Skip to main content
NCBI home page
Phenomics logoLink to Phenomics
. 2022 May 7;2(4):242–253. doi: 10.1007/s43657-022-00052-3

Causal Effect of Genetically Determined Blood Copper Concentrations on Multiple Diseases: A Mendelian Randomization and Phenome-Wide Association Study

Xiuming Feng 1,2,#, Wenjun Yang 1,2,3,#, Lulu Huang 1,4, Hong Cheng 1,2, Xiaoting Ge 1,2,5, Gaohui Zan 1,2, Yanli Tan 1,2, Lili Xiao 2, Chaoqun Liu 6, Xing Chen 7, Zengnan Mo 1,8, Longman Li 1,2,8,✉, Xiaobo Yang 1,2,5,✉
PMCID: PMC9590500  PMID: 36939787

Abstract

Exposures to copper have become a health concern. We aim to explore the broad clinical effects of blood copper concentrations. A total of 376,346 Caucasian subjects were enrolled. We performed a Mendelian randomization and phenome-wide association study (MR-PheWAS) to evaluate the causal association between copper and a wide range of outcomes in UK Biobank, and we constructed a protein–protein interaction network. We found association between blood copper concentrations and five diseases in the overall population and nine diseases in male. MR analysis implicated a causal role of blood copper in five diseases (overall population), including prostate cancer (OR = 0.87, 95% CI 0.77–0.98), malignant and unknown neoplasms of the brain and nervous system (OR = 0.58, 95% CI 0.38–0.89), and hypertension (OR = 0.94, 95% CI 0.90–0.98), essential hypertension (OR = 0.94, 95% CI 0.90–0.98) and cancer of brain and nervous system (OR = 0.63, 95% CI 0.41–0.98). For male, except for dysphagia being newly associated with blood copper (OR = 1.39, 95% CI 1.18–1.63), other MR results were consistent with the overall population. In addition, the PPI network showed possible relationship between blood copper and four outcomes, namely brain cancer, prostate cancer, hypertension, and dysphagia. Blood copper may have causal association with prostate cancer, malignant and unknown neoplasms of the brain and nervous system, hypertension, and dysphagia. Considering that copper is modifiable, exploring whether regulation of copper levels can be used to optimize health outcomes might have public health importance.

Supplementary Information

The online version contains supplementary material available at 10.1007/s43657-022-00052-3.

Keywords: Copper, Disease, Mendelian randomization, Phenome-wide association study

Introduction

Copper (Cu) is an essential trace mineral for human physiology. As a cofactor for various enzymes, Cu is critically involved in respiration, activation of neuroendocrine peptides, pigmentation, catecholamine synthesis and clearance, free radical defense and enzymes involved in many other cellular processes (Czlonkowska et al. 2018). Normal dietary consumption and absorption of Cu are mainly through foods (e.g. legumes, potatoes, nuts and seeds, chocolate, beef, organ meat and shellfish) and drinking water (Ma and Betts 2000).

In the case of Cu, it is likely to be U-shaped dose–response association, which means that both deficiency and excess can produce adverse health effects. Currently, epidemiological and experimental evidence have shown that Cu status was associated with multisystem diseases. It is known that extrahepatic Cu toxicity caused damage to nerve tissue further led to neurologic symptoms of Wilson disease (Czlonkowska et al. 2018). Cu levels were associated with mostly aberrant cognitive functions, greater regional gray matter volume in extensive areas, greater mean diffusivity in subcortical structures in 924 healthy young adults (Takeuchi et al. 2019). Results of a meta-analysis enrolled 15,274 cardiovascular disease outcomes in aggregate indicated that exposure to Cu was associated with an increased risk of cardiovascular disease and coronary heart disease (Chowdhury et al. 2018). A case–control study suggested that environmental exposure to Cu might be related with the prevalence of hypertension (Wu et al. 2018). Transgenic Adenocarcinoma of the Mouse Prostate model demonstrated that increasing intracellular bioavailable Cu can selectively kill cancerous prostate cells in vitro and in vivo (Cater et al. 2013).

Epidemiological studies have explored the association between Cu and a wide range of diseases. However, there were many potential confounding factors could affect the casual estimation between blood Cu concentrations and diseases. Recently, mendelian randomization and phenome-wide association study (MR-PheWAS) analysis have been proposed to assess causal relationships between exposure and outcomes (Denny et al. 2013; Li et al. 2018). For example, the MR-PheWAS of higher serum magnesium (Mg) discovered that Mg had a causal relationship with five disease groups and six disease outcomes (Li et al. 2021). And genetically determined iron status played a protective role on hypercholesterolemia and a detrimental role on risk of skin and skin structure infections (Gill et al. 2019).

Our study applied the MR-PheWAS method to find the association between a wide range of disease outcomes and Cu in the UK Biobank. And we further build a PPI network to validate the potential association between Cu and diseases.

Materials and Methods

Study Population

This study was based on data obtained from a prospective cohort study, UK Biobank. At baseline, a total of 502,505 participants aged 40–69 years were recruited from 22 assessment centers throughout the UK Biobank between 2006 and 2010 (Sudlow et al. 2015). The cohort provided a wide range of self-reported baseline information, including the data of health, lifestyle, disease outcomes and genome-wide genotyping (Bycroft et al. 2018). Our study was restricted to participants of Caucasian descent, and one of the relatives who have a kinship coefficient of  more than 0.0884 was randomly excluded.

Genetic Instruments for Cu

The genome-wide association study (GWAS) performed by the Queensland Institute of Medical Research has found that two Cu-associated single nucleotide polymorphisms (SNPs) on European subjects (N = 2,603), which were used in our analysis (Evans et al. 2013). These two identified SNPs were: rs1175550 in small integral membrane protein 1 (SMIM1) gene and rs2769264 in selenium binding protein 1 (SELENBP1) gene, with p < 5E−8, not in linkage equilibrium and F-statistics > 10. Two Cu-associated SNPs collectively explain approximately 3.99% of the variation in blood Cu (Supplementary Table 1) (Guo et al. 2020; Kodali et al. 2018).

PheWAS

The case and control groups were defined using the 10th WHO's International Classification of Diseases (ICD) coding systems. The study was limited to the incident and prevalent instances (both inpatient hospital episode records and cancer registry data), except for self-reported diagnosis. And we employed two Cu-associated SNPs independently in the PheWAS analysis (Denny et al. 2013). The PheWAS methods identified subjects who have at least one event as the cases, and those without the outcome or related phecodes were excluded in the control group (Li et al. 2018). Then a logistic regression analysis was generated to estimate the association between Cu-related SNPs with all Phecodes. The covariates considered adjusted were age, sex, body mass index (BMI), and the first 15 genetic principal components. And we only considered cases with 200 or more in the PheWAS analysis. In addition, 10% false discovery rate (FDR) was used as the threshold (Verma et al. 2018).

MR, Pleiotropy and Sensitivity Analyses

Due to two Cu-associated SNPs in the GWAS analysis, the inverse-variance weighted (IVW) method was applied in the MR estimates (Burgess et al. 2015; Palmer et al. 2012), with Cochran’s Q statistic to evaluate heterogeneity (Burgess et al. 2017; Egger et al. 1997; Huang et al. 2019). In order to exclude possible pleiotropic effects, we removed SNPs from the MR analyses through searching the Ensemble database (http://grch37.ensembl.org/Homo_sapiens/Info/Index). In sensitivity analyses, we performed PheWAS and MR analysis by gender stratification to account for any difference between male and female. And we further validated the MR results in MR-base platform (http://app.mrbase.org/) through uploading the exposure data of Cu-SNPs. In addition, we conducted bi-directional MR in MR base-platform to rule out the reverse causation.

The Protein–Protein Interaction (PPI) Network

To reveal the interactions of blood Cu-related genes and disease-related genes, we used the instrumental variable (IV) related genes of blood Cu, the top ten most related genes of blood Cu and outcomes from GeneCards (Stelzer et al. 2016) to model network. For visualizing the PPI network, we used the STRING (Szklarczyk et al. 2019) and Cytoscape (Shannon et al. 2003) software to analyze the interaction between genes.

All statistical analyses were conducted using the R software (version 4.0.2; The R Foundation for Statistical Computing, Vienna, Austria).

Results

The Demographic Characteristics

After filtering out those not meeting the inclusion criteria, a total of 376,346 UK Biobank participants were included, consisting of 202,177 females (53.72%) and 174,169 males (46.28%). The mean age was 57.96 (SD 7.95) years old and BMI was 27.41 (4.75) kg/m2. Drinking status was divided into never drinking, previous drinking and current drinking, accounting for 3.08%, 3.41% and 93.51%, respectively. Smoking status was classified as never smoking (54.62%), previous smoking (35.27%) and current smoking (10.12%). The demographic data are shown in Table 1.

Table 1.

Demographic characteristics of the sampled UK Biobank participants

Characteristic Mean ± SD/n (%)
Age (years) 57.96 ± 7.95
BMI 27.41 (4.75)
Sex
 Female 202,177 (53.72)
 Male 174,169 (46.28)
Drinking status
 Never 11,595 (3.08)
 Previous 12,834 (3.41)
 Current 351,917 (93.51)
Smoking status
 Never 205,541 (54.61)
 Previous 132,721 (35.27)
 Current 38,084 (10.12)
Open in a new tab

BMI, body mass index; SD, standard deviation

PheWAS

After mapping ICD codes to phecodes, there were 17 diseases categories and 576 unique diseases phecode (including cases ≥ 200) in Table 2. And we computed minimum, median, mean and maximum cases in 17 diseases systems (cases range from 201 to 84,148). In the overall PheWAS analysis, five pairs of genotype–phenotype associations were significant after FDR correction with adjustment for covariates (q < 0.10) (Table 3 and Fig. 1). The sex-stratified PheWAS analysis revealed nine pairs of genotype–phenotype associations in males and none in females. When compared with the overall PheWAS analysis, four new pairs of associations were identified from the sex-stratified PheWAS analysis. In general, the PheWAS analyses identified nine unique disease groups/outcomes that shared genetic risk loci with blood Cu levels, which included four disease groups (hypertension, prostate cancer, dysphagia and malignant and unknown neoplasms of the brain and nervous system) and five specific disease outcomes (essential hypertension, decreased white blood cell count, neutropenia, cancer of the brain and nervous system, and brain cancer). All the results of the PheWAS for each instrument SNP are provided in Supplementary Tables 2, 3.

Table 2.

The number of phenotypes and the number of cases in each disease category

Disease categories Phenotypes Cases
Minimum Median Mean Maximum
Circulatory system 58 217 2,463 7,974 84,148
Congenital anomalies 16 203 406 548 1,316
Dermatologic 33 272 1,236 2,148 8,259
Digestive 56 278 3,345 8,484 50,327
Endocrine / metabolic 23 210 1,838 4,344 21,094
Genitourinary 62 218 1,722 3,458 15,577
Hematopoietic 20 213 1,082 2,680 13,121
Infectious diseases 14 261 475 1,380 8,987
Injuries & poisonings 25 229 2,482 3,442 19,157
Mental disorders 24 201 779 1,079 5,472
Musculoskeletal 30 228 1,663 4,917 22,293
Neoplasms 79 212 972 3,004 28,281
Neurological 32 270 848 2,022 12,535
Pregnancy complications 12 208 832 1,033 2,299
Respiratory 38 245 1,936 4,103 27,679
Sense organs 44 214 994 2,287 24,571
Symptoms 10 245 5,300 6,710 15,527
Open in a new tab

Notes: After mapping diagnostic ICD-10 codes to phecodes (the mappings of phecodes is available at http://phewascatalog.org.), 856 different phecodes were summarized and 585 phecodes were included in PheWAS after eliminating disease outcomes with low prevalence (cases < 200)

Table 3.

Genotype–phenotype associations identified from PheWAS analysis

SNPs Phenotypes Description Groups Total (n) Cases (n) β SE OR (95% CI) p Fdr-q
Overall population
 rs1175550 401 Hypertension Circulatory system 376,346 84,148 − 0.026 0.007 0.97 (0.96, 0.99) 2.04 × 10–4 0.064
401.1 Essential hypertension 376,346 83,955 − 0.026 0.007 0.97 (0.96, 0.99) 2.22 × 10–4 0.064
185 Cancer of prostate Neoplasms 131,610 7,671 − 0.066 0.022 0.94 (0.90, 0.98) 0.002 0.077
 rs2769264 191 Malignant and unknown neoplasms of brain and nervous system 131,116 547 − 0.289 0.086 0.75 (0.63, 0.89) 7.93 × 10–4 0.063
191.1 Cancer of brain and nervous system 131,082 513 − 0.273 0.088 0.76 (0.64, 0.90) 0.002 0.080
Male
 rs1175550 401 Hypertension Circulatory system 174,169 45,525 − 0.032 0.010 0.97 (0.95, 0.99) 0.001 0.049
401.1 Essential hypertension 174,169 45,525 − 0.032 0.010 0.97 (0.95, 0.99) 0.001 0.049
185 Cancer of prostate Neoplasms 61,879 7,671 − 0.066 0.022 0.94 (0.90, 0.98) 0.002 0.084
 rs2769264 288.1 Decreased white blood cell count Hematopoietic 174,008 1,531 − 0.151 0.049 0.86 (0.78, 0.95) 0.002 0.064
288.11 Neutropenia 174,008 1,531 − 0.151 0.049 0.86 (0.78, 0.95) 0.002 0.064
532 Dysphagia Digestive 168,809 3,232 0.102 0.032 1.11 (1.04, 1.18) 0.001 0.064
191 Malignant and unknown neoplasms of brain and nervous system Neoplasms 62,264 318 − 0.433 0.119 0.65 (0.51, 0.82) 2.85 × 10–4 0.010
191.1 Cancer of brain and nervous system 62,246 300 − 0.393 0.121 0.68 (0.53, 0.86) 0.001 0.021
191.11 Cancer of brain 62,237 291 − 0.372 0.122 0.69 (0.54, 0.88) 0.002 0.028
Open in a new tab

PheWAS, phenome-wide association study; CI, confidence interval; NOS, not otherwise specified; OR, odds ratio

Notes: The PheWAS analysis was adjusted by age, sex, BMI, and the first 15 genetic principal components

Fig. 1.

Fig. 1

Open in a new tab

Manhattan plot for the phenome-wide association analysis using rs1175550 (a) and rs2769264 (b)

MR

We conducted MR analysis using the IVW method to examine if there was any causal association between genetic blood Cu concentrations and the disease groups/outcomes identified by the PheWAS analysis (Table 4). In overall population, the MR IVW analysis showed that genetically higher blood Cu concentrations were associated with a decreased risk in three groups, including prostate cancer (OR = 0.87, 95% CI 0.77–0.98, p = 0.020), malignant and unknown neoplasms of the brain and nervous system (OR = 0.58, 95% CI 0.38–0.89, p = 0.013), and hypertension (OR = 0.94, 95% CI 0.90–0.98, p = 0.002), and two specific disease outcomes, including cancer of the brain and nervous system (OR = 0.63, 95% CI 0.41–0.98, p = 0.040) and essential hypertension (OR = 0.94, 95% CI 0.90–0.98, p = 0.003). When stratified by sex, blood Cu concentrations was statistically significant among three disease groups, including prostate cancer (OR = 0.87, 95% CI 0.77–0.98, p = 0.020), malignant and unknown neoplasms of the brain and nervous system (OR = 0.33, 95% CI 0.18–0.59, p < 0.001), and dysphagia (OR = 1.39, 95% CI 1.18–1.63, p < 0.001), and two specific disease outcomes, including cancer of the brain and nervous system (OR = 0.35, 95% CI 0.19–0.64, p = 0.001) and brain cancer (OR = 0.37, 95% CI 0.20–0.68, p = 0.001) in males.

Table 4.

Two-sample MR analyses for PheWAS signals and for replication of nominal PheWAS associations

Phecode PheWAS signals Groups MR-IVW Power (%)
β SE OR (95% CI) p pheterogeneity
Overall population
 185 Cancer of prostate Neoplasms − 0.143 0.061 0.87 (0.77, 0.98) 0.020 0.050 60
 191 Malignant and unknown neoplasms of brain and nervous system − 0.544 0.219 0.58 (0.38, 0.89) 0.013 0.023 50
 191.1 Cancer of brain and nervous system − 0.462 0.225 0.63 (0.41, 0.98) 0.040 0.016 39
 191.11 Cancer of brain − 0.455 0.231 0.63 (0.40, 1.00) 0.049 0.048 41
 288 Diseases of white blood cells Hematopoietic − 0.136 0.185 0.87 (0.61, 1.25) 0.462 0.454 10
 288.1 Decreased white blood cell count − 0.112 0.081 0.89 (0.76, 1.05) 0.166 0.134 26
 401 Hypertension Circulatory system − 0.062 0.020 0.94 (0.90, 0.98) 0.002 0.015 87
 401.1 Essential hypertension Circulatory system − 0.060 0.020 0.94 (0.90, 0.98) 0.003 0.015 87
 532 Dysphagia Digestive 0.088 0.056 1.09 (0.98, 1.22) 0.119 0.105 33
Male
 401 Hypertension Circulatory system − 0.051 0.028 0.95 (0.90, 1.00) 0.073 0.007 46
 401.1 Essential hypertension − 0.050 0.028 0.95 (0.90, 1.01) 0.080 0.007 46
 288.1 Decreased white blood cell count Hematopoietic − 0.182 0.127 0.83 (0.65, 1.07) 0.152 0.001 26
 288.11 Neutropenia − 0.182 0.127 0.83 (0.65, 1.07) 0.152 0.001 26
 532 Dysphagia Digestive 0.327 0.084 1.39 (1.18, 1.63) 0.90 × 10–4 0.945 99
 185 Cancer of prostate Neoplasms − 0.143 0.061 0.87 (0.77, 0.98) 0.020 0.050 58
 191 Malignant and unknown neoplasms of brain and nervous system − 1.116 0.302 0.33 (0.18, 0.59)  < 0.001 0.251 67
 191.1 Cancer of brain and nervous system − 1.049 0.308 0.35 (0.19, 0.64) 0.001 0.380 62
 191.11 Cancer of brain − 0.993 0.311 0.37 (0.20, 0.68) 0.001 0.410 58
Open in a new tab

MR, Mendelian randomization; PheWAS, phenome-wide association study; IVW, inverse variance weighted; OR, odds ratio

Sensitivity Analysis

We also analyzed causal relationships between blood Cu levels and the disease groups/outcomes, which were borderline significant (p < 0.01, FDR-q > 0.10) in the total population PheWAS analysis (Supplementary Table 4). The MR IVW analysis showed a causal role in four disease groups, including cancer of other male genital organs (OR = 0.64, 95% CI 0.43–0.98, p = 0.039), benign neoplasm of the lip/oral cavity/pharynx (OR = 1.45, 95% CI 1.08–1.94, p = 0.013), asthma (OR = 0.94, 95% CI 0.89–1.00, p = 0.046), and corns and callosities (OR = 1.74, 95% CI 1.05–2.87, p = 0.031), and two disease outcomes, including occlusion of the cerebral arteries (OR = 0.80, 95% CI 0.70–0.93, p = 0.003) and congenital deformities of the feet (OR = 2.08, 95% CI 1.14–3.79, p = 0.017). We further validated the MR results in an MR-based platform and only found GWAS summary data of prostate cancer and hypertension (Supplementary Table 5). We found no significant association of prostate cancer and hypertension in the MR-based platform. And we conducted sensitivity analysis to examine the causal association between erythrocyte Cu and blood pressure measurements on MR-base platform. And we also found erythrocyte Cu was negatively associated with DBP, which is consistent with the result of hypertension. However, there is no significant between SBP and erythrocyte Cu. Furthermore, the result of bi-directional MR shown in Supplementary Table S6. There is no significant association between diseases and copper.

The PPI Network

Based on MR-PheWAS analysis, we found the IV related genes of blood Cu, the top 10 most related genes of blood Cu and outcomes in the GeneCards website and used these genes to construct a PPI network (Supplementary Table 7). At the genetic level, the network showed the possible association and mechanism between Cu and four outcomes, including brain cancer, prostate cancer, hypertension and dysphagia (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Interactions between disease-related genes and Cu-related genes. We plot the PPI networks of four diseases and Cu. Red points indicates the disease-related genes and blue points indicates Cu-related genes (The dots of circle are the top ten genes most related to blood Cu, the dots of triangle are instrumental variable (IV) related genes of blood Cu). a interactions between cancer of brain and Cu; b interactions between cancer of prostate and Cu; c interactions between hypertension and Cu; d interactions between dysphagia

Discussion

In our analysis, we found evidence of a protective effect of higher blood Cu levels on the risk of hypertension, prostate cancer, and malignant and unknown neoplasms of the brain and nervous system, while a detrimental effect of higher blood Cu levels on the risk of dysphagia in males was observed. Our study also showed a suggestive causal role in increasing the risk of benign neoplasms of the lip/oral cavity/pharynx, corns and callosities, and congenital deformities of the feet and in decreasing the risk of cancer of other male genital organs, occlusion of cerebral arteries, and asthma.

Previous studies have estimated the association between Cu and hypertension, but the results were controversial. Some studies suggested that Cu might increase the risk of hypertension (Akanle et al. 1999; Canatan et al. 2004; Khan et al. 1984; Okoduwa et al. 2015; Olatunbosun et al. 1976; Pavão et al. 2006; Yao et al. 2018), while some studies reported inverse results (Guzel et al. 2018; Liu et al. 2004). However, other studies did not find significant associations (Bergomi et al. 1997; Kedzierska et al. 2005; Li et al. 1999; Taneja and Mandal 2007; Vivoli et al. 1995). We cannot rule out the individual difference, reverse causality or residual confounding. We found causal evidence for a protective effect of Cu on the risk for hypertension. And we validated Cu was negatively related with DBP on the MR-base platform. However, we could not validate the result of hypertension on the MR-base platform. Trace elements play an important role in the prevention of cardiovascular diseases through their antioxidant effects (Kedzierska et al. 2005; McDermott 2000). Cu, as one of the most abundant trace elements in the human body, has antioxidant and pro-oxidant properties. It appears that Cu likely plays a role by affecting the inflammatory response or vascular function (Klotz et al. 2003; Vlad et al. 1994).

Dysphagia, defined as discomfort or difficulty in swallowing, occurs in approximately 18% of patients with Wilson disease and in 50% of patients with neurologic symptoms (Członkowska et al. 2017; da Silva-Júnior et al. 2008; Lorincz 2010). It is known that extrahepatic Cu toxicity caused damage to nerve tissue, further lead to neurologic symptoms of Wilson disease. Cu levels are generally elevated in brain tissues and cerebrospinal fluid and its toxicity contains multiple mechanisms include DNA crosslink, mitochondrial toxicity, oxidative stress, cell membrane damage, and enzymes inhibition. Excess Cu is originally absorbed and buffered by oligodendrocytes and astrocytes, but eventually leads to blood–brain-barrier dysfunction and demyelination. Moreover, the neurologic symptoms of Wilson disease were mostly reversible with anti-Cu therapy (Dusek et al. 2019). These might support our findings. Beyond that, Cu can also cause swallowing difficulties through other non-neurologic approach such as esophagus cancer.

Our study provides causal evidence for a protective effect of blood Cu on the risk for cancer of prostate and malignant and unknown neoplasms of brain and nervous system. Cu is an essential metal for cell division in both normal and cancerous tissues (Chen et al. 2001). Alteration in the equilibrium of the optimum levels of trace elements such as Cu can influence biological approaches and cause various diseases including cancer (Järup 2003). Cu metabolism appears to be altered in human tumor systems, and blood Cu levels were increased in certain types of cancer, such as prostate, brain, colon, lung, and breast cancers (Tisato et al. 2010). Elevated levels of blood Cu have been linked to the severity of breast cancer disease (Finney et al. 2009). However, an experiment showed that increasing intracellular bioavailability of Cu can selectively kill prostate cancer cells in vivo and in vitro, suggesting the potential of bis (thiosemicarbazone) Cu complexes for the treatment of prostate cancer (Cater et al. 2013). Moreover, (Wang et al. 2014) found that Cu is a potential therapeutic transition metal that produces ROS at cytotoxic concentrations in cancer cells. Due to multiple homeostasis mechanisms of Cu import/export from human body, adverse effects may be inhibited in normal tissues (Wang et al. 2014). And we have not found evidence of reverse causation between Cu and hypertension, prostate cancer and dysphagia.

Although no epidemiological studies about the association between Cu and benign neoplasms of the lip/oral cavity/pharynx, corns and callosities, congenital deformities of the feet, cancer of other male genital organs, occlusion of cerebral arteries, and asthma were reported, our study showed suggestive evidence. Further research on these diseases is well worth exploring. Independent cohort studies and functional studies are warranted to verify these relatively novel associations. Previous observational studies have reported neither a gender difference in the association between Cu levels and the development of the above diseases nor have any studies used MR to eliminate the influence of environmental confounders to solve the gender difference. Our study identified a few diseases, including dysphagia and malignant and unknown neoplasms of the brain and nervous system, which were potentially causally associated with the genetic variation of Cu levels in males but not in females. However, hypertension was causally associated with Cu only in females. The biological mechanism by which Cu levels are linked to certain diseases is more pronounced in women than in men remains to be further studied.

Previous study has used the MR-PheWAS method to examine the association of genetic Cu with 12 diseases outcomes, and some MR studies have explored the causal association with the risk of nine chronic diseases, including type 2 diabetes, osteoporosis, rheumatoid arthritis, gout, Parkinson’s disease, Alzheimer's disease, major depressive disorder, bipolar disorder, and schizophrenia (Cheng et al. 2019; Zhou et al. 2020). However, the present study conducted a hypothesis-free investigation of the causal effects of Cu level more broadly across the human phenome, thus contributing to the discovery of new relations. Using SNPs strongly associated with blood Cu level as instrument variants which are randomly assorted at conception, we tested the cumulative lifetime effects of genetically determined variation across more than 600 disease outcomes. One of the main challenges to MR method is deciphering effects that are attributable to bias owing to pleiotropic SNPs, which we addressed by searching online for secondary effects of the variants. Through an online search, we found that rs1175550 for Cu status is nominally associated with red cell and hemoglobin parameters, which may be related to hypertension risk reported by previous studies (Shimizu et al. 2014, 2017). As so far, two studies have explored the association between Cu and diseases using MR-PheWAS analysis (Zhou et al. 2020, 2021), and they discovered that copper was associated with lower risks of iron deficiency anemia and lipid metabolism diseases (lipid metabolism disorders, hyperlipidemia and hypercholesterolemia) and higher risk of osteoarthritis. However, we could not validate their results in our study. The MR-PheWAS of lipid metabolism diseases was limited to phecodes that had enough cases to ensure more than 80% statistical power in the subsequent MR analyses (Zhou et al. 2020). And we considered Phecodes with more than 200 cases, which could not ensure more than 80% statistical power in the subsequent MR analyses. The MR analysis of osteoarthritis, with Phecode diagnosed, only performed logistic regression and MR analysis, not performed PheWAS analysis in a wide range of diseases (Zhou et al. 2021). We applied different quality control in the MR-PheWAS analysis. First, we only considered the PheWAS result which meets the threshold lower than 10% FDR in next MR analysis. And we further considered diseases with p < 0.01 in next MR analysis. Therefore, many potential diseases potentially related with Cu could not be further validated.

In recent decades, various genetic, cell biological, biochemical and statistical methods have been used to study interactions. A PPI network provided an opportunity to understand cellular pathways and explore the mechanisms of the development of diseases via mapping genes and proteins. In our study, we provided clues for further exploring genetic and environmental factors and their interactions. For example, excessive Cu could lead to lower levels of total antioxidant capacity (T-AOC), catalase (CAT) and hydroxyl radicals and increased levels of heat shock proteins (HSPs) and NF-κB pathway-related pro-inflammatory mediators (Nie et al. 2020). These changes further induced inflammatory response and apoptosis in chicken brain.

This study also has limitations. First, we only used genetic data of UK biobank participants, which produced misclassification bias (Padmanabhan et al. 2019). Second, some diseases have lower power (the power of MR analysis was less than 80%) included in our analysis. After adjusting for multiple testing, the previously reported association between Cu and risk of Alzheimer's disease, bipolar disorder and ischemic heart disease risk could not be validated in our study (Cheng et al. 2019; Kodali et al. 2018). Finally, we could not repeat our results on the MR-based platform and validate with other MR-PheWAS analysis of Cu, possibly due to different ethnic backgrounds and quality control measures in the GWAS analysis. Also, the MR-based platform could not guarantee the platform or any GWAS data would be accurate, complete, free from errors or omissions or secure or free from bugs or viruses.

Conclusions

In our analysis, we discover a causal effect of Cu levels with hypertension, dysphagia, prostate cancer, and malignant and unknown neoplasms of the brain and nervous system, which need to be confirmed in further studies. Regulation of blood Cu levels could be beneficial to optimize health outcomes.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (XLSX 126 kb) (126.8KB, xlsx)

Authors' contributions

Conceptualization, data curation and funding acquisition: XY and ZM. Original draft and formal analysis: XF, WY and LH. Methodology: LL and LH. Software: HC and XG. Visualization: GZ and YT. Review and editing: LX and CL. Validation: XC.

Funding

The authors acknowledge the UK Biobank and their participants for contributing the data used in this work (approval number: 56902). This work was supported by the National Key R&D Program of China (grant number 2020YFE0201600); the National Natural Science Foundation of China (grant number 82073504); the Guangxi Natural Science Fund for Innovation Research Team (grant number 2017GXNSFGA198003).

Availability of data and material

The genetic and phenotypic UK Biobank data are available upon application from the UK Biobank (www.ukbiobank.ac.uk/).

Code availability

The code that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Conflicts of interest

On behalf of all authors, the corresponding authors state that there is no conflict of interest.

Ethics approval

UK Biobank received ethical approval from the research ethics committee (reference 13/NW/0382). All participants provided informed consent to participate. The present analyses were conducted under UK Biobank application number 56902.

Consent to participate and to publish

Not applicable.

Footnotes

Xiuming Feng and Wenjun Yang joint first authors.

Contributor Information

Longman Li, Email: lilongman@gxmu.edu.cn.

Xiaobo Yang, Email: yangx@gxmu.edu.cn.

References

  1. Akanle OA, Akintanmide A, Durosinmi MA, Oluwole AF, Spyrou NM. Elemental analysis of blood of Nigerian hypertensive subjects. Biol Trace Elem Res. 1999;71–72:611–616. doi: 10.1007/bf02784250. [DOI] [PubMed] [Google Scholar]
  2. Bergomi M, Rovesti S, Vinceti M, Vivoli R, Caselgrandi E, Vivoli G. Zinc and copper status and blood pressure. J Trace Elements Med Biol Organ Soc Miner Trace Elements (GMS) 1997;11:166–169. doi: 10.1016/s0946-672x(97)80047-8. [DOI] [PubMed] [Google Scholar]
  3. Burgess S, Scott RA, Timpson NJ, Davey Smith G, Thompson SG. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol. 2015;30:543–552. doi: 10.1007/s10654-015-0011-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Burgess S, Bowden J, Fall T, Ingelsson E, Thompson SG. Sensitivity analyses for robust causal inference from mendelian randomization analyses with multiple genetic variants. Epidemiology. 2017;28:30–42. doi: 10.1097/ede.0000000000000559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, Motyer A, Vukcevic D, Delaneau O, et al. The UK Biobank resource with deep phenotyping and genomic data. Nature. 2018;562:203–209. doi: 10.1038/s41586-018-0579-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Canatan H, Bakan I, Akbulut M, Halifeoglu I, Cikim G, Baydas G, Kilic N. Relationship among levels of leptin and zinc, copper, and zinc/copper ratio in plasma of patients with essential hypertension and healthy normotensive subjects. Biol Trace Elem Res. 2004;100:117–123. doi: 10.1385/bter:100:2:117. [DOI] [PubMed] [Google Scholar]
  7. Cater MA, Pearson HB, Wolyniec K, Klaver P, Bilandzic M, Paterson BM, Bush AI, Humbert PO, La Fontaine S, et al. Increasing intracellular bioavailable copper selectively targets prostate cancer cells. ACS Chem Biol. 2013;8:1621–1631. doi: 10.1021/cb400198p. [DOI] [PubMed] [Google Scholar]
  8. Chen F, Vallyathan V, Castranova V, Shi X. Cell apoptosis induced by carcinogenic metals. Mol Cell Biochem. 2001;222:183–188. doi: 10.1023/A:1017970330982. [DOI] [PubMed] [Google Scholar]
  9. Cheng WW, Zhu Q, Zhang HY. Mineral nutrition and the risk of chronic diseases: a mendelian randomization study. Nutrients. 2019 doi: 10.3390/nu11020378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chowdhury R, Ramond A, O'Keeffe LM, Shahzad S, Kunutsor SK, Muka T, Gregson J, Willeit P, Warnakula S, et al. Environmental toxic metal contaminants and risk of cardiovascular disease: systematic review and meta-analysis. BMJ (clin Res Ed) 2018;362:k3310. doi: 10.1136/bmj.k3310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Członkowska A, Litwin T, Chabik G. Wilson disease: neurologic features. Handb Clin Neurol. 2017;142:101–119. doi: 10.1016/b978-0-444-63625-6.00010-0. [DOI] [PubMed] [Google Scholar]
  12. Czlonkowska A, Litwin T, Dusek P, Ferenci P, Lutsenko S, Medici V, Rybakowski JK, Weiss KH, Schilsky ML. Wilson disease. Nat Rev Dis Primers. 2018;4:21. doi: 10.1038/s41572-018-0018-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. da Silva-Júnior FP, Carrasco AE, da Silva Mendes AM, Lopes AJ, Nobre ESMA, de Bruin VM. Swallowing dysfunction in Wilson's disease: a scintigraphic study. Neurogastroenterol Motil. 2008;20:285–290. doi: 10.1111/j.1365-2982.2007.01036.x. [DOI] [PubMed] [Google Scholar]
  14. Denny JC, Bastarache L, Ritchie MD, Carroll RJ, Zink R, Mosley JD, Field JR, Pulley JM, Ramirez AH, et al. Systematic comparison of phenome-wide association study of electronic medical record data and genome-wide association study data. Nat Biotechnol. 2013;31:1102–1110. doi: 10.1038/nbt.2749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dusek P, Litwin T, Członkowska A. Neurologic impairment in Wilson disease. Ann Transl Med. 2019;7:S64. doi: 10.21037/atm.2019.02.43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ (clin Res Ed) 1997;315:629–634. doi: 10.1136/bmj.315.7109.629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Evans DM, Zhu G, Dy V, Heath AC, Madden PA, Kemp JP, McMahon G, St Pourcain B, Timpson NJ, et al. Genome-wide association study identifies loci affecting blood copper, selenium and zinc. Hum Mol Genet. 2013;22:3998–4006. doi: 10.1093/hmg/ddt239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Finney L, Vogt S, Fukai T, Glesne D. Copper and angiogenesis: unravelling a relationship key to cancer progression. Clin Exp Pharmacol Physiol. 2009;36:88–94. doi: 10.1111/j.1440-1681.2008.04969.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gill D, Benyamin B, Moore LSP, Monori G, Zhou A, Koskeridis F, Evangelou E, Laffan M, Walker AP, et al. Associations of genetically determined iron status across the phenome: A mendelian randomization study. PLoS Med. 2019;16:e1002833. doi: 10.1371/journal.pmed.1002833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Guo Y, Lu Y, Jin H. Appraising the role of circulating concentrations of micro-nutrients in epithelial ovarian cancer risk: a Mendelian randomization analysis. Sci Rep. 2020;10:7356. doi: 10.1038/s41598-020-63909-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Guzel S, Guzel EC, Kucukyalcin V, Kiziler AR, Cavusoglu C, Gulyasar T, Cinemre H, Aydemir B. Midkine levels and its relationship with atherosclerotic risk factors in essential hypertensive patients. Niger J Clin Pract. 2018;21:894–900. doi: 10.4103/njcp.njcp_309_17. [DOI] [PubMed] [Google Scholar]
  22. Huang L, Li L, Luo X, Huang S, Hou Q, Ge X, Lv Y, Mo Z, Yang X. The association between serum iron status and risk of asthma: a 2-sample Mendelian randomization study in descendants of Europeans. Am J Clin Nutr. 2019;110:959–968. doi: 10.1093/ajcn/nqz162. [DOI] [PubMed] [Google Scholar]
  23. Järup L. Hazards of heavy metal contamination. Br Med Bull. 2003;68:167–182. doi: 10.1093/bmb/ldg032. [DOI] [PubMed] [Google Scholar]
  24. Kedzierska K, Bober J, Ciechanowski K, Gołembiewska E, Kwiatkowska E, Noceń I, Dołegowska B, Dutkiewicz G, Chlubek D. Copper modifies the activity of sodium-transporting systems in erythrocyte membrane in patients with essential hypertension. Biol Trace Elem Res. 2005;107:21–32. doi: 10.1385/bter:107:1:021. [DOI] [PubMed] [Google Scholar]
  25. Khan SN, Rahman MA, Samad A. Trace elements in serum from Pakistani patients with acute and chronic ischemic heart disease and hypertension. Clin Chem. 1984;30:644–648. doi: 10.1093/clinchem/30.5.644. [DOI] [PubMed] [Google Scholar]
  26. Klotz LO, Kröncke KD, Buchczyk DP, Sies H. Role of copper, zinc, selenium and tellurium in the cellular defense against oxidative and nitrosative stress. J Nutr. 2003;133:1448s–s1451. doi: 10.1093/jn/133.5.1448S. [DOI] [PubMed] [Google Scholar]
  27. Kodali HP, Pavilonis BT, Schooling CM. Effects of copper and zinc on ischemic heart disease and myocardial infarction: a Mendelian randomization study. Am J Clin Nutr. 2018;108:237–242. doi: 10.1093/ajcn/nqy129. [DOI] [PubMed] [Google Scholar]
  28. Li J, Wang J, Chen X, Tong S, Fang J, Yu H. Cross-sectional survey of intralymphocytic and serum elements in hypertensive patients. Chin Med J. 1999;112:641–645. [PubMed] [Google Scholar]
  29. Li X, Meng X, Spiliopoulou A, Timofeeva M, Wei WQ, Gifford A, Shen X, He Y, Varley T, et al. MR-PheWAS: exploring the causal effect of SUA level on multiple disease outcomes by using genetic instruments in UK Biobank. Ann Rheum Dis. 2018;77:1039–1047. doi: 10.1136/annrheumdis-2017-212534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Li L, Yang W, Huang L, Feng X, Cheng H, Ge X, Zan G, Tan Y, Xiao L, et al. MR-PheWAS for the causal effects of serum magnesium on multiple disease outcomes in Caucasian descent. iScience. 2021;24:103191. doi: 10.1016/j.isci.2021.103191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Liu WM, ZG Zhu, HX Leng (2004) [Analysis of the contents of K, Na, Ca, Mg, Zn, Cu, Fe and Mn in serum of middle and old-aged hypertension patients]. Guang pu xue yu guang pu fen xi = Guang pu 24:360–2 [PubMed]
  32. Lorincz MT. Neurologic Wilson's disease. Ann N Y Acad Sci. 2010;1184:173–187. doi: 10.1111/j.1749-6632.2009.05109.x. [DOI] [PubMed] [Google Scholar]
  33. Ma J, Betts NM. Zinc and copper intakes and their major food sources for older adults in the 1994–96 continuing survey of food intakes by individuals (CSFII) J Nutr. 2000;130:2838–2843. doi: 10.1093/jn/130.11.2838. [DOI] [PubMed] [Google Scholar]
  34. McDermott JH. Antioxidant nutrients: current dietary recommendations and research update. J Am Pharm Assoc (washington, DC) 2000;40:785–99. doi: 10.1016/s1086-5802(16)31126-3. [DOI] [PubMed] [Google Scholar]
  35. Nie X, Wang Y, Zhao H, Guo M, Liu Y, Xing M. As(3+) or/and Cu(2+) exposure triggers oxidative stress imbalance, induces inflammatory response and apoptosis in chicken brain. Ecotoxicol Environ Saf. 2020;203:110993. doi: 10.1016/j.ecoenv.2020.110993. [DOI] [PubMed] [Google Scholar]
  36. Okoduwa SI, Umar IA, Ibrahim S, Bello F, Habila N. Age-dependent alteration of antioxidant defense system in hypertensive and type-2 diabetes patients. J Diabetes Metab Disord. 2015;14:32. doi: 10.1186/s40200-015-0164-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Olatunbosun DA, Bolodeoku JO, Cole TO, Adadevoh BK. Relationship of serum copper and zinc to human hypertension in Nigerians. Bull World Health Organ. 1976;53:134–135. [PMC free article] [PubMed] [Google Scholar]
  38. Padmanabhan S, Carty L, Cameron E, Ghosh RE, Williams R, Strongman H. Approach to record linkage of primary care data from Clinical Practice Research Datalink to other health-related patient data: overview and implications. Eur J Epidemiol. 2019;34:91–99. doi: 10.1007/s10654-018-0442-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Palmer TM, Lawlor DA, Harbord RM, Sheehan NA, Tobias JH, Timpson NJ, Davey Smith G, Sterne JA. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat Methods Med Res. 2012;21:223–242. doi: 10.1177/0962280210394459. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Pavão ML, Figueiredo T, Santos V, Lopes PA, Ferin R, Santos MC, Nève J, Viegas-Crespo AM. Whole blood glutathione peroxidase and erythrocyte superoxide dismutase activities, serum trace elements (Se, Cu, Zn) and cardiovascular risk factors in subjects from the city of Ponta Delgada, Island of San Miguel, The Azores Archipelago, Portugal. Biomark Biochem Indic Exposure Response Suscept Chem. 2006;11:460–471. doi: 10.1080/13547500600625828. [DOI] [PubMed] [Google Scholar]
  41. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498–2504. doi: 10.1101/gr.1239303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Shimizu Y, Nakazato M, Sekita T, Kadota K, Arima K, Yamasaki H, Takamura N, Aoyagi K, Maeda T. Association between the hemoglobin levels and hypertension in relation to the BMI status in a rural Japanese population: the Nagasaki Islands Study. Internal Med (tokyo, Jpn) 2014;53:435–440. doi: 10.2169/internalmedicine.53.1353. [DOI] [PubMed] [Google Scholar]
  43. Shimizu Y, Sato S, Koyamatsu J, Yamanashi H, Nagayoshi M, Kadota K, Kawashiri SY, Maeda T. Possible mechanism underlying the association between higher hemoglobin level and hypertension in older Japanese men. Geriatr Gerontol Int. 2017;17:2586–2592. doi: 10.1111/ggi.13068. [DOI] [PubMed] [Google Scholar]
  44. Stelzer G, Rosen N, Plaschkes I, Zimmerman S, Twik M, Fishilevich S, Stein TI, Nudel R, Lieder I, et al. The genecards suite: from gene data mining to disease genome sequence analyses. Curr Protoc Bioinform. 2016;54:1301–13033. doi: 10.1002/cpbi.5. [DOI] [PubMed] [Google Scholar]
  45. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, Downey P, Elliott P, Green J, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12:e1001779. doi: 10.1371/journal.pmed.1001779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, Simonovic M, Doncheva NT, Morris JH, et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;47:D607–D613. doi: 10.1093/nar/gky1131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Takeuchi H, Taki Y, Nouchi R, Yokoyama R, Kotozaki Y, Nakagawa S, Sekiguchi A, Iizuka K, Yamamoto Y, et al. Association of copper levels in the hair with gray matter volume, mean diffusivity, and cognitive functions. Brain Struct Funct. 2019;224:1203–1217. doi: 10.1007/s00429-019-01830-y. [DOI] [PubMed] [Google Scholar]
  48. Taneja SK, Mandal R. Mineral factors controlling essential hypertension–a study in the Chandigarh, India population. Biol Trace Elem Res. 2007;120:61–73. doi: 10.1007/s12011-007-8013-7. [DOI] [PubMed] [Google Scholar]
  49. Tisato F, Marzano C, Porchia M, Pellei M, Santini C. Copper in diseases and treatments, and copper-based anticancer strategies. Med Res Rev. 2010;30:708–749. doi: 10.1002/med.20174. [DOI] [PubMed] [Google Scholar]
  50. Verma A, Bradford Y, Dudek S, Lucas AM, Verma SS, Pendergrass SA, Ritchie MD. A simulation study investigating power estimates in phenome-wide association studies. BMC Bioinform. 2018;19:120. doi: 10.1186/s12859-018-2135-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Vivoli G, Bergomi M, Rovesti S, Pinotti M, Caselgrandi E. Zinc, copper, and zinc- or copper-dependent enzymes in human hypertension. Biol Trace Elem Res. 1995;49:97–106. doi: 10.1007/bf02788959. [DOI] [PubMed] [Google Scholar]
  52. Vlad M, Caseanu E, Uza G, Petrescu M. Concentration of copper, zinc, chromium, iron and nickel in the abdominal aorta of patients deceased with coronary heart disease. J Trace Elem Electrolytes Health Dis. 1994;8:111–114. [PubMed] [Google Scholar]
  53. Wang Y, Zeng S, Lin TM, Krugner-Higby L, Lyman D, Steffen D, Xiong MP. Evaluating the anticancer properties of liposomal copper in a nude xenograft mouse model of human prostate cancer: formulation, in vitro, in vivo, histology and tissue distribution studies. Pharm Res. 2014;31:3106–3119. doi: 10.1007/s11095-014-1403-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Wu W, Jiang S, Zhao Q, Zhang K, Wei X, Zhou T, Liu D, Zhou H, Zhong R, et al. Associations of environmental exposure to metals with the risk of hypertension in China. Sci Total Env. 2018;622–623:184–191. doi: 10.1016/j.scitotenv.2017.11.343. [DOI] [PubMed] [Google Scholar]
  55. Yao J, Hu P, Zhang D. Associations between copper and zinc and risk of hypertension in US adults. Biol Trace Elem Res. 2018;186:346–353. doi: 10.1007/s12011-018-1320-3. [DOI] [PubMed] [Google Scholar]
  56. Zhou J, Liu C, Francis M, Sun Y, Ryu MS, Grider A, Ye K. The causal effects of blood iron and copper on lipid metabolism diseases: evidence from phenome-wide mendelian randomization study. Nutrients. 2020 doi: 10.3390/nu12103174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhou J, Liu C, Sun Y, Francis M, Ryu MS, Grider A, Ye K. Genetically predicted circulating levels of copper and zinc are associated with osteoarthritis but not with rheumatoid arthritis. Osteoarthr Cartil. 2021;29:1029–1035. doi: 10.1016/j.joca.2021.02.564. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary file1 (XLSX 126 kb) (126.8KB, xlsx)

Data Availability Statement

The genetic and phenotypic UK Biobank data are available upon application from the UK Biobank (www.ukbiobank.ac.uk/).

The code that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Phenomics are provided here courtesy of Springer

RESOURCES