Introduction

Neuroticism is one of the five higher order factors taxonomically defining one’s personality. The modern conception of neuroticism was initially introduced by Hans Eysenck and his contemporaries when it was defined in Freudian theory terms (Ormel et al. 2013). Currently, the consensus has neuroticism as the tendency to experience negative emotions, including anxiety, fear, sadness, anger, guilt, disgust, irritability, loneliness, worry, self-consciousness, dissatisfaction, hostility, embarrassment, reduced self-confidence, and feelings of vulnerability, in reaction to various types of stress, and to place themselves into situations that foster negative affect (Jeronimus et al. 2014; Specht et al. 2011). The individual differences in neuroticism scores show moderate to high stability across much of the adult life course (Wray et al. 2007).

Mounting evidence suggests that neuroticism has profound significance for mental health (Lahey 2009; Ormel et al. 2013). Specifically, people who score higher in the neuroticism dimension have a greater propensity for psychiatric conditions, such as anxiety, somatoform, obsessive–compulsive disorder (OCD), as well as schizophrenia (SZ), bipolar disorder (BD), major depression disorder (MDD), and attention-deficit hyperactivity disorder (ADHD) (Kendler and Myers 2010; Kotov et al. 2010; Lonnqvist et al. 2009; Michielsen et al. 2014; Schirmbeck et al. 2015; Van Os and Jones 2001). Furthermore, several studies linked higher levels of neuroticism to an increase in all-cause mortality (Gale et al. 2016). Neuroticism also contributes to other phenomena that correlate strongly with psychological distress, for example, persistent low subjective well-being (SWB), and physical health problems (Heller et al. 2004; Okbay et al. 2016). Previous studies had quantified neuroticism’s cross-sectional association with common mental disorders (CMDs), but also revealed the strong predictive effect of neuroticism with CMDs in prospective association (Kotov et al. 2010; Lahey 2009; Lonnqvist et al. 2009; Van Os and Jones 2001). Discovered connections between neuroticism and CMDs have promoted the assumption that neuroticism is an independent, etiologically informative risk factor, which may actively contribute to the progress of CMDs.

A vital role of neuroticism in mental health has long been documented, but whether the association of neuroticism with psychopathology is determined by inherited genetic variants remains elusive. Previous studies indicated that the possible explanation for associations between neuroticism and these various mental disorders may be due to shared genetic components. Genetic correlations have been identified between neuroticism and anorexia nervosa, MDD, and SZ, and attributed to the genes with pleiotropic effects (Gale et al. 2016; Hill et al. 2019). In addition, some of the subfactors of neuroticism were also genetically associated with BD, ADHD, and autism spectrum disorder (ASD) (Hill et al. 2019). Furthermore, previous studies reported moderate to high genetic correlations between neuroticism and obsessive–compulsive symptoms (Bergin et al. 2014; Taylor et al. 2011). Twin studies have shown considerable overlaps between the genetic factors of neuroticism and those conferring risk for internalizing disorders, such as anxiety and depression (Hettema et al. 2006).

Recently, genome-wide association studies (GWAS) have begun to map the loci that are associated with neuroticism (Luciano et al. 2018; Smith et al. 2016), as well as ones contributing both to neuroticism and psychiatric disorders (Barlow et al. 2014; Lo et al. 2017; Smeland et al. 2017). Current evidence strongly supports the importance of neuroticism in the genetic etiology of mental disorders. However, causal relationships between neuroticism and a wide range of human health outcomes, especially physical conditions, remain largely unknown.

Mendelian randomization (MR) is an analytic technique that uses genetic variants as instrumental variables to test for causative association between an exposure and an outcome (Lawlor et al. 2008). MR is becoming increasingly efficient and cost-effective when applied to ever-growing data curated from recent GWAS. Recently, the GSMR (Generalized Summary-data-based MR) had been developed by leveraging power from multiple genetic variants accounting for linkage disequilibrium (LD) between the variants (Zhu et al. 2018).

Here, we hypothesize that genetic liability of neuroticism may confer risk to mental health and some kinds of physical diseases. In the present study, we used LD score regression to trace the genetic correlations of neuroticism to a panel of mental and physical conditions. We further used the GSMR to perform a multi-SNP MR analysis on summary results presented in GWAS datasets to test the causal associations between neuroticism and these conditions. Finally, we curated and compared genome-wide risk genes for neuroticism and 13 traits causally influenced by neuroticism. For further dissection, we have concentrated on the connection of neuroticism to MDD and cardiovascular diseases (CVDs), two major causes of morbidity in world populations.

Materials and methods

Datasets and quality control

Summary results of GWAS for 17 mental and 18 physical conditions were included in the analyses (Table 1). The mental conditions were selected based on each condition’s frequency and importance, and the physical conditions were selected based on each condition’s significance within human non-communicative chronic disorders. The trait psychiatric disorders included the collection of psychological/psychiatric problem, mania/bipolar disorder/manic depression, schizophrenia, anxiety/panic attacks. The cardiovascular disease (CVD) included a panel of cardiovascular conditions recruited by the UKB (Supplementary File) (Sudlow et al. 2015). In this study, we used CVD to specifically denote the collection of cardiovascular conditions used the UKB dataset, and used CVDs to denote all cardiovascular conditions. All participants were of European origins. Per condition sample sizes have ranged from 7556 to 694,649. Detailed information of these included studies is summarized in Table 1. We compared SNP alleles between the neuroticism dataset and each of the other datasets and performed the following quality control steps: (1) SNPs were fileted based on INFO ≥ 0.90 if it exists; (2) each SNP was compared between the two datasets and SNPs with conflicting alleles between each pair of datasets were excluded; (3) the alleles and effects of the datasets were harmonized. For palindromic SNPs, we employed allele frequency (FRQ) filtering to determine whether the alleles from the two GWASs were derived from the same reference panel or not (SNPs with FRQ difference between the two GWASs ≥ 0.20 were deemed as a mapping from opposite reference panels).

Table 1 Summary information of the datasets
Full size table

Genetic correlation analysis

GWAS summary results were utilized to analyze the genetic correlations of neuroticism with the 34 health conditions using LD score regression (Bulik-Sullivan et al. 2015a; b). The 1000 Genome project phase 3 (Genomes Project et al. 2015) was used to estimate the LD structure for European populations, which was obtained from the LD score regression website (Bulik-Sullivan et al. 2015a, b; Finucane et al. 2015). SNPs were filtered by 1.1 million variants, a subset of 1000 Genomes and HapMap3, with MAF above 0.05. Adjustment for multiple tests was achieved by calculating the false discovery rate (FDR).

Bidirectional MR analysis

Bidirectional causal associations between neuroticism and the 34 traits were inferred using GSMR (Zhu et al. 2018), and FDR was calculated for the 68 tests. This method utilizes summary-level data to test for putative causal associations between a risk factor (exposure) and an outcome using independent genome-wide significant SNPs as instrumental variables as an index of the exposure. Instrumental variants were selected based on default P ≤ 5 × 10–8 and further pruned using a clumping r2 cutoff of 0.01. When the threshold was surpassed by less than 10 SNPs, a P value threshold of 1 × 10–5 was used; when more than 10,000 SNPs surpassed the default threshold, further analysis was limited to the top 10,000 most significant SNPs. HEIDI outlier detection was used to filter genetic instruments that showed clear pleiotropic effects on the exposure phenotype and the outcome phenotype. We used a P value threshold of 0.01 for the outlier detection analysis in HEIDI, which removes 1% of SNPs by chance if there is no pleiotropic effect.

The method estimates a putative causal effect of the exposure on the outcome (bxy) as a function of the relationship between the SNP’s effects on the exposure (bzx) and the SNP’s effects on the outcome (bzy), given the assumption that the effect of non-pleiotropic SNPs on an exposure (x) should be related to their effect on the outcome (y) in an independent sample only via mediation through the phenotypic causal pathway (bxy). The estimated causal effect coefficients (bxy) are approximately equal to the natural log odds ratio for a case–control trait. An odds ratio of 2 can be interpreted as a doubled risk compared with the population prevalence of a binary trait for every standard deviation increase in the exposure trait. Criteria for significant causal association included both the absolute value of effect size bxy ≥ 0.10 and FDR < 0.05.

To explore the mechanisms underlying the causal associations between neuroticism and MDD, CVD, and HD, we mapped the instrumental SNPs into their host genes. FUMA was used to map SNPs to genes and identify LD-independent genomic regions (Watanabe et al. 2017). All genes located within 10 kb vicinity of each variant were mapped. Therefore, these mapped genes contributed to the causal effects between neuroticism and MDD, CVD, and HD.

Curation of GWAS genes and gene overlapping analysis

We obtained GWAS results (including meta-analysis) of the 14 traits with a causal association with neuroticism from the GWAS Catalog database (https://www.ebi.ac.uk/gwas/) (Buniello et al. 2019). Quality control steps: (1) studies with sample size of ≥ 2000 were retained; (2) studies with mixed phenotypes were removed; (3) genes in MHC regions were removed due to the high LD within the MHC region, which may inflate the gene overlapping analysis; (4) pharmacogenetic studies were excluded. The results of quality control steps are listed in Supplementary Table 1.

Analysis of overlaps between gene sets of each trait with those of neuroticism was conducted using R package SuperExactTest (Wang et al. 2015), with the total gene number in the genome being set as 30,000.

All the statistical analyses were conducted in R 3.6.1 or Python 3.7 environment. A more detailed description of the methods is provided in the Supplementary File.

Results

Genetic correlations

Genetic correlation analyses indicated that neuroticism has a significant genetic correlation with all the 16 mental conditions, and with 8 out of the 18 physical conditions studied (Supplementary Table 2, Fig. 1a). In particular, a relatively high genetic correlation was detected between neuroticism and SWB (rg = − 0.68) and MDD (0.67), followed by anxiety disorder, loneliness, psychiatric disorders, insomnia, alcohol dependence, and OCD (rg in range of 0.31–0.54). The average absolute rg for correlations between neuroticism and mental conditions was 0.34 ± 0.18, while the average absolute rg between neuroticism and physical conditions was 0.11 ± 0.05.

Fig. 1
figure 1

Relationships of neuroticism to each of correlating traits. AD Alzheimer’s disease, ADHD attention-deficit/hyperactivity disorder, ASD autism spectrum disorder, BD bipolar disorder, BMI waist-to-hip ratio adjusted for body mass index, CAD coronary artery disease, CVD cardiovascular disease, HAC hernia abdominopelvic cavity, IBD inflammatory bowel disease, MDD major depressive disorder, OCD obsessive–compulsive disorder, PTSD posttraumatic stress disorder, PVD peripheral vascular disease, SWB subjective well-being, SZ schizophrenia, T2D type 2 diabetes, TS Tourette syndrome, HAC hernia abdominopelvic cavity, PVD peripheral vascular disease, HD hypertensive disease, PD psychiatric disorders. a Genetic correlation of neuroticism with all traits assessed. Red bars indicate FDR ≤ 0.05, and blue bars indicate FDR > 0.05; the lines in each bar indicate standard error. b Genetic correlation coefficient and Mendelian randomization effect size for each trait. c Mendelian randomization analysis of neuroticism with all traits assessed. Arrowed lines show a causal effect from a source node to a targeted node; red lines show positive coefficient, and blue lines show negative coefficient; the thickness of lines is relative to the absolute value of the coefficient. d Pyramid depicting the hierarchy of the relationship between all traits connected to neuroticism

Full size image

MR analysis

In the MR analysis, neuroticism displayed a causal effect on 12 out the 16 mental conditions (MDD, insomnia, SWB, SZ, ADHD, alcohol dependence, loneliness, anorexia nervosa, anxiety disorder, BD, OCD, and psychiatric disorders) and two physical conditions (hypertensive disease (HD) and CVD) (Table 2 and Fig. 1c). Conversely, MDD, insomnia, and reduced SWB displayed a causal effect on neuroticism. The SNPs and genes contributing to the causal effect between neuroticism and MDD are listed in Supplementary Tables 3–4. The SNPs and genes contributing to the causal effect of neuroticism on CVD and HD are listed in Supplementary Tables 5–6. Figure 2 depicts outcomes of specific analysis which uncovered genes mediating the causal effects between neuroticism and MDD, CVD, and HD. The genetic correlation coefficient and MR effect size for each trait are shown in Fig. 1b.

Table 2 Mendelian randomization analysis of neuroticism with all traits
Full size table
Fig. 2
figure 2

Key genes mediating the causal effects between neuroticism and MDD, CVD, and HD. MDD major depressive disorder, CVD cardiovascular disease, HD hypertensive disease

Full size image

The mental and physical traits can be grouped into four classes based on their relationship with neuroticism (see Fig. 1d). At the first level, traits do not have any genetic relationship with neuroticism, while the traits residing at the second level and the neuroticism share some genetic background. Traits at the third level both share a genetic background with neuroticism, and are causally influenced by neuroticism. Fourth-level traits (at the top) display both high genetic correlations and mutually causal relationships with neuroticism.

Gene overlaps

GWAS results of the 14 traits are listed in Supplementary Tables 7–20. Except for OCD and anxiety disorder, each trait showed significant over-representation of overlapped genes with neuroticism (Supplementary Table 21). Especially, 304 out of 590 GWAS risk genes for SWB (significant overlapping, P = 0; FDR = 0) and 218 out of 791 GWAS risk genes for MDD (significant overlapping, P = 1.47 × 10–166; FDR = 1.03 × 10–165) were overlapped with those for neuroticism.

Discussion

The profound significance of neuroticism for public health is evidenced by the fact that the estimated economic costs of neuroticism exceed those of common mental disorders (Cuijpers et al. 2010). Higher scores for neuroticism are associated with an increased risk for various mental disorders, and, somewhat less consistently, with certain aspects of physical health. A well-known case is the type A behavior pattern (TABP) as a coronary risk factor. An overall association between the TABP and CHD has been well established (Smith and MacKenzie 2006). The prevailing model to explain this association is that type A individuals show higher levels of cardiovascular and neuroendocrine responses to stressors compared with their more relaxed Type B counterparts. Type D (i.e., distressed) personality is another kind of personality associated with CHD (Denollet 2005; Denollet et al. 1996). One of the two factors of this construct—negative affectivity—corresponds closely to neuroticism (Denollet 2005).

The evidence supporting the association of neuroticism with mental health and cardiovascular risk is convincing, the emphasis at present is to determine the mechanisms involved. The concept that personality influences health seems straightforward, but the explication of mechanisms underlying mind–body associations is conceptually complex and methodologically challenging due to inconsistent findings and alternative interpretations (Smith and MacKenzie 2006).

We outline the models of mechanisms for neuroticism into two broad categories, physiological reactions to stressors and health-related behavior alterations. Within the psychological reaction category, neuroticism influences the appraisal of potentially stressful life circumstances and coping responses and leads to various negative affections, including depression, anxiety, hostility, and anger, which are also risk factors for CVDs (Barth et al. 2004; Chida and Steptoe 2009). Physiological mechanisms underlying psychological reactions include neuroendocrine responses, inflammation, autonomic function (Schneiderman et al. 2005). Another category concerns patterns of behavioral changes. Personality influences health-relevant daily habits (e.g., smoking, diet, and exercise) relate to negative health behaviors, which could mediate the association between personality and disease (Booth-Kewley and Vickers 1994). Higher neuroticism could lead to unhealthy behaviors such as poor diet, smoking, sleep disturbances, physical inactivity, or lower treatment adherence (Chida and Steptoe 2009). Neuroticism has higher comorbidity with alcohol and drug dependence, conduct disorders, and antisocial personality (Khan et al. 2005). Of note is that the two aspects are not independent and mutually exclusive, but rather act together to link neuroticism to elevated cardiovascular risk.

Our results indicate that neuroticism may confer risk to mental health and CVDs through genetic variation, therefore, reveal a novel mechanism by which neuroticism influences the risk for mental disorders and CVDs. It is a more fundamental way compared with theories derived from psychological effects and health behaviors. The evidence derived from genetic variants as instrumental variables indicates that the mediating effects are most likely etiologically causal in nature compared with other mechanisms. Undoubtedly, the psychological, behavioral, and genetic mechanisms and pathways operate in a synergistic and integrative way to promote cardiovascular changes and related clinical manifestations.

Recently, genetic correlations of neuroticism with some health conditions have been reported. Gale et al. examined genetic correlations of neuroticism with 9 mental traits and 8 physical health-related outcomes, and reported significant positive associations between neuroticism and MDD, anorexia nervosa, and SZ (Gale et al. 2016). Nagel et al. investigated 9 mental traits and 26 other traits, and detected significant positive genetic correlations between neuroticism and 11 health characteristics, including MDD, ADHD, anorexia nervosa, SZ, intracranial volume, and significant negative genetic correlations between neuroticism and SWB, height, age of first child, intelligence quotient (IQ), and educational attainment (Nagel et al. 2018). Using GWAS results from the UK Biobank (UKB), Hill et al. (2019) found that neuroticism significantly correlates with 20 out of the 32 studied conditions; the resultant list of correlates included SZ, MDD, anorexia nervosa, ADHD, and SWB (negatively).

In our study, we found that neuroticism shows a correlation with all mental conditions studied and nearly half of physical conditions and characteristics, with relatively higher correlations between neuroticism and MDD, SWB (negatively), anxiety, loneliness, and insomnia. In addition to replicating previously reported correlations of neuroticism with SZ, anorexia nervosa, AD, and ADHD, our study detected novel genetic connections between neuroticism and loneliness, insomnia, alcohol dependence, OCD, and TS. For physical conditions, our study identified genetic correlations of neuroticism with asthma, CVD, dyslipidemia, osteoarthritis, coronary artery disease, body mass index (BMI), inflammatory bowel disease, HD, and type 2 diabetes. Predictably, the average correlation of neuroticism with mental conditions was stronger than that observed for neuroticism and physical conditions.

Even if previous studies have detected correlations between neuroticism and health-related traits, the causality of these relationships has rarely been tested. Nagel et al. reported a significant causal effect of neuroticism on depression, anxiety disorder, ADHD, SWB, SZ, IQ, and educational attainment, as well as a reverse causal effect of some traits on neuroticism, including depression, ADHD, SWB, SZ, childhood obesity, intracranial volume, height, age of first child, IQ, and educational attainment.

Presented results suggest that neuroticism has a close genetic relationship with all major mental conditions, and in most cases, these relationships are causal. We augmented the study of Nagel et al. by identifying additional traits genetically influenced by neuroticism, including insomnia, alcohol dependence, loneliness, anorexia nervosa, BD, and OCD. Thus, etiological evidence for the previous observation that high scores for neuroticism predate the onset and development of any CMD, and that their effects rarely decline over time (Jeronimus et al. 2016) is provided. Furthermore, our study shows a causal relationship between neuroticism and CVDs, pointing at the strength of the mind–body relationship underlining at least some physical conditions. In this light, it is important to note a handful of genes and protein products recently shown to contribute both to psychological traits, including neuroticism, and physical diseases. The causal role of neuroticism in the development of a physical disease may inform a pathway to truly personalized, prevention-driven medicine.

In agreement with the above results, we observed significant enrichment of overlapping genes between neuroticism and the majority of the 14 traits tested. Especially, half of GWAS risk genes of SWB and more than a quarter of risk genes of MDD were overlapped with those of neuroticism.

Given together, neuroticism showed the closest relationship with MDD and SWB, as evidenced by strong genetic correlation, large bidirectional causal effect, and striking enrichment of overlapping genes between neuroticism and the above two mentioned characteristics. This might probably attribute to the shared genetic factors, and these shared genetic factors have vertical genetics pleiotropy, which may cause some related intermedia phenotypes to be causally correlated.

A strong association between neuroticism and an increased risk of MDD has been previously reported in two large population cohorts (Navrady et al. 2017), consistent with previous cross-sectional observations that neuroticism is most strongly associated with depressive disorders (Kotov et al. 2010). However, molecular mechanisms underlying the associations of neuroticism and MDD remain elusive. In this study, we revealed a panel of genes contributing to both the causal effect of neuroticism on MDD and those contributing to the causal effect of MDD on neuroticism, including DCC, LINC00461, DKFZP686K1684, LOC105377703, RBFOX1, RERE, SOX5, and TCF4. Each of these genes has been previously identified as a genome-wide gene for each of the two traits. Therefore, these genes are plausible key contributors to the bidirectional causal associations of the two conditions.

The transcription factor 4 encoding gene, TCF4, was one of the first genome-wide risk loci for SZ (Stefansson et al. 2009). Later, in some works, this gene was implicated in MDD (Wray et al. 2018), insomnia (Jansen et al. 2019), Pitt–Hopkins syndrome, and other neurodevelopmental disorders (Forrest et al. 2014). It plays a role in lymphoid tissue development and epithelial–mesenchymal transition (Forrest et al. 2014). DCC plays a key role in axon guidance and nerve regeneration (Finci et al. 2015), and is a genome-wide risk gene for MDD (Wray et al. 2018), neuroticism (Nagel et al. 2018; Okbay et al. 2016), intelligence (Savage et al. 2018), cognitive ability (Lee et al. 2018), and educational attainment (Lee et al. 2018). LINC00461 locus encodes one of the most highly conserved lncRNA, which is predominantly expressed in the brain (Deguchi et al. 2017). Previous GWASs have implicated it in MDD (Nagel et al. 2018) and neuroticism (Baselmans et al. 2019; Luciano et al. 2018; Nagel et al. 2018). RBFOX1 regulates tissue-specific alternative splicing, and have being implicated in neurodevelopmental and neuropsychiatric disorders, including MDD (Hyde et al. 2016; Wray et al. 2018), ASD (Grove et al. 2019), and neuroticism (Baselmans et al. 2019; Luciano et al. 2018; Okbay et al. 2016). DKFZP686K1684 and LOC105377703 are two ncRNAs with unknown functions. The function of RERE is discussed later.

SWB is an indicator of happiness or satisfaction. The highest negative correlation between neuroticism and SWB observed in our study was similar to the results of two previous studies (Hill et al. 2019; Nagel et al. 2018), which is not surprising as all three studies, including the present one, utilized the same dataset for SWB (Okbay et al. 2016). The high negative correlation between neuroticism and SWB and the negative causal effect of neuroticism on SWB points towards an overall detrimental effect of neuroticism on mental health.

In addition to its association with mental and physical diseases, neuroticism also influences one’s social life. This effect is evident from a high genetic correlation between neuroticism and loneliness and an extremely high causal effect of neuroticism on loneliness (bxy = 0.74). These results are consistent with the previously reported genetic correlation between neuroticism and social deprivation (Hill et al. 2019), suggesting an extended role of neuroticism in one’s social behavior and activity.

Personality is commonly seen as an inherited trait that remains stable across the lifespan. Due to the general belief that neuroticism is a risk factor for various mental disorders and distress, previous work heavily emphasized the effect of neuroticism on various medical conditions. In contrast, our present work, along with the study of Nagel et al., indicates that the scores for neuroticism may be affected by an individual’s mental status, too, stressing on bidirectionality of the relationship between neuroticism and mental disorders. The discovery of neuroticism-promoting influence on insomnia and reduced SWB is of clinical relevance. In the case of insomnia, the discovery of its bidirectional causal connections to neuroticism may bring hope for the patients seeking personality improvement, since the sleeping problems are more amenable to modification than the neuroticism itself. The negative genetic correlation and the bidirectional negative causal connections between neuroticism and SWB indicate that either mutually detrimental effects or reciprocally inhibiting mechanisms shared between these conditions are at play. It is tempting to speculate that an individual’s neuroticism may be improved by increasing one’s feeling of happiness, which can be achieved by social interventions or in course of psychotherapy. Together, our results point towards a potential avenue for possible amelioration of maladaptive personality traits, and therefore, for decreasing the risks for the relevant mental disorders by improving the quality of sleep, and/or by enhancing the feeling of happiness or satisfaction.

For physical conditions, neuroticism may genetically exert an effect on CVD and HD. CVDs account for a large proportion of the total disability and morbidity (DALYs and Collaborators 2018). Hypertension is a major risk factor for CVDs, contributing to a large proportion of cardiovascular deaths (Lim et al. 2012). Neuroticism has been reported to be correlated with stroke and increased coronary heart disease mortality (Jokela et al. 2014), but biological mechanisms underlying their associations are poorly understood. Our findings may have implications in the prevention and personalized treatment of CVDs. The American Heart Association (AHA) advocated screening depression in all patients with CADs (Lichtman et al. 2008). Our results suggest that it may be useful to screen neuroticism in patients with CVDs, since improved treatment may be reached by including psychological or psychiatric interventions for subgroups with higher neuroticism.

Here, we explored critical biological molecules and potential mechanisms contributing to the links between neuroticism and CVDs. Of particular interest are MAD1L1, ARNTL, RERE, LINC01122, and SOX6. All these genes were found both to mediate the causal effect of neuroticism to CVD, and to confer genome-wide risks for the two conditions. None of these genes is a novice in the field of neurodevelopment. The arginine-glutamic acid dipeptide (RE) repeats gene (RERE) is located in the proximal 1p36 region critical for neurodevelopment (Zoltewicz et al. 2004). Heterozygous variants in the RERE gene have been reported to be associated with a neurodevelopmental disorder with or without anomalies of the brain, eye, or heart (NEDBEH) (Jordan et al. 2018). In addition, RERE is also one of the candidate genes for CHARGE syndrome (Issekutz et al. 2005) and 1p36 deletion syndrome (Zaveri et al. 2014), sharing the features of the developmental delay, intellectual disability, congenital heart disease, cardiovascular malformations, and cardiomyopathy (Fregeau et al. 2016; Issekutz et al. 2005; Jordan et al. 2018; Zaveri et al. 2014). Mouse models implicate RERE in cerebellar foliation and the migration and maturation of Purkinje cells (Kim and Scott 2014) as well as in ventricular septal defects (Kim et al. 2018). LINC01122 is a non-coding RNA that was linked to depression (Nagel et al. 2018), insomnia (Jansen et al. 2019), Tourette syndrome (Yu et al. 2019), neuroticism (Nagel et al. 2018), and CVDs (Kichaev et al. 2019). MAD1L1 play a role in cell cycle control and has been implicated in the genetic susceptibility for MDD (Howard et al. 2018), SZ (Ripke et al. 2013), BD (Psychiatric 2011), neuroticism (Luciano et al. 2018; Nagel et al. 2018), and CVDs (Kichaev et al. 2019). The protein encoded by SOX6 is a transcriptional activator that plays a versatile regulatory role in vertebrate development, including normal development of the central nervous system and maintenance of cardiac and skeletal muscle cells (Hagiwara 2011). GWASs implicated it in neuroticism (Baselmans et al. 2019; Luciano et al. 2018; Nagel et al. 2018), hypertension or blood pressure (Ehret et al. 2016; Giri et al. 2019; Kichaev et al. 2019; Lu et al. 2015; Takeuchi et al. 2018), and CVDs (Kichaev et al. 2019; van der Harst and Verweij 2018), while molecular physiology research pointed at its direct contribution to the prevention of the cardiac hypertrophy (Huang et al. 2015). The presence of both neurodevelopmental and cardiovascular phenotypes in the spectrum of the effects pertained to each of these genes is remarkable.

Our study had several strengths. First, we analyzed the data for a larger set of traits, especially for mental traits. Second, for each trait, only well-powered datasets were included, and a threshold of MR effect coefficient was used to select the strongest factors. Furthermore, to avoid potential population heterogeneity across the studies, we limited our analysis to individuals of European ancestry.

However, several limitations should also be noted. First, to focus attention on mental conditions, a relatively small number of physical traits was included in the study, thus, necessitating future exploring of the association of neuroticism with physical conditions in detail. Second, amounts of samples comprising different datasets varied substantially—from a minimum of 7,556 samples for loneliness to 694,649 samples for BMI; this imbalance may result in uneven power across the traits. Finally, as our analysis was limited to a genetic component of each trait, presented results should be interpreted cautiously, with the understanding that human traits result from a complex web of interactions among a plethora of psycho-social-environmental factors.

Conclusions

In summary, our findings reveal genetic relationships between neuroticism and various health conditions and shed light on mechanisms underlying their phenotypic relationships.