010.

Table1. Overview of data sources.

Traits	Data source	Ancestery	Cases(patients/controls)
Exposure
Frequency of tiredness / lethargy in last 2 weeks	UK Biobank (17)	European	449,019
Frequency of depressed mood in last 2 weeks	UK Biobank (17)	European	442,480
Frequency of tenseness / restlessness in last 2 weeks	UK Biobank (17)	European	445,194
Frequency of unenthusiasm / disinterest in last 2 weeks	UK Biobank (17)	European	447,403
Worrier / anxious feelings	UK Biobank (17)	European	450,765
Irritable mood	GWAS by Nagel M. (18)	European	366,726
Outcome
CAD	CARDIoGRAM and CARDIoGRAMplusC4D (19)	Trans-Ethnics	184,305(60,801/123,504)
MI	CARDIoGRAM and CARDIoGRAMplusC4D (19)	Trans-Ethnics	171,873(43,676/128,197)
All stroke	MEGASTROKE consortium (20)	European	446,696(40,585/406,111)
Any ischemic stroke	MEGASTROKE consortium (20)	European	440,328(34,217/406,111)
Large artery stroke	MEGASTROKE consortium (20)	European	409,484(3,373/406,111)
Small vessel stroke	MEGASTROKE consortium (20)	European	411,497(5,386/406,111)
Cardioembolic stroke	MEGASTROKE consortium (20)	European	413,304(7,193/406,111)
Intracerebral hemorrhage	ISGC (21)	European	6,948(3,723/3,725)
Subarachnoid hemorrhage	GWAS by Bakker et al. (22)	European	77,092(5,140/71,952)
Body composition
BMI	GIANT (23)	European	339,224
WHR	GIANT (23)	European	224,459
Blood pressure
SBP	ICBP (24)	European	757,601
DBP	ICBP (24)	European	757,601
Serum lipids
HDL-C	GLGC (25)	European	187,167
LDL-C	GLGC (25)	European	173,082
TC	GLGC (25)	European	187,365
TG	GLGC (25)	European	177,861
Other related traits
Years of schooling	SSGAC (26)	European	766,345
Cigarettes smoked per day	GSCAN (27)	European	249,752
HbAlc	Within family GWAS consortium (28)	European	45,734
Physical activity	Within family GWAS consortium (28)	European	78,007

CAD, coronary artery disease; MI, myocardial infarction; BMI, body mass index; WHR, waist-to-hip ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; HbAlc, glycosylated hemoglobin.

Table 2 Multivariable Mendelian randomization results for association of tiredness with CAD/MI.

Traits	Tiredness-CAD				Tiredness-MI
Traits	SNPs	P-value	Beta(95%CI)	OR(95%CI)	SNPs	P-value	Beta(95%CI)	OR(95%CI)
BMI	96	0.018	0.553(0.095-1.0124)	1.740(1.100-2.752）	96	0.006	0.660(0.192-1.128）	1.935(1.211-3.091）
WHR	53	0.046	0.586(0.0102-1.163)	1.797(1.010-3.198）	53	0.017	0.691(0.123-1.260）	2.000(1.131-3.526)
HDL	104	0.034	0.794(0.062-1.526)	2.212(1.064-4.600)	104	0.038	0.752(0.043-1.461)	2.121(1.044-4.309)
LDL	98	0.035	0.696(0.049-1.342)	2.005(1.050-3.829)	98	0.017	0.842(0.150-1.534)	2.321(1.162-4.637)
TC	101	0.118	0.593(-0.150-1.336)	1.809(0.860-3.805)	101	0.060	0.768(-0.033-1.569)	2.155(0.967-4.804)
TG	75	0.017	0.733(0.133-1.333)	2.081(1.142-3.793)	101	0.014	0.748(0.148-1.348)	2.113(1.160-3.851)
Systolic blood pressure	420	0.002	0.897(0.329-1.465)	2.451(1.389-4.325)	420	0.001	1.019(0.453-1.585)	2.772(1.574-4.882)
Diastolic blood pressure	429	0.231	0.324(-0.207-0.855)	1.383(0.813-2.351)	243	0.151	0.429(-0.156-1.013)	1.535(0.856-2.754)
HbAlc	68	0.006	0.535(0.151-0.919)	1.707(1.163-2.507)	68	0.007	0.567(0.153-0.981)	1.763(1.166-2.666)
Years of schooling(Education)	304	0.087	0.295(-0.043-0.634)	1.344(0.958-1.884)	304	0.007	0.527(0.147-0.907)	1.694(1.159-2.477)
Cigarettes smoked per day(Smoke)	53	0.001	0.653(0.270-1.036)	1.921(1.310-2.818)	53	0.003	0.684(0.230-1.138)	1.982(1.258-3.120)
Physical activity	36	0.012	0.687(0.154-1.221)	1.988(1.167-3.389）	36	0.020	0.706(0.112-1.300）	2.027(1.119-3.670）
All traits(BMI, LDL, SBP, HbAlc, Smoke, Education, Physcical activity)	373	0.045	0.682(0.015-1.350)	1.979(1.015-3.857)	373	0.012	0.878(0.197-1.559)	2.406(1.217-4.756)

CAD, coronary artery disease; MI, myocardial infarction; BMI, body mass index; WHR, waist-to-hip ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; HbAlc, glycosylated hemoglobin.

Table 3 Multivariable Mendelian randomization results for association of depressed mood with CAD/MI.

Traits	Depressed-CAD				Depressed-MI
Traits	SNPs	P-value	Beta(95%CI)	OR(95%CI)	SNPs	P-value	Beta(95%CI)	OR(95%CI)
BMI	87	0.276	0.458(-0.366-1.282)	1.582(0.694-3.606)	87	0.420	0.341(-0.488-1.170)	1.406(0.614-3.223)
WHR	42	0.264	0.662(-0.501-1.825)	1.939(0.606-6.203)	42	0.409	0.452(-0.622-1.526)	1.572(0.537-4.602)
HDL	92	0.935	-0.063(-1.592-1.465)	0.938(0.203-4.327)	92	0.696	-0.290(-1.745-1.165)	0.938(0.203-4.327)
LDL	89	0.474	0.486(-0.844-1.816)	1.626(0.430-6.148)	89	0.676	0.303(-0.844-1.816)	1.354(0.327-5.601)
TC	90	0.612	0.382(-1.093-1.857)	1.465(0.335-6.406)	90	0.625	0.392(-1.182-1.967)	1.481(0.307-7.152)
TG	65	0.666	0.248(-0.879-1.376)	1.282(0.415-3.961)	65	0.775	0.164(-0.964-1.293)	1.179(0.381-3.644)
Systolic blood pressure	417	0.027	0.949(0.109-1.788)	2.582(1.115-5.979)	417	0.011	1.082(0.248-1.915)	2.950(1.282-6.789)
Diastolic blood pressure	428	0.211	0.521(-0.295-1.337)	1.684(0.745-3.808)	430	0.200	0.559(-0.295-1.412)	1.748(0.744-4.106)
HbAlc	44	0.019	1.034(0.173-1.896)	2.815(1.189-6.661)	44	0.043	0.892(0.030-1.754)	2.440(1.030-5.777)
Years of schooling(Education)	300	0.239	0.365(-0.242-0.971)	1.440(0.785-2.640)	300	0.134	0.519(-0.160-1.199)	1.681(0.852-3.316)
Cigarettes smoked per day(Smoke)	35	0.003	1.073(0.361-1.784)	2.924(1.435-5.955)	35	0.078	0.787(-0.088-1.663)	2.197(0.916-5.274)
Physical activity	14	0.029	0.984(0.100-1.869)	2.676(1.105-6.482)	14	0.032	0.879(0.074-1.684)	2.409(1.077-5.388)
All traits(BMI, LDL, SBP, HbAlc, Smoke, Education, Physcical activity)	372	0.223	0.646(-0.392-1.654)	1.908(0.675-5.390)	372	0.223	0.646(-0.392-1.685)	1.908(0.675-5.390)

CAD, coronary artery disease. CAD, coronary artery disease; MI, myocardial infarction; BMI, body mass index; WHR, waist-to-hip ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; HbAlc, glycosylated hemoglobin.

No competing interests reported.

SupplementaryTable.docx

Causal effects between negative moods and cardiovascular diseases: a Mendelian randomization study

Status:

Version 1

Abstract

Background.

Methods.

Results.

Conclusions.

Figures

Introduction

Methods

Result

Discussion

Conclusion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1