Introduction
Hypertension, as a prevalent cardiovascular disease, has exhibited a persistent rise in incidence, emerging as a crucial global public health issue (1). Statistics indicated that from 1990 to 2019, the population of hypertensive patients aged 30 to 79 doubled. Significant disparities exist among different countries and regions, with the prevalence in certain middle-income countries surpassing that of high-income countries (2). Hypertension not only predisposes individuals to serious complications such as cardiovascular disease and stroke but also heightens the risk of organ diseases such as those affecting the heart and kidneys, posing a significant threat to patients’ quality of life and health. Presently, the treatment for hypertension primarily comprises two main categories: pharmacological and non-pharmacological interventions. Pharmacological treatment commonly utilizes antihypertensive medications, such as β-blockers, ACE inhibitors, etc., to reduce blood pressure levels (3). Non-pharmacological interventions encompass lifestyle changes, including dietary control and increased physical activity. Despite the availability of multiple treatment methods, the management of hypertension continues to encounter numerous challenges, including suboptimal treatment outcomes and severe side effects. This challenge has prompted researchers to turn their attention to the potential role of micronutrients in the onset and treatment of hypertension, aiming to identify novel treatment avenues and strategies.
Micronutrients are essential nutrients for the human body, including calcium, iron, magnesium, vitamin A, and other elements. They play a vital role in the human body, encompassing cellular metabolism, nerve transmission, enzyme activity, and various other aspects. In recent years, Bastola et al. found a significant positive correlation between serum selenium levels and hypertension, but serum zinc and copper were not significantly positively correlated (4). Lewandowska et al. found that lower copper levels in early pregnancy were associated with a higher risk of hypertension. Zinc levels did not affect the risk of hypertension (5). Xiong et al. found a significant negative correlation between dietary folate, vitamin B6, and vitamin B12 and hypertension, indicating a protective effect of these nutrients against hypertension (6). However, current research primarily relies on observational studies, which are subject to numerous limitations. These limitations include challenges in establishing causality, susceptibility to recall bias and information bias, and the inability to adequately control for confounding factors. Consequently, there is an urgent requirement for more stringent research to elucidate the precise role of micronutrients in the onset and progression of hypertension.
In the current context, Mendelian randomization (MR) studies have emerged as a prominent and highly regarded method. Its uniqueness lies in the use of genetic variations such as single nucleotide polymorphisms (SNPs) as instrumental variables (IVs) to eliminate the influence of confounding factors, allowing for a more precise assessment of the effects of micronutrients on hypertension (7). However, the absence of MR studies on circulating micronutrients and hypertension to date underscores the importance of our research. Hence, we employed a two-sample multivariable Mendelian randomization (MVMR) study to examine the potential causal relationships between 15 micronutrients (copper, selenium, zinc, calcium, magnesium, potassium, iron, folate, carotenoids, vitamin A, vitamin B6, vitamin B12, vitamin C, vitamin D, and vitamin E) and hypertension. Our study aims to fill this knowledge gap and provide theoretical insights for future research on hypertension prevention and treatment.
Methods
Study design
Our study followed the STROBE-MR statement for reporting MR studies (8). We utilized a two-sample, MVMR method to explore the potential causal relationship between circulating micronutrient concentrations and hypertension. MR relies on three critical assumptions: (1) IV is closely associated with circulating micronutrient concentrations; (2) IV should remain unaffected by known or unknown confounding factors; (3) IV affects hypertension solely through circulating micronutrient concentrations (9). All studies included in genome-wide association studies (GWAS) have received approval from the respective review committees; hence, ethical approval and informed consent are deemed unnecessary.
Outcome data sources
Genotypic data for hypertension were extracted from FinnGen, including 55,917 cases of hypertension and 162,837 controls. FinnGen is a large-scale public–private partnership aimed at collecting and analyzing genomic and health data from 500,000 participants in the Finnish biobank. Our study included only individuals of European ancestry to minimize any confounding effects due to ancestry.
Selection of instrumental SNPs
Initially, we searched PubMed (accessed on April 12, 2024) to retrieve published observational studies or meta-analyses related to micronutrients and hypertension. The preliminary catalog comprises 21 micronutrients: copper, selenium, zinc, calcium, magnesium, sodium, potassium, iron, arsenic, cadmium, mercury, manganese, lead, folate, carotenoids, vitamin A, vitamin B6, vitamin B12, vitamin C, vitamin D, and vitamin E (4, 6, 10, 11, 12, 13, 14, 15, 16, 17). Following this, we conducted additional searches in the GWAS catalog (https://www.ebi.ac.uk/gwas, accessed on April 17, 2024) to acquire GWAS data on circulating micronutrient concentrations in European populations. Sodium, arsenic, cadmium, mercury, manganese, and lead were excluded due to the absence of GWAS data. Ultimately, this study evaluated a total of 15 micronutrients, comprising copper, selenium, zinc, calcium, magnesium, potassium, iron, folate, carotenoids, vitamin A, vitamin B6, vitamin B12, vitamin C, vitamin D, and vitamin E. Detailed information on these micronutrients is provided in Table 1. Eligible genetic instruments were chosen based on the following criteria: first, each SNP had to have a genome-wide significant P-value< 5 × 10–6 and an F-statistic > 10 to satisfy the relevance assumption. Second, SNPs with linkage disequilibrium (LD; 10 MB clustering window and an R2 threshold of 0.001) were excluded to meet the independence assumption. Third, PhenoScanner V2 also excluded SNPs associated with the outcome or potential confounders to fulfill the exclusion restriction assumption.
Table 1
Detailed information of the GWAS datasets used in the present study.
| EXPOSURE | GWAS ID | STUDY OR CONSORTIUM | SAMPLE SIZE | SNP | YEAR | ANCESTRY |
|---|---|---|---|---|---|---|
| Copper | ieu-a-1073 | Evans DM et al. | 2,603 | 2,543,646 | 2013 | European |
| Selenium | ieu-a-1077 | Evans DM et al. | 2,603 | 2,543,617 | 2013 | European |
| Zinc | ieu-a-1079 | Evans DM et al. | 2,603 | 2,543,610 | 2013 | European |
| Calcium | ukb-b-8951 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Iron | ukb-b-20447 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Magnesium | ukb-b-7372 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Potassium | ukb-b-17881 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Carotene | ukb-b-16202 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Folate | ukb-b-11349 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Vitamin A | ukb-b-9596 | Ben Elsworth et al. | 460,351 | 9,851,867 | 2018 | European |
| Vitamin B6 | ukb-b-7864 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Vitamin B12 | ukb-b-19524 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Vitamin C | ukb-b-19390 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Vitamin D | ukb-b-18593 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
| Vitamin E | ukb-b-6888 | Ben Elsworth et al. | 64,979 | 9,851,867 | 2018 | European |
[i] GWAS, genome-wide association study.
Univariable MR analysis
For each phenotype of micronutrients, after harmonizing exposure and outcome directions, the inverse variance weighted (IVW) method was employed as the primary MR analysis. In addition to IVW, sensitivity analyses were performed using the MR-Egger, weighted median, weighted mode, and simple mode methods. To ensure the reliability of the results, it was required that the results of the IVW method be statistically significant, and the results of the other four methods should be consistent with the IVW results in direction. The MR-Egger intercept test is employed to detect horizontal pleiotropy, where a P intercept > 0.05 indicates the absence of such pleiotropy. Additionally, we evaluate heterogeneity using the IVW method and MR-Egger regression, where P < 0.05 indicates its existence. Cochran’s Q statistic is employed for assessing heterogeneity.
Multivariate MR analysis
Considering potential interactions between different micronutrients, we have performed MVMR analysis on micronutrients that have exhibited significance in univariable Mendelian randomization (UVMR) analysis. This enables a more accurate estimation of their direct effects on hypertension. In MVMR analysis, the instrumental SNPs chosen must meet the criteria previously described for UVMR selection.
All statistical analyses in this study were performed using the ‘TwoSampleMR’ and ‘MVMR’ packages in R software (version 4.1.0) for MR analysis. The causal effects of exposure on outcomes were presented using odds ratios (ORs) and 95% confidence intervals (CIs).
Results
Instrumental variables
Based on the three main assumptions of MR, a series of SNPs were obtained to predict the traits in the study. Among these SNPs, 6 were utilized for predicting copper, 8 were employed for predicting zinc, and the number of SNPs used for predicting other traits varied from 6 to 19 (Table S1). Furthermore, the F-statistics of SNPs are listed in Table S1, and they are all >10.
UVMR analysis of the causal relationship between micronutrients and hypertension
The results of the IVW method suggest a potential influence of copper on hypertension, with an OR of 1.052 (95% CI: 1.006–1.099; P = 0.025). Similarly, the OR for zinc is 1.083 (95% CI: 1.007–1.165; P = 0.031; Figure 1). Conversely, micronutrients such as selenium, calcium, magnesium, folic acid, and vitamin A have no significant impact on hypertension (all P > 0.05). As shown in Figure 2 and Table S2. These findings were corroborated by other MR analysis methods. Sensitivity analysis revealed heterogeneity between calcium, iron, potassium, and hypertension (Table 2). Therefore, we employed the multiplicative random-effects IVW method in this study. MR-Egger regression analyses did not detect potential horizontal pleiotropy (allP > 0.05), indicating that IVs do not significantly influence the outcome through pathways other than exposure (Table 2).
Figure 1
A forest plot showing associations between genetically determined levels of copper, zinc and hypertension based on IVW MR analysis. SNPs single nucleotide polymorphism, IVW inverse-variance weighted, OR odds ratio, CI confidence interval.
Figure 2
Circular heat map of suggestive genetic correlation between micronutrients and hypertension. IVW, Inverse variance weighted.
Table 2
Heterogeneity and pleiotropy analysis in MR analysis.
| EXPOSURE | MR METHOD | COCHRAN Q STATISTIC | EGGER INTERCEPT | HETEROGENEITY P-VALUE | PLEIOTROPY P-VALUE |
|---|---|---|---|---|---|
| Copper | MR Egger | 0.11 | –0.011 | 0.999 | 0.392 |
| Copper | IVW | 1.03 | 0.960 | ||
| Selenium | MR Egger | 3.86 | 0.018 | 0.425 | 0.320 |
| Selenium | IVW | 5.15 | 0.398 | ||
| Zinc | MR Egger | 5.60 | 0.010 | 0.470 | 0.709 |
| Zinc | IVW | 5.75 | 0.569 | ||
| Folate | MR Egger | 14.80 | 0.008 | 0.139 | 0.682 |
| Folate | IVW | 15.06 | 0.180 | ||
| Carotene | MR Egger | 12.53 | –0.010 | 0.484 | 0.471 |
| Carotene | IVW | 13.09 | 0.520 | ||
| Potassium | MR Egger | 24.01 | 0.021 | 0.020 | 0.426 |
| Potassium | IVW | 25.37 | 0.021 | ||
| Vitamin D | MR Egger | 8.45 | 0.040 | 0.672 | 0.136 |
| Vitamin D | IVW | 11.03 | 0.526 | ||
| Vitamin C | MR Egger | 4.88 | –0.005 | 0.770 | 0.774 |
| Vitamin C | IVW | 4.97 | 0.837 | ||
| Vitamin B12 | MR Egger | 8.48 | 0.013 | 0.205 | 0.606 |
| Vitamin B12 | IVW | 8.90 | 0.260 | ||
| Iron | MR Egger | 20.25 | 0.023 | 0.016 | 0.408 |
| Iron | IVW | 21.94 | 0.015 | ||
| Vitamin E | MR Egger | 13.68 | –0.014 | 0.188 | 0.397 |
| Vitamin E | IVW | 14.75 | 0.194 | ||
| Magnesium | MR Egger | 11.25 | –0.008 | 0.735 | 0.531 |
| Magnesium | IVW | 11.66 | 0.767 | ||
| Vitamin B6 | MR Egger | 10.23 | 0.013 | 0.805 | 0.350 |
| Vitamin B6 | IVW | 11.16 | 0.800 | ||
| Calcium | MR Egger | 36.73 | –0.019 | 0.004 | 0.488 |
| Calcium | IVW | 37.81 | 0.004 | ||
| Vitamin A | MR Egger | 7.25 | –0.015 | 0.612 | 0.553 |
| Vitamin A | IVW | 7.63 | 0.665 |
[i] MR, Mendelian randomization; IVW, inverse variance weighted.
MVMR analysis of the causal relationship between micronutrients and hypertension
Due to the results of UVMR analysis showing that only copper and zinc are causally related to hypertension, the other 13 micronutrients were excluded from MVMR analysis. Ultimately, after adjusting for copper, zinc continued to demonstrate a positive direct effect on hypertension (OR = 1.087, 95% CI: 1.026–1.151, P = 0.005). Conversely, after adjusting for zinc, copper no longer had a direct impact on hypertension (OR = 1.026, 95% CI: 0.987–1.066, P = 0.193; Figure 3).
Figure 3
Multivariable Mendelian randomization analysis of the impact of copper, zinc on hypertension. OR odds ratio, CI confidence interval.
Discussion
Hypertension is a common health issue worldwide, presenting significant challenges to global socioeconomic development. To further explore potential preventive measures, we conducted an MR study investigating the causal relationship between the levels of 15 circulating micronutrients and hypertension. Our analysis identified that circulating zinc levels may be a potential risk factor for hypertension. However, there was little evidence observed for the association of other micronutrients with the risk of hypertension.
Zinc is an essential nutrient for the human body, involved in maintaining protein structure, enzyme catalysis, and information transmission (18). However, there are multiple viewpoints and research findings regarding the relationship between zinc and hypertension. First, some studies supported the protective role of zinc in hypertension (19, 20). This perspective argues that zinc deficiency can cause arteries to become stiff, fragile, and prone to inflammation, rather than flexible, which may lead to elevated blood pressure, particularly systolic blood pressure (21). However, increased zinc levels may have an inhibitory effect on inflammatory responses, helping to reduce the inflammatory response of vascular endothelial cells, thereby protecting vascular function and reducing the risk of hypertension (22, 23). Additionally, zinc may lower blood pressure levels by promoting the generation of nitric oxide, which promotes vasodilation (24). Second, other studies maintain the opposite perspective, suggesting that zinc may be a risk factor for hypertension. They found that circulating zinc levels may be correlated with the occurrence and development of hypertension, indicating that higher zinc levels may be linked to an increased risk of hypertension (25, 26, 27). This discovery has prompted a reevaluation of the role of zinc in the pathophysiology of hypertension. This may be attributed to the role of zinc in antioxidative stress and calcium ion regulation. Zinc deficiency can result in an increase in superoxide levels, damaging endothelial structure, increasing vascular tension, and ultimately promoting the development of hypertension (28, 29). Furthermore, the inhibitory effect of zinc on ATP-dependent calcium pumps may lead to an increase in vascular wall tension, further exacerbating the progression of hypertension (30, 31). These perspectives offer new insights into understanding the relationship between zinc and hypertension. Apart from the above two perspectives, some studies have not found a clear association between zinc and hypertension (5). For instance, a cross-sectional study conducted by Yao et al. found no independent correlation between zinc and hypertension in the adult population in the United States (32). Bastola et al. and Darroudi et al. also reported similar findings (4, 33).
Our UVMR study indicated a risk association between circulating zinc levels and hypertension. After adjusting with MVMR methods, we still observed the association between zinc and hypertension, further supporting our initial findings. This suggests that even when considering other potential influencing factors, zinc may still be one of the important risk factors for the onset of hypertension. Besides the physiological processes already mentioned, such as vascular tone regulation, antioxidative stress, and anti-inflammatory effects, some studies have found that zinc may also influence blood pressure levels through pathways such as affecting kidney function and inhibiting the sympathetic nervous system (34, 35). Nevertheless, the precise mechanisms underlying the association between zinc and hypertension remain incompletely understood. The implications of these findings suggest that reducing zinc intake within a healthy range may help lower the risk of hypertension, thus informing public health strategies and clinical guidelines for hypertension prevention and management. For future research, exploring the specific pathways and molecular mechanisms through which zinc influences hypertension is critical. This could aid in the development of targeted interventions, potentially involving zinc modulation as part of personalized hypertension treatment plans. Moreover, nonlinear relationships between zinc and hypertension should be investigated, as such dynamics might better explain varying outcomes across different populations and levels of zinc intake.
Although our study did not find a clear causal relationship between other micronutrients and hypertension, there are still some possible mechanisms worthy of attention. For example, copper may affect the blood pressure regulatory system through pathways such as participating in angiogenesis and regulating vascular constriction and dilation (36). Iron, as a component of hemoglobin and myoglobin, may be related to vascular function and oxygen delivery, thereby affecting blood pressure levels (37). Magnesium is involved in multiple biochemical reactions, including neural–muscular conduction and cell signal transduction, potentially exerting a regulatory effect on blood pressure regulation (38). Furthermore, vitamin A participates in retinal generation and cell differentiation, potentially influencing vascular health and consequently affecting blood pressure regulation (11). Vitamin B6 plays a role in amino acid metabolism and neural transmission, which may affect vascular function and the nervous system, thereby influencing blood pressure (39). Despite the lack of direct evidence supporting the causal relationship between these micronutrients and hypertension in our study, future research can continue to explore their potential roles in the pathogenesis of hypertension and verify their associations with hypertension. In conclusion, our study offers some new insights into understanding the relationship between micronutrients and hypertension, while also posing numerous questions and challenges for future research to delve into.
The primary strength of this study lies in the novel application of MVMR methods to investigate the causal relationship between various circulating micronutrients and hypertension. This approach controls for confounding factors, enabling a more precise evaluation of the impact of micronutrients on hypertension and offering a fresh perspective for hypertension prevention and treatment. It is worth noting that this study has some limitations. First, the MR method assumes a linear relationship between exposure and outcome, which may not hold when nonlinear associations exist. Micronutrients may be beneficial within certain ranges, but deviations could lead to different outcomes. Our study does not account for nonlinear effects or provide specific zinc thresholds for hypertension, limiting its clinical applicability. Future research should determine optimal zinc concentrations and thresholds for intervention. Second, although MR helps to minimize confounding and reverse causality, it cannot fully eliminate bias from pleiotropy, where genetic variants affect the outcome through pathways unrelated to the exposure of interest. This may introduce some residual confounding in the results. Additionally, our study relies on data from publicly available GWAS datasets, which may lack certain individual-level details such as ethnicity, age, or sex-specific effects. These factors could influence the observed associations, thus potentially limiting the generalizability of the findings to broader populations. Finally, we only considered the effects of circulating micronutrients and did not take into account other factors that may influence hypertension, such as diet and lifestyle. These factors may have significant implications for the onset and development of hypertension, so future research should consider these factors holistically to gain a more comprehensive understanding of the pathogenesis of hypertension.
Conclusion
In summary, our MVMR study highlights zinc’s potential as a risk factor for hypertension, while other micronutrients showed no significant association. These findings suggest that reducing zinc intake within a healthy range may help lower the risk of hypertension. Future research should delve deeper into understanding the role of zinc and explore nonlinear associations for a more comprehensive understanding.
Data Accessibility Statement
All summary-level data necessary to conduct this MR analysis were obtained from https://gwas.mrcieu.ac.uk/.
Additional File
The additional file for this article can be found as follows:
Abbreviations
MR, Mendelian randomized; GWAS, genome-wide association studies; SNP, single nucleotide polymorphism; UVMR, univariable Mendelian randomization; IVW, inverse variance weighting; MVMR, multivariate Mendelian randomization; OR, odds ratio; CI, confidence interval; IV, instrumental variable; LD, linkage disequilibrium.
Funding Information
This study was supported by the National Natural Science Foundation of China (81974548), Young Qi Huang Scholars Support Program, State Administration of Traditional Chinese (20201 A2180), and China Postdoctoral Science Foundation (2021MD703841).
Competing Interests
The authors have no competing interests to declare.
Author Contributions
Yuting Liu: study conception and design, acquisition of data, analysis, and interpretation of data, writing the manuscript; Chenggong Bao: study conception and design, writing the manuscript, critical revision of the manuscript; Han Wang: study conception and design, acquisition of data; Dongsheng Wei: study conception and design, analysis and interpretation of data; Zhe Zhang: writing manuscript, acquisition of data and critical revision of the manuscript.
Yuting Liu and Chenggong Bao have contributed equally to this work and share first authorship.