Joëlle Schutten


Health aspects of magnesium status and supplementation Joëlle C. Schutten

Health aspects of magnesium status and supplementation Financial support for this thesis was generously provided by Nedmag B.V, the University of Groningen and the Graduate School of Medical Sciences/University Medical Center Groningen. Cover design: Joëlle C. Schutten Lay-out: Publiss | Print: Ridderprint | © Copyright 2022: Joëlle C. Schutten, The Netherlands All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, by photocopying, recording, or otherwise, without the prior written permission of the author.

Health aspects of magnesium status and supplementation Proefschrift ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen op gezag van de rector magnificus prof. dr. C. Wijmenga en volgens besluit van het College voor Promoties. De openbare verdediging zal plaatsvinden op maandag 21 november 2022 om 11.00 uur door Joëlle Catharina Schutten geboren op 31 oktober 1992 te Schoonebeek

Promotores Prof. dr. S.J.L. Bakker Prof. dr. M.H. de Borst Beoordelingscommissie Prof. dr. G.J. Navis Prof. dr. I.P. Kema Prof. dr. G. Rimbach

Paranimfen R.M. Douwes Y. van der Veen

Contents Chapter 1 General Introduction 9 Chapter 2 Lower plasma magnesium, measured by nuclear magnetic resonance spectroscopy, is associated with increased risk of developing type 2 diabetes mellitus in women: Results from a Dutch prospective cohort study 23 J Clin Med 2019; 8: 169 Chapter 3 Comparison of two methods for the assessment of intraerythrocyte magnesium and its determinants: Results from the LifeLines cohort study 47 Clin Chim Acta 2020; 510: 772-780 Chapter 4 Magnesium and blood pressure: A physiology-based approach 77 Adv Chronic Kidney Dis 2018; 25: 244-250 Chapter 5 Long-term magnesium supplementation improves glucocorticoid metabolism: A post-hoc analysis of an intervention trial 93 Clin Endocrinol 2021; 94 :150–157 Chapter 6 Effects of magnesium citrate, magnesium oxide and magnesium sulfate on arterial stiffness: A randomized double-blind placebocontrolled intervention trial 111 J Am Heart Assoc 2022; 11: e021783 Chapter 7 General Discussion 157 Nederlandse samenvatting 171 Dankwoord 177 About the author 183 List of publications 187


General Introduction

Chapter 1 10 Food and nutrition surveys in the United States and Europe consistently show that dietary magnesium intake is insufficient. In the US, dietary magnesium intake is below the recommended daily allowance (RDA) and the estimated average requirement (EAR) in almost all age and sex groups 1. In fact, approximately 50% of the US population consumes less than the EAR of magnesium 2. A more recent study among post-menopausal women showed that approximately 75% had a magnesium intake below the RDA 3. Among Dutch adults, the proportion with intakes below the EAR is between 16 and 35% 4. Low consumption of vegetables, decreased magnesium content in vegetables compared to historic contents 5 and increased consumption of processed foods, which contain low amounts of magnesium have been identified as important contributors. Early signs of lowmagnesium intake include fatigue and muscle cramps 6,7. Lowmagnesium intake may also lead to an increased risk of cardiovascular disease (CVD) 8. As CVD is the leading cause of death globally 9, preventive measures to lower the risk of CVD are needed and may include adjustments in magnesium intake. At the same time, large well-designed randomized controlled trials (RCTs) addressing the potential effects of increased dietary magnesium intake (by means of supplementation) on cardiovascular health outcomes are lacking.To date, several clinical studies investigated effects of magnesium supplementation on blood pressure, however, showing inconclusive results 10–14. It has been suggested that these inconclusive results are the consequence of large differences in study populations, large differences in study durations and different magnesium formulations that were used. Interestingly, a meta-analysis that summarized the effect of magnesium supplementation on blood pressure found a reduction in blood pressure in a dose- and time-dependent manner, suggesting that studies with higher magnesium dosages and longer durations are more effective 15. Because magnesium has been suggested to antagonize calcification of soft tissues, including the vascular wall, from the perspective of vascular health magnesium supplementation may not only be of interest for reducing blood pressure, but also for improving arterial stiffness by reducing vascular calcification. The effect of magnesium supplementation on arterial stiffness, which is a validated marker of cardiovascular health 16, has rarely investigated, but Joris et al. recently found a significant reduction in arterial stiffness after 24 weeks of magnesium citrate supplementation 17. Magnesium in foods and supplements Magnesium can either be ingested via foods or via dietary supplements that contain magnesium. Foods that are high in magnesium include grains, (green) vegetables, nuts, and dairy products. When consuming an average portion of 200 g of spinach, the total

General Introduction 11 1 magnesium ingested is 156 mg, which is about 52% of the total daily requirement for women and 45% for men, respectively. Given the presumed health effects and the observed low intake of magnesium with a normal contemporary diet, an increasing interest in dietary magnesium supplements has emerged over the past decades. Results from a recent study showed that magnesium was the most popular topic when it comes to dietary supplements among global Google users and was even more popular than vitamin D and iron 18. Numerous magnesium supplements are commercially available, of which the organic formulation magnesium citrate is most commonly preferred because of its suggested high bioavailability. Inorganic formulations, such as magnesium oxide and magnesium sulfate, are often considered less desirable because of their poorer solubility in water 19. To date, several studies investigated the bioavailability of different magnesium formulations, in which in vitro solubility and/or in vivo gastro-intestinal absorbability were tested. Three studies have suggested a higher bioavailability of the magnesium citrate formulation based on a higher renal excretion of magnesium 20,21 or higher serum magnesium levels 22 when compared to other inorganic compounds. The opposite has, however, been suggested by a recent meta-analysis, which found that inorganic magnesium, particularly magnesium oxide, exhibited a greater increase in serum magnesium compared to organic magnesium formulations 15. As a consequence, no consensus has been reached with respect to the best magnesium formulation to be used. Magnesium homeostasis and status Magnesium homeostasis mainly depends on the interplay between intestinal absorption and renal excretion. The predominant sites of intestinal absorption are the distal jejunum and the ileum, via a paracellular pathway, whereas a small part is absorbed in the colon,which mainly involves a transcellular pathway 23. Under normal circumstances, approximately 30-50% of the magnesium is absorbed in the intestine. However, during a period of low magnesium intake, fractional absorption can reach values of as high as 80-90%, whereas it decreases significantly to values as low as 10-20% when intake is high. In the kidney, the vast majority of magnesium is reabsorbed in the loop of Henle, mainly in the thick ascending limb. Renal magnesium handling is regulated by several magnesium transporters, including the transient receptor potential channel melastatin member 6 (TRPM6) 24. The body is able to maintain normal plasma levels by increasing the intestinal absorption and decreasing renal excretion when low intakes occur. It is worth noting that the plasma level of magnesium is not representative of the total body magnesium, as magnesium is mainly located in cells and bones and the plasma

Chapter 1 12 level is highly regulated. Although it poorly represents total body magnesium, many observational studies have focused on the plasma or serum total concentration and found weak to moderate associations with CVD and hypertension 25,26. The impact of magnesium deficiency, a state in which whole body magnesium content is low, may therefore be grossly underestimated 24. Magnesium and health outcomes Type 2 diabetes (T2D) The global prevalence of T2D is rapidly growing. In fact, it was estimated in 2019 that approximately 460 million (19.3%) adults aged 20-79 years suffered from T2D 27. Patients with T2D have a two-fold higher risk of CVD mortality 28. Since 1947, T2D has been linked to the incidence of hypomagnesaemia 29, which is defined as a plasma magnesium concentration below 0.70 mmol/L. Although T2D has been associated with low plasma magnesium concentrations, subjects with T2D do not necessarily consume less magnesium than subjects without T2D 30. Hyperglycemia, on the other hand, leads to increased urinary magnesium loss, which in turn can result in decreased plasma magnesium concentrations. Because of this, magnesium supplementation is sometimes recommended in T2D patients with magnesium deficiency 25,31. Although hypomagnesaemia may be a consequence of T2D, several studies have suggested that hypomagnesaemia might also be a risk factor for developing T2D 32,33, in which the mechanistic pathway may at least in part be mediated through insulin resistance. In this respect, these studies investigated the prospective association of total serum/ plasma magnesium with T2D. To date, prospective associations of ionized plasma magnesium, the biologically active fraction, with risk of developing T2D have been poorly documented. Blood pressure Hypertension is currently one of the strongest risk factors for CVD. Hypertension, defined as a systolic blood pressure of ≥ 140 mmHg and/or a diastolic blood pressure of ≥ 90 mmHg, is a global public health problem, being responsible for 12.8% of all deaths worldwide 34. Risk factors for hypertension include, among others, age, race, family history of hypertension, obesity, physical inactivity, smoking, excessive intake of sodium, as well as insufficient dietary intake of potassium and magnesium 35–37. Epidemiological studies reported dose-response associations between dietary magnesium intake and risk of hypertension 26. In a large prospective cohort study, Joosten et al. showed a

General Introduction 13 1 lower risk of developing hypertension among individuals with a higher 24-h urinary magnesium excretion 38. To date, some, but not all randomized controlled trials (RCTs) have shown small effects of oral magnesium supplementation on both systolic and diastolic blood pressure 15,39,40, with greater blood pressure effects in studies using higher dosages of magnesium. Arterial stiffness Arterial stiffness causes dysfunction of the large arteries and is strongly associated with increased risk of coronary heart disease and stroke 41. It is one of the earliest detectable changes in vascular structure and function 42.Arterial stiffness is a non-invasive vascular functionmarker and can bemeasured bymeans of carotid-to-femoral pulsewave velocity (c-fPWV), which is currently the gold standard for the quantification of arterial stiffness 43. c-fPWV is calculated as the time it takes a pulse wave to travel from the carotid to the femoral site divided by the distance. Ageing is a non-modifiable risk factor for arterial stiffness as the large arteries lose their elasticity over time 44, whereas smoking, alcohol use, physical inactivity, and an unhealthy diet are modifiable risk factors 45. One of the most important causes of arterial stiffness is vascular calcification. In-vitro studies have suggested that a high dietary magnesium diet may prevent vascular calcification 46,47, which may in turn improve arterial stiffness. RCTs that addressed effects of oral magnesium supplementation on arterial stiffness are scarce and show inconclusive results 10,13,17. Joris et al. showed a significant improvement of arterial stiffness by 1.0 m/s following oral magnesium supplementation in healthy overweight and slightly obese adults using a total daily magnesium dose of 350 mg/d 17. However, the study was based on a single comparison between magnesium citrate and placebo, and thus, it remains unknown whether the effect was induced by magnesium itself or by the anion citrate. Furthermore, no head-to-head comparison between various magnesium formulations in terms of effects of arterial stiffness has been performed. In this respect, a comparison between magnesium citrate and other formulations might be of interest to investigate whether the effect was induced by magnesium itself or by the anion citrate and whether other formulations have similar effects. Glucocorticoid metabolism Physiological stress is a major, yet modifiable risk factor for CVD 48. During the stress response, the hypothalamic-pituitary-adrenal (HPA) axis is activated, which results in the secretion of cortisol, a glucocorticoid hormone. Elevated cortisol levels cause

Chapter 1 14 increased blood pressure and suppression of immune system and inflammatory reactions. 11β-hydroxysteroid dehydrogenases (11β-HSDs) are enzymes that catalyze the conversion of active cortisol to inert cortisone and vice versa 49,50. Disturbed activity of these enzymes has been linked to hypertension and insulin resistance, which have been implicated to adversely affect cardiovascular health 51,52. Enzyme activity of 11β-HSDs can be quantified by the ratios of 24-h urinary cortisol, cortisone, and their metabolites. Several preclinical studies have shown a relationship between a magnesium deficient diet and a disturbed glucocorticoid metabolism 53–55. Whether oral magnesium supplementation can modify glucocorticoid metabolism in humans is unknown. OUTLINE OF THE THESIS Previous research mainly focused on total plasma or serum magnesium as a marker of magnesium status. In addition, RCTs addressing potential effects of magnesium supplementation on cardiovascular end points mainly focused on blood pressure, while effects on arterial stiffness have rarely been studied. This thesis aims to compare several analytic methods for the assessment of magnesium status and to investigate health aspects of magnesium status and magnesium supplementation, with emphasis on T2D, glucocorticoid metabolism, blood pressure, and arterial stiffness. Under physiological conditions, magnesium homeostasis is maintained by adjusting urinary excretion equaling the net uptake of magnesium in the gut, with plasma values ranging within the normal range of 0.70 - 1.00 mmol/L 56. Of the total amount of circulating magnesium, 70-80% is available in the free, ionized form, whereas the rest is bound to proteins, citrate, phosphate, and other compounds. So far, most epidemiological studies that focused on the relationship between low total plasma or serum magnesium and disease risk reported weak or absent associations 38,57–59, whereas strong associations of 24-h urinary magnesium excretion, a marker of intestinal magnesium uptake, with hypertension and CVD risk have been reported repeatedly 38,57. Furthermore, total plasma magnesium seems to correlate weakly with 24-h urinary magnesium excretion. Plasma ionized magnesium, on the other hand, may have additional clinical implications, since this form is biologically active and may therefore better reflect the intracellular concentration 60. In fact, several studies have suggested that the measurement of plasma ionized magnesium is superior to the measurement of total plasma magnesium 60–63, although no proper method comparison between plasma ionized and total plasma magnesium has been performed so far 62,63. In addition, not

General Introduction 15 1 much is known about potential health aspects of low plasma ionized magnesium levels. Therefore, in Chapter 2, we study the performance of a nuclear magnetic resonance (NMR)-based assay, that quantifies ionized magnesium in EDTA plasma samples and we prospectively examine the association of plasma magnesium with risk of developing T2D in the Prevention of Renal and Vascular End-stage Disease (PREVEND) study using data from 5747 subjects. Also, there is increasing interest in the measurement of intracellular magnesium concentrations, as a significant amount of the total body magnesium is located in cells. Particularly, the measurement of magnesium in white and red blood cells has become more popular, because these cells are relatively easy to obtain 64. The relevance of intra-erythrocyte measurement has been further emphasized by a previous study showing that the prevalence of hypomagnesemia based on the intra-erythrocyte level was high among geriatric patients, whereas these patients had normal total plasma magnesium concentrations 65. Unfortunately, no simple and rapid technique exists to measure intracellular magnesium concentrations. A laborious direct method has been established to measure magnesium concentrations in erythrocytes. A less laboriously indirect method can calculate intra-erythrocyte magnesium from whole blood magnesium and plasma magnesium by taking into account the hematocrit and is therefore much easier to apply. Whether this method is representative of the intra-erythrocyte magnesium concentration is not well known. Deuster et al. were the first to report a method comparison between a direct and an indirect method for the assessment of intra-erythrocyte magnesium 66, however, using only 10 samples. A larger sample is therefore required in order to perform a valid method comparison. In Chapter 3, we compare a direct and an indirect method for the assessment of intra-erythrocyte magnesium and additionally explore determinants of intra-erythrocyte magnesium using data from the LifeLines cohort study. As magnesium is mainly located in cells, and due to its proposed role in T2D, it was hypothesized that intra-erythrocyte magnesium correlates with T2D markers, such as fasting glucose and HbA1c. Several clinical trials already reported small, yet inconclusive effects on blood pressure following magnesium supplementation. The exact mechanism by which magnesium potentially lowers blood pressure is unknown. In Chapter 4, we review available data on potential mechanisms linking magnesium with blood pressure. Here, we use a physiology-based approach, structuring our analyses based on blood pressure and its immediate determinants, i.e. cardiac output and peripheral resistance.

Chapter 1 16 As alluded to above, cortisol excess may affect cardiovascular health. To gain more mechanistic insight, we investigate the effects of oral magnesium supplementation on glucocorticoid metabolism in Chapter 5, as a potential underlying mechanism by which increased dietary magnesium intake may beneficially affect cardiovascular health. To do so, we perform post-hoc analyses using data from the previous RCT on magnesium citrate supplementation that we performed in collaboration with researchers from the University of Maastricht 17. In Chapter 6, we investigate the effects of oral magnesium supplementation on arterial stiffness in an RCT of 164 healthy overweight and slightly obese Dutch adults. Our aim is two-fold: i) to replicate the effect of oral magnesium citrate supplementation on arterial stiffness that was found in the previous study 17 and ii) to examine whether other commercially available inorganic magnesium supplements, including magnesium oxide and magnesium sulfate, are non-inferior to effects of magnesium citrate supplements on arterial stiffness. We additionally assess effects on blood pressure, plasma and urinary magnesium, and gastro-intestinal tolerability.

General Introduction 17 1 References 1. Ford ES, Mokdad AH. Dietary magnesium intake in a national sample of US adults. J Nutr. 2003;133(9):2879-2882. 2. Moshfegh A, Goldman J, Ahuja J, Rhodes D, LaComb R.What we eat in America, NHANES 20052006. US Dep Agric. 2009. 3. Wu WC, Huang M, Taveira TH, et al. Relationship between dietary magnesium intake and incident heart failure among older women: The whi. J Am Heart Assoc. 2020. doi:10.1161/ JAHA.119.013570 4. van Rossum CTM, Fransen H, Verkaik-Kloosterman J, Buurma-Rethans E, Ocke M. National Food Consumption Survey: Dutch National Food Consumption Survey 2007-2010.; 2011. 5. Worthington V. Nutritional Quality of Organic Versus Conventional Fruits, Vegetables, and Grains. J Altern Complement Med. 2001;7(2):161-173. doi:10.1089/107555301750164244 6. DiNicolantonio JJ, O’Keefe JH, Wilson W. Subclinical magnesium deficiency: A principal driver of cardiovascular disease and a public health crisis. Open Hear. 2018. doi:10.1136/ openhrt-2017-000668 7. Bilbey DLJ, Prabhakaran VM. Muscle cramps and magnesium deficiency: Case reports. Can Fam Physician. 1996. 8. Dinicolantonio JJ, Liu J, O’Keefe JH. Magnesium for the prevention and treatment of cardiovascular disease. Open Hear. 2018. doi:10.1136/openhrt-2018-000775 9. Roth GA, Johnson C, Abajobir A, et al. Global, Regional, and National Burden of Cardiovascular Diseases for 10 Causes, 1990 to 2015. J Am Coll Cardiol. 2017. doi:10.1016/j.jacc.2017.04.052 10. Cosaro E, Bonafini S, Montagnana M, et al. Effects of magnesium supplements on blood pressure, endothelial function and metabolic parameters in healthy young men with a family history of metabolic syndrome. Nutr Metab Cardiovasc Dis. 2014;24(11):1213-1220. doi:10.1016/j.numecd.2014.05.010 11. AfitskaK,Clavel J,KistersK,Vormann J,WernerT.Magnesiumcitrate supplementationdecreased blood pressure and HbA1c in normomagnesemic subjects with metabolic syndrome: a 12-week, placebo-controlled, double-blinded pilot trial. Magnes Res. 2021;34(3):130-139. doi:10.1684/mrh.2021.0489 12. Rodríguez-Morán M, Simental-Mendía LE, Gamboa-Gómez CI, Guerrero-Romero F. Oral Magnesium Supplementation and Metabolic Syndrome: A Randomized Double-Blind Placebo-Controlled Clinical Trial. Adv Chronic Kidney Dis. 2018;25(3):261-266. doi:10.1053/j. ackd.2018.02.011 13. Cunha AR,D’El-Rei J,Medeiros F, et al.Oral magnesium supplementation improves endothelial function and attenuates subclinical atherosclerosis in thiazide-treated hypertensive women. J Hypertens. 2017;35(1):89-97. doi:10.1097/HJH.0000000000001129 14. Rodríguez-Morán M, Guerrero-Romero F. Oral magnesium supplementation improves insulin sensitivity and metabolic control in type 2 diabetic subjects: A randomized double-blind controlled trial. Diabetes Care. 2003. doi:10.2337/diacare.26.4.1147 15. Zhang X, Li Y, Del Gobbo LC, et al. Effects of Magnesium Supplementation on Blood Pressure: A Meta-Analysis of Randomized Double-Blind Placebo-Controlled Trials. Hypertens (Dallas, Tex 1979). 2016;68(2):324-333. doi:10.1161/HYPERTENSIONAHA.116.07664 16. Cohn JN, Quyyumi AA, Hollenberg NK, Jamerson KA. Surrogate markers for cardiovascular disease: functional markers. Circulation. 2004;109(25 Suppl 1):IV31-46. doi:10.1161/01. CIR.0000133442.99186.39

Chapter 1 18 17. Joris PJ, Plat J, Bakker SJ, Mensink RP. Long-term magnesium supplementation improves arterial stiffness in overweight and obese adults: results of a randomized, double-blind, placebo-controlled intervention trial. Am J Clin Nutr. 2016;103(5):1260-1266. doi:10.3945/ ajcn.116.131466 18. Kamiński M, Kręgielska-Narożna M, Bogdański P. Determination of the popularity of dietary supplements using google search rankings. Nutrients. 2020. doi:10.3390/nu12040908 19. Blancquaert L, Vervaet C, Derave W. Predicting and testing bioavailability of magnesium supplements. Nutrients. 2019. doi:10.3390/nu11071663 20. Kappeler D, Heimbeck I, Herpich C, et al. Higher bioavailability of magnesium citrate as compared to magnesium oxide shown by evaluation of urinary excretion and serum levels after single-dose administration in a randomized cross-over study. BMC Nutr. 2017;3(1):7. doi:10.1186/s40795-016-0121-3 21. Lindberg JS, Zobitz MM, Poindexter JR, Pak CYC. Magnesium bioavailability from magnesium citrate and magnesium oxide. J Am Coll Nutr. 1990;9(1):48-55. doi:10.1080/07315724.1990. 10720349 22. Walker AF, Marakis G, Christie S, Byng M. Mg citrate found more bioavailable than other Mg preparations in a randomised, double-blind study. Magnes Res. 2003;16(3):183-191. 23. Voets T, Nilius B, Hoefs S, et al. TRPM6 Forms the Mg2+ Influx Channel Involved in Intestinal and Renal Mg2+ Absorption. J Biol Chem. 2004. doi:10.1074/jbc.M311201200 24. de Baaij JHF, Hoenderop JGJ, Bindels RJM. Magnesium in man: implications for health and disease. Physiol Rev. 2015;95(1):1-46. doi:10.1152/physrev.00012.2014 25. Fang X,Wang K, Han D, et al. Dietary magnesium intake and the risk of cardiovascular disease, type 2 diabetes, and all-cause mortality: a dose-response meta-analysis of prospective cohort studies. BMC Med. 2016;14(1):210. doi:10.1186/s12916-016-0742-z 26. Han H, Fang X, Wei X, et al. Dose-response relationship between dietary magnesium intake, serum magnesium concentration and risk of hypertension: a systematic review and metaanalysis of prospective cohort studies. Nutr J. 2017;16(1):26. doi:10.1186/s12937-017-0247-4 27. Internation Diabetes Federation. International Diabetes Federation; Diabetes Atlas Ninth Edition, 2019.; 2019. 28. Morrish NJ, Wang SL, Stevens LK, Fuller JH, Keen H. Mortality and causes of death in the WHO multinational study of vascular disease in diabetes. Diabetologia. 2001. doi:10.1007/ PL00002934 29. Martin HE,Wertman M. SERUM POTASSIUM,MAGNESIUM,AND CALCIUM LEVELS IN DIABETIC ACIDOSIS. J Clin Invest. 1947. doi:10.1172/JCI101799 30. McClure ST, Schlechter H, Oh S, et al. Dietary intake of adults with and without diabetes: results from NHANES 2013–2016. BMJ Open Diabetes Res Care. 2020;8(1):e001681. doi:10.1136/bmjdrc-2020-001681 31. Garber A, Blackard WG, Feinglos M, et al. Magnesium Supplementation in the Treatment of Diabetes. Diabetes Care. 1992;15(8):1065-1067. doi:10.2337/diacare.15.8.1065 32. Gommers LMM, Hoenderop JGJ, Bindels RJM, de Baaij JHF. Hypomagnesemia in Type 2 Diabetes: A Vicious Circle? Diabetes. 2016;65(1):3-13. doi:10.2337/db15-1028 33. van der Burgh AC, Moes A, Kieboom BCT, et al. Serum magnesium, hepatocyte nuclear factor 1β genotype and post-transplant diabetes mellitus: a prospective study. Nephrol Dial Transplant. 2019. doi:10.1093/ndt/gfz145

General Introduction 19 1 34. World Health Organization (WHO). A global brief on hypertension. bitstream/10665/79059/1/WHO_DCO_WHD_2013.2_eng.pdf. Published 2013. Accessed August 3, 2017. 35. Houston MC, Harper KJ. Potassium, magnesium, and calcium: their role in both the cause and treatment of hypertension. J Clin Hypertens (Greenwich). 2008;10(7 Suppl 2):3-11. 36. Forman JP, Stampfer MJ, Curhan GC. Diet and lifestyle risk factors associated with incident hypertension in women. JAMA - J Am Med Assoc. 2009. doi:10.1001/jama.2009.1060 37. Kieneker LM, Gansevoort RT, Mukamal KJ, et al. Urinary potassium excretion and risk of developing hypertension: the prevention of renal and vascular end-stage disease study. Hypertens (Dallas,Tex 1979).2014;64(4):769-776.doi:10.1161/HYPERTENSIONAHA.114.03750 38. Joosten MM, Gansevoort RT, Mukamal KJ, et al. Urinary magnesium excretion and risk of hypertension: the prevention of renal and vascular end-stage disease study. Hypertens (Dallas, Tex 1979). 2013;61(6):1161-1167. doi:10.1161/HYPERTENSIONAHA.113.01333 39. Kass L, Weekes J, Carpenter L. Effect of magnesium supplementation on blood pressure: a meta-analysis. Eur J Clin Nutr. 2012;66(4):411-418. doi:10.1038/ejcn.2012.4 40. Jee SH, Miller ER 3rd, Guallar E, Singh VK, Appel LJ, Klag MJ. The effect of magnesium supplementation on blood pressure: a meta-analysis of randomized clinical trials. Am J Hypertens. 2002;15(8):691-696. 41. Mattace-Raso FUS, van der Cammen TJM, Hofman A, et al. Arterial Stiffness and Risk of Coronary Heart Disease and Stroke. Circulation. 2006;113(5):657-663. doi:10.1161/ CIRCULATIONAHA.105.555235 42. Tanaka H.Various Indices of Arterial Stiffness: Are They Closely Related or Distinctly Different? Pulse. 2017. doi:10.1159/000461594 43. Laurent S, Cockcroft J, Van Bortel L, et al. Expert consensus document on arterial stiffness: methodological issues and clinical applications. Eur Heart J. 2006;27(21):2588-2605. doi:10.1093/eurheartj/ehl254 44. Tesauro M, Mauriello A, Rovella V, et al. Arterial ageing: from endothelial dysfunction to vascular calcification. J Intern Med. 2017;281(5):471-482. doi:10.1111/joim.12605 45. Angoff R, Mosarla RC, Tsao CW. Aortic Stiffness: Epidemiology, Risk Factors, and Relevant Biomarkers. Front Cardiovasc Med. 2021;8. doi:10.3389/fcvm.2021.709396 46. ter Braake AD, Tinnemans PT, Shanahan CM, Hoenderop JGJ, de Baaij JHF. Magnesium prevents vascular calcification in vitro by inhibition of hydroxyapatite crystal formation. Sci Rep. 2018;8(1):2069. doi:10.1038/s41598-018-20241-3 47. ter Braake AD, Smit AE, Bos C, et al. Magnesium prevents vascular calcification in Klotho deficiency. Kidney Int. 2020. doi:10.1016/j.kint.2019.09.034 48. Steptoe A, Kivimäki M. Stress and cardiovascular disease. Nat Rev Cardiol. 2012. doi:10.1038/ nrcardio.2012.45 49. Tomlinson JW, Walker EA, Bujalska IJ, et al. 11β-Hydroxysteroid Dehydrogenase Type 1: A Tissue-Specific Regulator of Glucocorticoid Response. Endocr Rev. 2004;25(5):831-866. doi:10.1210/er.2003-0031 50. Edwards CR, Stewart PM, Burt D, et al. Localisation of 11 beta-hydroxysteroid dehydrogenase- -tissue specific protector of the mineralocorticoid receptor. Lancet (London, England). 1988;2(8618):986-989. doi:10.1016/s0140-6736(88)90742-8

Chapter 1 20 51. Ferrari P. The role of 11β-hydroxysteroid dehydrogenase type 2 in human hypertension. Biochim Biophys Acta - Mol Basis Dis. 2010. doi:10.1016/j.bbadis.2009.10.017 52. Schnackenberg CG, Costell MH, Krosky DJ, et al. Chronic inhibition of 11 β -hydroxysteroid dehydrogenase type 1 activity decreases hypertension, insulin resistance, and hypertriglyceridemia in metabolic syndrome. Biomed Res Int. 2013. doi:10.1155/2013/427640 53. Takaya J, Iharada A, Okihana H, Kaneko K. Down-regulation of hepatic phosphoenolpyruvate carboxykinase expression in magnesium-deficient rats. Magnes Res. 2012;25(3):131-139. doi:10.1684/mrh.2012.0321 54. Takaya J, Iharada A, Okihana H, Kaneko K. Magnesium deficiency in pregnant rats alters methylation of specific cytosines in the hepatic hydroxysteroid dehydrogenase-2 promoter of the offspring. Epigenetics. 2011;6(5):573-578. doi:10.4161/epi.6.5.15220 55. Thomas AE, Inagadapa PJN, Jeyapal S, Merugu NM, Kalashikam RR, Manchala R. Maternal Magnesium Restriction Elevates Glucocorticoid Stress and Inflammation in the Placenta and Fetus of WNIN Rat Dams. Biol Trace Elem Res. 2018;181(2):281-287. doi:10.1007/s12011-0171058-3 56. Blaine J, Chonchol M, Levi M. Renal control of calcium, phosphate, and magnesium homeostasis. Clin J Am Soc Nephrol. 2015. doi:10.2215/CJN.09750913 57. Joosten MM, Gansevoort RT, Mukamal KJ, et al. Urinary and plasma magnesium and risk of ischemic heart disease. Am J Clin Nutr. 2013;97(6):1299-1306. doi:10.3945/ajcn.112.054114 58. Chiuve SE, Korngold EC, Januzzi JLJ, Gantzer M Lou, Albert CM. Plasma and dietary magnesium and risk of sudden cardiac death in women. Am J Clin Nutr. 2011;93(2):253-260. doi:10.3945/ ajcn.110.002253 59. Akarolo-Anthony SN, Jiménez MC, Chiuve SE, Spiegelman D, Willett WC, Rexrode KM. Plasma magnesium and risk of ischemic stroke among women. Stroke. 2014. doi:10.1161/ STROKEAHA.114.005917 60. Huijgen HJ, Van Ingen HE, Kok WT, Sanders GTB. Magnesium fractions in serum of healthy individuals and CAPD patients, measured by an ion-selective electrode and ultrafiltration. Clin Biochem. 1996. doi:10.1016/0009-9120(96)84729-B 61. Ordak M,Maj-Zurawska M,Matsumoto H, et al. Ionized magnesium in plasma and erythrocytes for the assessment of low magnesium status in alcohol dependent patients. Drug Alcohol Depend. 2017. doi:10.1016/j.drugalcdep.2017.04.035 62. Saris NE, Mervaala E, Karppanen H, Khawaja JA, Lewenstam A. Magnesium. An update on physiological, clinical and analytical aspects. Clin Chim Acta. 2000;294(1-2):1-26. 63. Zhan J,Wallace TC, Butts SJ, et al. Circulating ionized magnesium as a measure of supplement bioavailability: Results from a pilot study for randomized clinical trial. Nutrients. 2020. doi:10.3390/nu12051245 64. Arnaud MJ. Update on the assessment of magnesium status. Br J Nutr. 2008;99(S3):S24-S36. doi:10.1017/S000711450800682X 65. Ulger Z, Ariogul S, Cankurtaran M, et al. Intra-erythrocyte magnesium levels and their clinical implications in geriatric outpatients. J Nutr Heal Aging. 2010. doi:10.1007/s12603-010-0121-y 66. Deuster PA, Trostmann UH, Bernier LL, Dolev E. Indirect vs direct measurement of magnesium and zinc in erythrocytes. Clin Chem. 1987;33(4):529-532.

General Introduction 21 1


Lower plasma magnesium,measured by nuclear magnetic resonance spectroscopy, is associated with increased risk of developing type 2 diabetes mellitus in women: Results from a Dutch prospective cohort study Joëlle C. Schutten António W. Gomes-Neto Gerjan Navis Ron T. Gansevoort Robin P.F. Dullaart Jenny. E. Kootstra-Ros Richard M. Danel Frans Goorman Rijk O.B. Gans Martin H. de Borst Elias J. Jeyarajah Irina Shalaurova James. D. Otvos Margery A. Connelly Stephan J.L. Bakker J Clin Med 2019; 8: 169

Chapter 2 24 Abstract Background Low circulating magnesium (Mg) is associated with increased risk of developing type 2 diabetes mellitus (T2D). We aimed to study the performance of a nuclear magnetic resonance (NMR)-based assay that quantifies ionized Mg in EDTA plasma samples and prospectively investigate the association of Mg with risk of T2D. Methods The analytic performance of an NMR-based assay for measuring plasma Mg was evaluated. We studied 5747 subjects free of T2D at baseline in the Prevention of Renal and Vascular End-stage Disease (PREVEND) study. Results Passing-Bablok regression analysis, comparing NMR-measured ionized Mg with total Mg measured by the Roche colorimetric assay, produced a correlation of r=0.90, with a slope of 1.08 (95% CI: 1.00-1.13) and an intercept of 0.02 (95% CI: -0.02-0.08). During median follow-up period of 11.2 (IQR: 7.7-12.0) years, 289 (5.0%) participants developed T2D. The association of NMR-measured ionized Mg with T2D risk was modified by sex (Pinteraction=0.007). In women, we found an inverse association between Mg and risk of developing T2D, independent of adjustment for potential confounders (HR: 1.80; 95% CI:1.20-2.70). In men, we found no association between Mg and risk of developing T2D (HR: 0.90; 95%: 0.67-1.21). Conclusion Lower NMR-measured plasma ionized Mg was independently associated with a higher risk of developing T2D in women, but not in men.

Lower plasma magnesium, measured by nuclear magnetic resonance spectroscopy, is associated with increased risk of developing type 2 diabetes mellitus in women: Results from a Dutch prospective cohort study 25 2 Introduction The global prevalence of type 2 diabetes mellitus (T2D) has increased over the past few decades 1, and certain modifiable risk factors, including obesity and insulin resistance as well as inadequate intake of vitamins and minerals, have received considerable interest 2,3. Magnesium (Mg) is an essential cofactor for multiple enzymatic pathways involved in energy metabolism and modulation of insulin-mediated glucose uptake 4 and has been associated with inflammation and endothelial dysfunction 5,6. Not surprisingly, Mg levels have been linked to several cardiovascular diseases, including ischemic heart disease, stroke and hypertension, but also to T2D 7–13. Nearly 99% of the magnesium in the body is found in the bone, muscle and soft tissues 14,15. Only about 0.3-1% is present in serum, with a mean Mg concentration of nearly 0.85 mmol/L. Of this 70-80% is available in the free ionized form and the rest is bound to proteins, phosphate, citrate and other compounds. In current clinical laboratories, Mg is measured largely as total Mg with the predominant techniques being: 1) photometry, which uses a number of chromogenic substances such as xylidyl blue, and 2) atomic absorption spectroscopy 15. Determination of ionized Mg has been problematic and ion-selective electrodes for measuring ionized Mg potentiometrically have historically suffered from a lack of selectivity as well as relatively long response times. In recent years, efforts have been underway to optimize measurement of ionized Mg in plasma and serum due to numerous publications promoting the relevance of ionized Mg in different clinical situations and potential superiority of ionized Mg over total Mg concentrations 15. Recently, a clinical nuclear magnetic resonance spectroscopy (NMR) instrument (Vantera® Clinical Analyzer) was developed that addresses the limiting factors of research NMR instruments and allows for the simultaneous quantification of lipoprotein particles, metabolites and an inflammatory marker in the clinical laboratory 16–19. The aim of the current study was to develop and validate an assay for quantifying ionized Mg in plasma using NMR spectra collected for routine lipoprotein quantification on a clinical laboratory instrument. In this way, Mg can be measured in addition to routine lipoprotein quantification without incurring extra costs. With this newly developed NMR-based assay, we further aimed to determine the prospective association of NMRmeasured Mg and the risk of developing T2D in a large Dutch cohort study.

Chapter 2 26 Materials and methods Study design For the analyses of the present study, the Prevention of Renal and Vascular Endstage Disease (PREVEND) study was used, which is a prospective Dutch cohort. Details are described elsewhere 20. In brief, from 1997 to 1998, all inhabitants of Groningen, the Netherlands, aged 28 to 75 years (N=85,421), were sent a short questionnaire on demographic characteristics and renal and cardiovascular morbidity and a vial to collect a first morning void urine sample. Those who were unable or unwilling to participate, pregnant women, and individuals using insulin were not allowed to participate. Altogether, 40,856 people (48%) responded. Subjects with a urinary albumin concentration of ≥10 mg/L (n=7768) were invited to participate, of whom 6000 subjects were enrolled. In addition, a randomly selected group with a urinary albumin concentration of <10 mg/L (n=3394) was invited to participate in the cohort and 2592 subjects of the initially 3394 invited subjects were enrolled. 8592 subjects participated to the PREVEND cohort and completed an extensive examination in 1997 and 1998 (baseline). Participants were invited to the outpatient clinic of the University Medical Center Groningen for measurements approximately every 3 years. The second screening took place from 2001 through 2003 (n=6894), which was the starting point of the presQnt evaluation. For the present study,we excluded subjects with diabetes at baseline or unknown diabetes status or with no follow-up data available for diabetes (n=545) and subjects with missing Mg data (n=602), leaving 5747 participants for the analyses (Figure 1). The PREVEND study was approved by the Medical Ethics Committee of the University Medical Center Groningen. Written informed consent was obtained from all participants and was performed according to the principles outlined in the Declaration of Helsinki. All participants provided written informed consent. Laboratory analysis Venous blood was obtained at each screening round after an overnight fast. EDTA plasma and lithium heparin plasma samples were prepared by centrifugation at 4 °C and stored at -80°C until thawed for testing. EDTA plasma samples from the second screening were sent frozen to LipoScience, (now LabCorp, Morrisville, USA) for testing on the Vantera Clinical Analyzer (Morrisville, North Carolina) and lithium heparin plasma was tested on the Roche Modular system.

Lower plasma magnesium, measured by nuclear magnetic resonance spectroscopy, is associated with increased risk of developing type 2 diabetes mellitus in women: Results from a Dutch prospective cohort study 27 2 Figure 1. Flowchart of the PREVEND study participants included or excluded for the purposes of this study. NMR-based ionized Mg assay As is customary for collecting NMR spectra for the NMR LipoProfile test, EDTA plasma samples were diluted 1:1 with phosphate buffer (pH 7.4) containing 5 mmol/L EDTA. The extra EDTA in the buffer ensured complete chelation of free ionized Mg present in the plasma specimens, as well as any circulating Mg that may not be tightly bound to proteins, citrates or phosphates. Proton NMR spectra were collected on 400 MHz Vantera Clinical Analyzers at 47°C as described previously 16,21. The NMR acquisition time was 48 seconds, with a total sample to sample turnaround time of 90 seconds.

Chapter 2 28 The proton NMR spectra were deconvoluted using proprietary software as follows. The singlet peak emanating from four equivalent protons of the ethylene moiety in the MgEDTA complex (-N-CH2-CH2-N-) appearing at 2.66 ppm in the NMR spectrum was used for quantitation. As the Mg-EDTA NMR signal overlaps with a signal from circulating proteins, the deconvolution method included the protein signal encompassing approximately 50 Hz, and a 16 Hz wide region of the Mg-EDTA peak was integrated. The relation between Mg-EDTA signal area and Mg concentrations were established by standard addition experiments on dialyzed serum, and the conversion factor thus obtained was applied to transform Mg-EDTA signal areas to concentrations expressed in mmol/L. The Mg concentrations were standardized against a 25.0 mM solution of ACS Reagent Grade MgCl2.6H2O (MilliporeSigma, US). Defined amounts of the standard MgCl2 solution were spiked into dialyzed serum devoid of ionized Mg. Accuracy was ascertained through recovery experiments done from 0 to 4.0 mM Mg concentrations. Similar to other NMR assays on Vantera, the commercial assay would also involve running 2 levels of serum controls serving daily check on accuracy and guarding against drift with time. We tested for imprecision in the NMR-measured ionized Mg assay as per CLSI guidelines. Pooled samples with two varying concentrations of Mg (low and high) were tested to determine within-lab (inter-assay) precision. Roche modular total Mg assay and assay comparison The Roche Modular assay is a colorimetric end point assay that measures total Mg in a serum, heparin plasma or urine sample. The method is based on the reaction of Mg with xylidyl blue in an alkaline solution containing ethylene glycol-bis(β-aminoethyl ether)- N,N,N’,N’-tetraacetic acid (EGTA), which has a lower affinity for Mg, in order to mask the calcium in the sample. In the alkaline solution, Mg forms a purple complex with the xylidyl blue diazonium salt and the concentration of Mg is determined photometrically via the decrease in the xylidyl blue absorbance (505/600 nm). In order to understand the differences between the two assays, we compared values from the NMR-based ionized Mg assay with total Mg measured on a Roche Modular Analyzer (Roche Diagnostics, Mannheim, Germany) in 799 samples of appropriate specimen types from the second screening of the PREVEND cohort. The Roche Mg assay has an inter-assay coefficient of variation of 1.3%. Assessment of covariates Body mass index (BMI) was calculated as weight (kg) divided by height squared (m2). Smoking status was defined as self-reported never smoker, former smoker or current

Lower plasma magnesium, measured by nuclear magnetic resonance spectroscopy, is associated with increased risk of developing type 2 diabetes mellitus in women: Results from a Dutch prospective cohort study 29 2 smoker 22. Blood pressure was measured with an automatic Dinamap XL Model 9300 series device (Johnson-Johnson Medical, Tampa, FL, USA). Hypertension was defined as a systolic blood pressure (SBP) >140mmHg or a diastolic blood pressure (DPB) >90mmHg, and/or the use of anti-hypertensive drugs. Information on medication use was combined with information on drug use from the database, containing pharmacy-dispensing data from community pharmacies in the Netherlands 23. Estimated glomerular filtration rate (eGFR) based on serum creatinine and serum cystatin C was calculated from the Chronic Kidney Disease Epidemiology Collaboration equation 24. Urinary albumin, sodium, urea, and creatinine excretion and circulating albumin, sodium, potassium, calcium and creatinine, total cholesterol, high-density lipoprotein cholesterol, triglycerides, high sensitivity C-reactive protein (hsCRP) and glucose were determined as previously described 25–28. Assessment of T2D risk Incident T2D was ascertained if one or more of the following criteria were met: 1) fasting plasma glucose >7.0 mmol/L; 2) random sample plasma glucose >11.1 mmol/L; 3) self-reporting of a physician diagnosis; 4) initiation of glucose-lowering medication use retrieved from a central pharmacy registry 29. Incident T2D was defined as T2D that occurred after the second screening. Statistical analysis Analytic validation data was calculated using Analyze-it (Analyze-it Software, Ltd. Leeds, UK). Passing-Bablok regression analysis was used to test agreement between NMRmeasured Mg in EDTA plasma and heparin plasma Mg measured on the Roche Modular Analyzer. Bland-Altman plots were used to visualize bias. Baseline characteristics are reported in terms of means (SD) when normally distributed or medians (interquartile range) in the case of non-normally distributed data. Categorical data are presented as frequencies (percentages). We prospectively examined the association between chelated Mg and risk of developing T2D using Cox proportional hazards regression models.We used chelated Mg as a continuous variable in these models and additionally, we examined the association in tertiles of chelated Mg. Person-time of follow-up was calculated for each participant from the first visit (baseline) until the last visit, the incidence of T2D, death, or relocation to an unknown destination, whichever came first. Multivariable Cox models were adjusted for age, sex, BMI, smoking (2 categories), alcohol intake (2 categories), triglyceride to high-density lipoprotein cholesterol ratio, hypertensive treatment, parental history of T2D, plasma levels of albumin, potassium

Chapter 2 30 and calcium and urinary albumin excretion, and fasting glucose levels, CRP and eGFR. In addition, we performed sensitivity analyses in which we replaced adjustment for antihypertensive treatment in Cox regression analyses by adjustment for presence of cardiovascular disease and in which we replaced adjustment for eGFR in Cox regression analyses by adjustment for presence of chronic kidney disease. Hazard ratios (HRs) are reported with 95% confidence intervals (CIs). Restricted cubic splines with three knots were performed to show the association between ionized Mg and risk of T2D using Cox regression analyses. We evaluated potential effect modification in the analyses of plasma ionized Mg and risk of T2D by fitting models containing both main effects and their cross-product terms. Interaction terms were considered statistically significant at p<0.10. Missing data (present in 0.0-13.4%) in covariables were handled by multiple imputation 30. Results are reported for imputed data, except for the baseline characteristics and the analytic validation data. We considered a two-sided p value <0.05 as statistically significant. Data were analyzed using SPSS Statistics version 23.0 (SPSS Inc, Chicago, IL). Results Analytical performance of the NMR-measured ionized Mg assay The coefficient of variation for the NMR ionized Mg assay ranged from 4.6-7.1% for within-lab imprecision (Table 1). We compared NMR-measured Mg in EDTA plasma specimens with lithium heparin plasma total Mg measured by a Roche Modular colorimetric assay in 799 samples from the PREVEND study. We found a strong linear relationship between NMR-measured ionized Mg and colorimetrically measured total Mg (r=0.90). Bland-Altman analysis showed a systematic bias of 0.07 mmol/L with chemically measured total Mg concentrations being slightly higher than the ionized Mg quantified by NMR (Figure 2). Passing-Bablok regression analysis revealed an intercept of 0.02 (95% CI: -0.02-0.08) and a slope of 1.08 (95% CI: 1.00-1.13) (Figure 3). 2 pools of EDTA plasma with low and high Mg concentrations were tested twice a day in duplicate for 20 days on one instrument.

Lower plasma magnesium, measured by nuclear magnetic resonance spectroscopy, is associated with increased risk of developing type 2 diabetes mellitus in women: Results from a Dutch prospective cohort study 31 2 Table 1. Within-lab imprecision of ionized Mg measured on the Vantera Clinical Analyzer. NMR-measured Mg (mmol/L) Within-lab Low High Mean 0.489 0.892 SD 0.035 0.041 CV 7.1% 4.6% Figure 2. Bias plots for total Mg from the Roche Modular and NMR-measured ionized Mg. Figure 3. Passing and Bablok regression analysis.

Chapter 2 32 Association of ionized Mg with risk of developing T2D The baseline characteristics of the 5747 participants are shown in Table 2. Mean age was 53.0 ± 11.9 years, mean NMR-measured ionized Mg was 0.75 ± 0.05 mmol/L and 50.4% of the participants were female. Higher ionized Mg was associated with a slightly lower BMI, lower fasting glucose levels and higher HDL-cholesterol levels. In addition, subjects in the highest tertile of ionized Mg were more likely to be non-smokers. During a median follow-up period of 11.2 (IQR: 7.7-12.0) years, 289 (5.0%) participants developed T2D. We found an association between the levels of NMR-measured ionized Mg and the risk of developing T2D in the total population (HR: 1.50; 95% CI: 1.191.89). However, after multivariable adjustment, the association lost significance (HR: 1.16; 95% CI: 0.91-1.47). The association of NMR-measured Mg with risk of T2D was modified by sex (Pinteraction=0.007). In men, the association between NMR-measured Mg and risk of developing T2D was non-significant in the crude model (HR: 1.25; 95% CI: 0.94-1.67) and in a fully adjusted multivariable model (HR: 0.90; 95%: 0.67-1.21) (Table 3). In women, on the other hand, we found an association between NMR-measured Mg and the risk of developing T2D in the crude model (HR: 2.02, 95% CI: 1.37-2.99). After adjustment for lifestyle factors, including BMI, alcohol consumption, smoking status, triglyceride to HDL cholesterol ratio, use of antihypertensive drugs, and parental history of T2D (Model 1), the hazard ratio was slightly attenuated, but not substantially different (HR: 1.66; 95% CI: 1.11-2.47) (Table 3). When we further adjusted for the variables in Model 3, including fasting glucose, CRP and eGFR, the association remained similar (HR: 1.80; 95% CI: 1.20-2.70). Furthermore, the HR for women in the lowest tertile of NMR-measured Mg was 1.65 (95% CI: 1.02-2.66) in the crude analysis and 1.72 (95% CI: 1.03-2.86) after multivariable adjustment. A restricted spline curve confirmed the loglinear inverse association of NMR-measured Mg and risk of developing T2D in women (Figure 4).