Genetically determined watercress intake and risk for bone mineral density

doi:10.21203/rs.3.rs-2983664/v1

Purpose: It has been previously shown that watercress intake increases the expression of type I collagen and promotes human osteoblast-like MG-63 cells of osteoblastic proliferation and differentiation. However, there have been no randomized controlled clinical trials of watercress intake in bone mineral density (BMD). It remains to be clarified if watercress intake leads to an increase in BMD. In this study, we used two-sample Mendelian randomization (MR) to estimate the causal effect of watercress intake on susceptibility to BMD.

Methods: Single nucleotide polymorphisms (SNP) associated with watercress intake phenotypes, used as instrumental variables, were extracted from public genome-wide association study (GWAS) summary statistics. We take the intake of watercress as the exposure and BMD as the outcome. By using MR, it is possible to investigate whether there is a causal association between the exposure and the outcome.

Results: Using the IVW-fixed(Inverse variance weighted, fixed effects) and IVW-random methods(Inverse variance weighted, multiplicative random effects), we found the evidence to report the causal associations between the watercress intake and femoral neck BMD , but the intake of watercress has nothing to do with lumbar spine BMD and forearm BMD（Table 1）.

Conclusion: The present study suggested a harmful role of watercress intake on the risk for femoral neck BMD. Further studies elucidating the underlying mechanism are warranted.

Osteoporosis, one of the most frequent form of metabolic bone diseases, is defined as a skeletal disorder characterized by low bone mass and micro-architectural deterioration of bone tissue, predisposing a person to an increased risk of fracture [1, 2]. There are 200 million women suffering from osteoporosis worldwide and 1/3 women aged over 50 years will have an osteoporotic fracture in their lifetime, which becomes an important disease factor threating women both physical and mental health as well as quality of life [3]. Osteoporosis represents a major public health problem through its association with fragility fractures, primarily of the hip, spine and distal forearm [4]. The U.N. (Unit-ed Nations) predicted that by 2050, the number of people over 60 years old will reach 2.5 billion, accounting for about 21.3% of the population [5]. In the United States alone, a total of 340,000 hip fractures occur each year, and the number of hip fractures is predicted to more than triple worldwide from 1.66 million in 1990 to 6.26 million in 2050 [4, 6]. Strikingly, the mortality rate is approximately 20% during the first year following a hip fracture [7]. Osteoporosis and osteoporotic fractures seriously affect the health and quality of life of the elderly, bringing a heavy burden to society and their family.

Watercress, a cruciferous vegetable, is a perennial aquatic or semi-aquatic plant with high nutritional value [8]. Watercress originated in Europe and is widely used in local recipes. Fresh watercress leaves can be used as salad vegetables [9]. In traditional medicine, it has also been investigated for treating hypercholesterolemia, hyperglycemia, hypertension, arthritis, diuresis, toothache and gangrene [10]. Watercress is used as a medicine to relieve abdominal pain in Turkish folk medicine [11]. It can be seen that the intake frequency of watercress is very high in some people. A previous study showed that rutin, the main component of watercress, is converted into quercetin in the large intestine by the glycosidase enzyme of intestinal bacteria [12]. On the contrary, quercetin inhibits the proliferation, differentiation and mineralization of osteoblast like cells in rat calvaria [13, 14]. Whether watercress is a risk factor for BMD is controversial. However, few studies have investigated the potential public health implications of watercress. Therefore, large-scale randomized controlled trials are needed to address these issues, but RCTs are relatively costly.

Mendelian randomization (MR) is a genetic epidemiology method that utilizes genetic variation with the modifiable exposure (e.g., watercress intake) as proxies for the exposure to assess whether the exposure is causally related to the outcome (e.g., bone mineral density). Compared with observational studies, MR can avoid both reverse causation bias and potential confounding bias [15]. Randomized controlled trial (RCT) can directly infer the causal relationship between watercress intake and BMD, which is helpful to prevent and early intervene osteoporosis and secondary fracture in high-risk population. However, in some cases, such as high costs, human resources, time-consuming and ethical constraints, RCT is difficult to perform and may not be well accepted by doctors and patients [16]. As an alternative, the Mendelian randomization study design is similar to a randomized controlled trial, a convenient and popular method using exposure-related genetic variants to study causality [17].

Table 1

Two-sample MR analyses of watercress intake and BMD levels. IVW- random stands for Inverse variance weighted (multiplicative random effects). IVW- fixed stands for Inverse variance weighted (fixed effects). FN-BMD, femoral neck mineral density BMD; LS-BMD, lumbar spine BMD; FA-BMD, forearm BMD. Watercress intake is related to the reduction of bone mineral density of the femoral neck (p < 0.05).
Outcomes	IVW-random		IVW-fixed		Weighted median		MR egger		Simple median
	β	P-value	β	P-value	β	P-value	β	P-value	β	P-value
FN-BMD	0.17	0.02	-0.17	0.03	-0.17	0.12	-0.26	0.10	-0.17	0.11
LS-BMD	-0.09	0.36	-0.09	0.33	-0.03	0.80	-0.41	0.03	-0.04	0.77
FA-BMD	-0.04	0.83	-0.04	0.81	0.005	0.98	-0.34	0.33	0.01	0.97

Study design

In order to confirm the causal relationship between watercress intake and BMD, we con-ducted a two-sample Mendelian randomization (MR) study. Both the exposure and out-comes data were obtained from public genome wide association study (GWAS) database. The research priorities of the present study are BMD at various skeletal sites, including forearm BMD, femoral neck BMD and lumbar spine BMD. All the Date used for the study were publicly available.

Instrumental Variable Selection

In principle, the instrumental variables used in MR analysis must satisfy the following three assumptions [18]: ① IVs must be relevant to exposure (watercress intake); ② IVs must be independent of any confounder; ③ IVs must be associated with outcome (BMD) only via exposure (watercress intake) (Fig. 1). From the GWAS summary data of watercress intake, we conducted a series of quality control steps to select eligible instrumental SNPs. First, we retrieved SNPs related to watercress intake from the GWAS(p<5×10 − 5, Supplementary Table S1 ). Second, the SNPs constituting each genetic instrument were checked for strong linkage disequilibrium (LD). Strong LD between SNPs may exaggerate the significance level of linkage and thus generate false positive results, and thus we performed a clumping process(r²<0.001; kb = 10000)with the European samples from the 1000 genomes project which were used to estimate LD between SNPs. The Instrumental Variable SNPs were extracted from the full GWAS data of outcomes (BMDs). For the second assumption, relevant traits associated with the SNPs were also searched in the PhenoScannerV2 (an expanded tool for searching human genotype phenotype associations) database at a significance level of 5×10 − 5 for potential pleiotropy [19]. We removed three SNPs (rs1912995, rs113812291, rs143792188) associated with any potential confounders of BMD. For the third key assumption (exclusion restriction assumption), we performed the MR Steiger filtering test to identify the SNPs suggestive of causality in the reverse direction [20]. We harmonized the effects of the SNP on the outcome and the exposure by ensuring that they refer to the same allele, correcting the strand for non-palindromic SNPs, and dropping all palindromic SNPs from the analysis. These rigorously selected SNPs were used as the final instrumental SNPs for the subsequent MR analysis [21]. Minor allele frequency (MAF) for each SNP was calculated (MAF > 0.01) in the Instrumental Variable SNPs [21]. SNPs with minor allele frequency (MAF) < 0.01 were removed [22]. For Mendelian randomization, the F statistic is an indicator of the strength of the instrument variable, with values over 10 reflecting strong instruments. We use the following equation: F = R² x (N-2)/(1-R²), where R²=2 x β² x EAF x (1-EAF)/ [ 2 x β² x EAF x (1-EAF) + 2 x SE² x N x EAF x (1-EAF)]. N represents the number of participants, EAF stands for effective allele frequency and β is the estimated effect of the SNP to assess its ability to uniquely predict the outcome [23–25].

Data Sources

GWAS data of watercress intake (Sample size = 64949) can be downloaded from IEU website. (IEU OpenGWAS project (mrcieu.ac.uk))BMDs statistics are from the Genetic Factors for Osteoporosis Consortium which includes Summary statistics on BMDs of three different parts (femoral neck BMD, n = 32,735;lumbar spine BMD, n = 28,498;forearm BMD, n = 8143)[19]( Home | GEFOS). All participants are European (Table 2).

Table 2

Detailed information of studies and datasets used for analyses.
	Date source	Phenotype	Sample size	Population
Exposure	IEU	Watercress intake	64949	European
Outcomes	GEFOS	FN-BMD LS-BMD FA-BMD	32735 28498 8143	European

Statistical Analysis

MR analysis uses genetic variation as an Instrumental Variable to estimate the causal im-pact of exposure variables on results. In this study, we used different methods to estimate the causal relationship between the intake of watercress and BMDs (FN-BMD, LS-BMD and FA-BMD) by combining the summary statistics. The causal associations were estimated with inverse variance weighted (IVW) with fixed effects (IVW-fixed) and random effects (IVW-random), weighted median [26], MR-Egger [27] and Simple median estimate methods. The IVW as the main method was used to provide a combined estimate of the causal estimate from each SNP [28]. The IVW method is the main method used to evaluate the causal effect of MR analysis of two samples [21]. The MR-Egger analysis can identify unbalanced and horizontal pleiotropy through its intercept test [29]. MR-PRESSO is a method for correcting outliers in IVW. Three MR-PRESSO tests were applied: the global test was used to detect the presence of horizontal pleiotropy, the outlier test to identify variants with significant horizontal pleiotropic effect, and the distortion test to estimate the distortion caused by significant horizontal pleiotropic outlier variants [21, 30, 31]. The Chi-square-based Q statistic test (Cochran’s Q statistic) and the I² statistic were calculated to determine between-study heterogeneity [32]. All the MR analyses in this study were conducted in software R (version 4.2.1) with R packages including Two Sample MR and MR-PRESSO [32, 33]. The whole process was plotted as a flow diagram (Fig. 2).

The effect of watercress intake on BMD may differ depending on the skeletal site.

In MR analysis based on IVW model, 92 effective SNPs were significantly related to watercress intake. The results were essentially unchanged after removing SNPs associated with any potential confounders. In the multiplicative random effect IVW model, the intake of watercress has a causal relationship with BMDs of the femoral neck, that is, the intake of watercress is related to the reduction of bone mineral density of the femoral neck (β=-0.17, p=0.02). But there is no causal relationship with forearm and lumbar bone mineral density (Figure 3 and 4). Different analysis methods were used for the same data, mainly to enhance the reliability of the IVW analysis results. Our research results show that the IVW results are reliable, and the causality is the same as the main analysis direction (Figure 3 and 4).

Sensitivity Analysis

Sensitivity analysis was conducted to verify the reliability of IVW results [34]. Heterogeneity was assessed using Cochran’s Q statistic test and the I² (𝐼²=100%×（Q-DF）/Q, take 0 when I² is negative)test, where Q is Cochran’s heterogeneity statistic and DF the degrees of freedom. When there is heterogeneity, the IVW random effect model is used. If there is no heterogeneity, the IVW fixed effect model or the random effect model is used [35,36]. Our results showed no heterogeneity (Table 3). The MR-Egger regression intercept and MR-PRESSO global test were also implemented to examine horizontal pleiotropy [37]. The results of both agreed to show that there was no pleiotropic between watercress intake and femoral neck bone density (Table 3). We also performed a leave-one-out method analysis, observing that the remaining SNPs were removed one by one, and then the remaining SNPs were analyzed.

Table 3. Evaluation of heterogeneity and directional pleiotropy using different methods. I²＝0：No heterogeneity; I²＜31%：Homogeneous ;31%＜I²＜56%：Heterogeneity cannot be ruled out ;I²＞56%：Great heterogeneity

		Heterogeneity		Pleiotropy
Exposure	Outcomes	I² (%)	Cochran’s Q p	MR−Egger Intercept p	MR−PRESSO Global Test p
	FN-BMD	0	0.71	0.53	0.71
watercress	LS-BMD	13.29	0.16	0.05	0.17
	FA-BMD	23.26	0.03	0.31	0.08

We found that decreased BMD of the femoral neck was associated with watercress intake, but the intake of watercress has nothing to do with lumbar spine BMD and forearm BMD. The main component of watercress is rutin and previous studies have shown that rutin can improve bone microstructure and stimulate osteoblast differentiation [38]. Some studies have also shown that quercetin in rutin can inhibit the proliferation, differentiation and mineralization of osteoblasts in rat skull Cell-like cells [14]. Therefore, the influence of rutin on bone microstructure is still controversial. Our study aims to determine whether there is a causal relationship between watercress intake and bone mineral density in the crowd. In this study, we used GWAS summary statistics to determine the causal relationship between watercress intake and BMD at different bone sites. The results showed that the intake of watercress had a differential impact on diverse bone sites, and its intake had a causal relationship with the decline of bone mineral density of the femoral neck, but had no relationship with the bone mineral density of the forearm and lumbar spine. We speculate that the different effects of watercress intake on BMD of different bone parts may be related to the different proportions of cortical and trabecular bone and to the metabolic activity in each type of bone [39]. The exact cause for this is unclear, but it may be complex. Although the MR method is more resistant to reverse causation and confounding than traditional observational epidemiological studies, there are limitations to this approach. These include population stratification, pleiotropy (genes influencing multiple phenotypes), canalization (the ability of a population to produce the same phenotype regardless of variability of its environment or genotype), inadequate power and linkage disequilibrium [40]. Mendelian randomization the gold standard for inferring causality is a randomized control trial [41], but no RCT showed a negative effect of watercress intake on BMD. Our research also has some limitations. For example, we have not calculated power. Since no scholars have done observational studies in this study, some data for calculating the power are missing, making the power impossible to calculate.

Decreased BMD of the femoral neck was associated with watercress intake. For people who like to eat watercress for a long time, BMD was routinely and regularly assessed to detect low BMD as early as possible with the aim of preventing fractures. However, more investigations are necessary to determine whether watercress may be a potential risk factor that affect bone mineralization.

Ethical Approval

Our analysis uses published research or publicly available GWAS summary data. There is no draft to collect the original data, so the approval of the Ethics Committee is not required.

Funding

This work was supported by the Shanghai Natural Science Foundation (19ZR1440700); Medical-Industrial Interdisciplinary Research Fund of Shanghai Jiaotong University (YG2017MS02); Youth Science and Technology Innovation Studio of Shanghai Jiao Tong University School of Medicine (JYKCGZS8); Practice training base for interdisciplinary innovative talents of Shanghai Jiao Tong University (SJTUJXCX-3). National Natural Science Foundation of China Projects (82103410). Shanghai Young Science and Technology Talent Sailing Program (21YF1437400).

Availability of data and materials

The summary GWAS data for watercress intake from [IEU OpenGWAS project (mrcieu.ac.uk)]. The bone mineral density summary data were accessed from [Home | GEFOS].

Author contributions

Q-H.Z. conceived and designed the project. Z-C.D. L-S.L. and Y.L. contributed to the acquisition and analysis of data. E-J.H. and H-J.W. provided constructive comments and discussion. W-H.G. and Z-Y. G. contributed to the data curation and formal analysis. C.C. contributed to the analysis of data. Q-H.Z. and Z-C.D. wrote the manuscript. All authors discussed the results and commented on the paper.

Conflict of Interest

Zhicheng Dai, Longshuai Lin, Yuan Liu, Enjun He, Weihong Guo, Chen Chen, Zengya Guo, Hongjie Wang and Qinghua Zhao declare that they have no conflict of interest.

Acknowledgments

This work was supported by the Shanghai Natural Science Foundation (19ZR1440700); Medical-Industrial Interdisciplinary Research Fund of Shanghai Jiaotong University (YG2017MS02); Youth Science and Technology Innovation Studio of Shanghai Jiao Tong University School of Medicine (JYKCGZS8); Practice training base for interdisciplinary innovative talents of Shanghai Jiao Tong University (SJTUJXCX-3). National Natural Science Foundation of China Projects (82103410). Shanghai Young Science and Technology Talent Sailing Program (21YF1437400). We thank the consortiums and investigators involved in the GWASs for access to their data.

Lin L, Guo Z, He E, Long X, Wang D, Zhang Y, Guo W, Wei Q, He W, Wu W, Li J, Wo L, Hong D, Zheng J, He M, Zhao Q. SIRT2 regulates extracellular vesicle-mediated liver-bone communication. Nat Metab. 2023 May 15. doi: 10.1038/s42255-023-00803-0. Epub ahead of print. PMID: 37188819..
Föger-Samwald U, Dovjak P, Azizi-Semrad U, Kerschan-Schindl K, Pietschmann P. Osteoporosis: Pathophysiology and therapeutic options. EXCLI J. 2020 Jul 20;19:1017-1037.
Wright NC, Looker AC, Saag KG, Curtis JR, Delzell ES, Randall S, Dawson-Hughes B. The recent prevalence of osteoporosis and low bone mass in the United States based on bone mineral density at the femoral neck or lumbar spine. J Bone Miner Res. 2014 Nov;29(11):2520-6.
Gourlay ML, Brown SA. Clinical considerations in premenopausal osteoporosis. Arch Intern Med. 2004 Mar 22;164(6):603-14.
UNDESA World Population Ageing 2017
Clynes MA, Harvey NC, Curtis EM, Fuggle NR, Dennison EM, Cooper C. The epidemiol-ogy of osteoporosis. Br Med Bull. 2020 May 15;133(1):105-117.
Cooper C, Atkinson EJ, Jacobsen SJ, O'Fallon WM, Melton LJ 3rd. Population-based study of survival after osteoporotic fractures. Am J Epidemiol. 1993 May 1;137(9):1001-5.
Klimek-Szczykutowicz M, Szopa A, Ekiert H. Chemical composition, traditional and professional use in medicine, application in environmental protection, position in food and cosmetics industries, and biotechnological studies of Nasturtium officinale (watercress) - a review. Fitoterapia. 2018 Sep;129:283-292.
Gonçalves EM, Cruz RMS, Abreu M, Brandão TRS, Silva CLM. Biochemical and colour changes of watercress (Nasturtium officinale R. Br.) during freezing and frozen storage. J Food Eng. 2009;93:32-39.
Panahi Kokhdan E, Khodabandehloo H, Ghahremani H, Doustimotlagh AH. A Narrative Review on Therapeutic Potentials of Watercress in Human Disorders. Evid Based Com-plement Alternat Med. 2021 May 7;2021:5516450.
Ozen T. Investigation of antioxidant properties of Nasturtium officinale (watercress) leaf extracts. Acta Pol Pharm. 2009 Mar-Apr;66(2):187-93.
Tamura G, Gold C, Ferro-Luzzi A, Ames BN. Fecalase: a model for activation of dietary glycosides to mutagens by intestinal flora. Proc Natl Acad Sci U S A. 1980 Aug;77(8):4961-5.
Prouillet C, Mazière JC, Mazière C, Wattel A, Brazier M, Kamel S. Stimulatory effect of naturally occurring flavonols quercetin and kaempferol on alkaline phosphatase activity in MG-63 human osteoblasts through ERK and estrogen receptor pathway. Biochem Phar-macol. 2004 Apr 1;67(7):1307-13.
Notoya M, Tsukamoto Y, Nishimura H, Woo JT, Nagai K, Lee IS, Hagiwara H. Quercetin, a flavonoid, inhibits the proliferation, differentiation, and mineralization of osteoblasts in vitro. Eur J Pharmacol. 2004 Feb 6;485(1-3):89-96.
Davey Smith G, Hemani G. Mendelian randomization: genetic anchors for causal in-ference in epidemiological studies. Hum Mol Genet. 2014 Sep 15;23(R1):R89-98.
Lilli C, Biggeri A, Zingaretti C, Vertogen B, Frassineti V, Vespignani R, Grossi V, Florescu C, Matteucci L, Pazzi C, Bongiovanni A, Limarzi F, Fausti V, Bertoni L, Donati C, Galardi F, Gentili N, Mazza F, Martinelli G, Nanni O. Is it possible to conduct clinical trials during a pandemic? The example of a trial of hydroxychloroquine. Epidemiol Prev. 2021 Jan-Apr;45(1-2):28-36. English.
Trajanoska K, Rivadeneira F. Using Mendelian Randomization to Decipher Mechanisms of Bone Disease. Curr Osteoporos Rep. 2018 Oct;16(5):531-540.
Xiong J, Yang L, Deng YQ, Yan SY, Gu JM, Li BH, Zi H, Ming DJ, Zeng XT, Wang YB. The causal association between smoking, alcohol consumption and risk of bladder cancer: A univariable and multivariable Mendelian randomization study. Int J Cancer. 2022 Dec 15;151(12):"2136-2143".
Kamat MA, Blackshaw JA, Young R, Surendran P, Burgess S, Danesh J, Butterworth AS, Staley JR. PhenoScanner V2: an expanded tool for searching human genotype-phenotype associations. Bioinformatics. 2019 Nov 1;35(22):4851-4853.
Hemani G, Tilling K, Davey Smith G. Orienting the causal relationship between impre-cisely measured traits using GWAS summary data. PLoS Genet. 2017 Nov 17;13(11):e1007081.
Wu F, Huang Y, Hu J, Shao Z. Mendelian randomization study of inflammatory bowel disease and bone mineral density. BMC Med. 2020 Nov 10;18(1):312.
Below JE, Gamazon ER, Morrison JV, Konkashbaev A, Pluzhnikov A, McKeigue PM, Parra EJ, Elbein SC, Hallman DM, Nicolae DL, Bell GI, Cruz M, Cox NJ, Hanis CL. Ge-nome-wide association and meta-analysis in populations from Starr County, Texas, and Mexico City identify type 2 diabetes susceptibility loci and enrichment for expression quantitative trait loci in top signals. Diabetologia. 2011 Aug;54(8):2047-55.
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 Jun;21(3):223-42.
Levin MG, Judy R, Gill D, Vujkovic M, Verma SS, Bradford Y; Regeneron Genetics Center, Ritchie MD, Hyman MC, Nazarian S, Rader DJ, Voight BF, Damrauer SM. Genetics of height and risk of atrial fibrillation: A Mendelian randomization study. PLoS Med. 2020 Oct 8;17(10):e1003288.
Gill D, Efstathiadou A, Cawood K, Tzoulaki I, Dehghan A. Education protects against coronary heart disease and stroke independently of cognitive function: evidence from Mendelian randomization. Int J Epidemiol. 2019 Oct 1;48(5):1468-1477.
Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet Epidemiol. 2016 May;40(4):304-14.
Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017 May;32(5):377-389.
Burgess S, Dudbridge F, Thompson SG. Combining information on multiple instru-mental variables in Mendelian randomization: comparison of allele score and summarized data methods. Stat Med. 2016 May 20;35(11):1880-906.
Shim H, Chasman DI, Smith JD, Mora S, Ridker PM, Nickerson DA, Krauss RM, Stephens M. A multivariate genome-wide association analysis of 10 LDL subfractions, and their re-sponse to statin treatment, in 1868 Caucasians. PLoS One. 2015 Apr 21;10(4):e0120758.
Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018 May;50(5):693-698.
Chen J, Bacelis J, Sole-Navais P, Srivastava A, Juodakis J, Rouse A, Hallman M, Teramo K, Melbye M, Feenstra B, Freathy RM, Smith GD, Lawlor DA, Murray JC, Williams SM, Ja-cobsson B, Muglia LJ, Zhang G. Dissecting maternal and fetal genetic effects underlying the associations between maternal phenotypes, birth outcomes, and adult phenotypes: A mendelian-randomization and haplotype-based genetic score analysis in 10,734 moth-er-infant pairs. PLoS Med. 2020 Aug 25;17(8):e1003305.
Chen H, Chen S, Ye H, Guo X. Protective Effects of Circulating TIMP3 on Coronary Ar-tery Disease and Myocardial Infarction: A Mendelian Randomization Study. J Cardiovasc Dev Dis. 2022 Aug 18;9(8):277.
Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, Laurin C, Burgess S, Bowden J, Langdon R, Tan VY, Yarmolinsky J, Shihab HA, Timpson NJ, Evans DM, Relton C, Martin RM, Davey Smith G, Gaunt TR, Haycock PC. The MR-Base platform supports sys-tematic causal inference across the human phenome. Elife. 2018 May 30;7:e34408.
Xu J, Zhang S, Tian Y, Si H, Zeng Y, Wu Y, Liu Y, Li M, Sun K, Wu L, Shen B. Genetic Causal Association between Iron Status and Osteoarthritis: A Two-Sample Mendelian Randomization. Nutrients. 2022 Sep 6;14(18):3683.
Bowden J, Del Greco M F, Minelli C, Davey Smith G, Sheehan N, Thompson J. A framework for the investigation of pleiotropy in two-sample summary data Mendelian randomization. Stat Med. 2017 May 20;36(11):1783-1802.
Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in me-ta-analyses. BMJ. 2003 Sep 6;327(7414):557-60.
Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instru-ments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015 Apr;44(2):512-25.
Xiao Y, Wei R, Yuan Z, Lan X, Kuang J, Hu D, Song Y, Luo J. Rutin suppresses FNDC1 expression in bone marrow mesenchymal stem cells to inhibit postmenopausal osteopo-rosis. Am J Transl Res. 2019 Oct 15;11(10):6680-6690.
Dempster DW, Cosman F, Parisien M, Shen V, Lindsay R. Anabolic actions of parathy-roid hormone on bone. Endocr Rev. 1993 Dec;14(6):690-709.
Smith GD, Ebrahim S. 'Mendelian randomization': can genetic epidemiology contribute to understanding environmental determinants of disease? Int J Epidemiol. 2003 Feb;32(1):1-22.
Cornish AJ, Tomlinson IPM, Houlston RS. Mendelian randomisation: A powerful and inexpensive method for identifying and excluding non-genetic risk factors for colorectal cancer. Mol Aspects Med. 2019 Oct;69:41-47.

No competing interests reported.

SupplementaryTableS1.docx

Genetically determined watercress intake and risk for bone mineral density

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Statistical Analysis

Results

Discussion

Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1