European Journal of Clinical Nutrition
SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Contact Springer Nature
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
May 2002, Volume 56, Number 5, Pages 442-447
Table of contents    Previous  Article  Next   [PDF]
Original Communication
Magnesium bioavailability from mineral water. A study in adult men
M Verhas1, V de La Guéronnière2, J-M Grognet3, J Paternot1, A Hermanne4, P Van den Winkel4, R Gheldof1, P Martin1, M Fantino5 and Y Rayssiguier6

1CHU Brugmann-Free University of Brussels, Brussels, Belgium

2SEV, 92340 Bourg la Reine, France

3Département des Sciences du Vivant, CEA, Gif sur Yvette, France

4Cyclotron Eenheid-Vrije Universiteit Brussels, Brussels, Belgium

5Laboratoire de Physiologie CHU Dijon, Dijon, France

6Laboratoire de Physiologie, INRA, Clermont-Ferrand-Theix, France

Correspondence to: M Verhas, Department of Nuclear Medicine, Brugmann Hospital, Free University of Brussels, 4 Place Van Gehuchten, 1020, Brussels, Belgium. E-mail:

Guarantor:M Verhas.

Contributors: MV was responsible for conception and realisation of the protocol and redaction. VLG was responsible for initiation of the study, discussion and redaction. JMG was responsible for conception of protocol. JP, RG and PM were responsible for protocol realisation. AH and PVW were responsible for isotope production. MF was responsible for dietary consumption collection and analysis. YR was responsible for results analysis and discussion.


Objective: To assess magnesium enteral absorption from a magnesium-rich mineral water.

Design: Experimental study.

Setting: Department of Nuclear Medicine, Brugmann Hospital, Brussels, Belgium.

Subjects: Ten healthy male volunteers in the age range 25-42 y.

Intervention: Each subject completed two sessions in a random order. At one session, they received an oral load of 300 ml of water (containing 1.2 mmol Mg), traced with 28Mg, and at the other session they received an intravenous injection of 28Mg, in order to take into account the metabolism of endogenous magnesium. The dietary consumption was further noted on a weekly diary.

Results: The mean bioavailability was 59.1% (s.d.±13.6). Magnesium absorption and age were significantly inversely correlated (r=-0.68, P=0.035).

Conclusion: Magnesium-rich mineral water is a reliable source of magnesium. Our observation of decreased magnesium absorption with age deserves further investigations.

Sponsorship: The study was sponsored by SEV, Bourg la Reine, France.

European Journal of Clinical Nutrition (2002) 56, 442-447. DOI:10.1038/sj/ejcn/1601333


magnesium; bioavailability; mineral waters; whole-body counting


Magnesium (Mg) plays an essential role in a wide variety of fundamental cellular reactions (Shils, 1988). Neuromuscular hyperexcitability is the initial problem cited in individuals who have or are developing magnesium deficiency (Durlach et al, 1997). Many recent findings have augmented the significance of magnesium in human health and disease. There is mounting evidence that alterations in magnesium metabolism have a negative impact on cardiovascular biology (Shils, 1988). The recommended dietary intake of magnesium for the French population has been recently set at 6 mg/kg/body weight/day for both men and women. The American recommendation for Mg intake has recently been augmented and it is now close to the French recommendation (Rayssiguier et al, 2000). In industrialised countries reduction in energy intake and increase in consumption of foods containing only energy without minerals have contributed to decreasing Mg intake. In France in 1997, more than 70% of subjects had dietary Mg intake lower than the RDA and 20% consumed less than two-thirds of the RDA (Galan et al, 1997). The Mg intake is lower in the USA (Cleveland et al, 1996). From a public health viewpoint it is important not only to identify the possible sources of Mg, but also to assess their respective bioavailabilities. Mineral water may represent a substantial source of dietary magnesium, with levels of up to 4 mmol/l in some mineral waters. It could thus provide one-quarter to one-third of the magnesium RDA per litre intake. Epidemiological studies underlie the importance of mineral water as a source of dietary Mg and suggest that Mg in drinking water is associated with lower mortality from cardiovascular diseases (Rubenowitz et al, 1998).

In the present study, we measured the magnesium absorption rate in humans of a single product, magnesium-rich water (3.9 mmol/l). Magnesium has two stable isotopes, 25Mg and 26Mg. Because their high natural abundance (10.00 and 11.01%, respectively) reduces the sensitivity with which they can be measured in body fluids (Schwartz et al, 1984; Schuette et al, 1990; Benech & Grognet, 1995; Schwartz, 1982), we used the radioactive isotope 28Mg instead, which is more suited to assessing the bioavailability of a 1.2 mmol Mg load in a 300 ml sample of water, as already described (Danielson et al, 1979).



Ten male volunteers (mean age 29.4 y; range 25-42 y), deemed healthy on the basis of a complete medical history and examination, normal blood count, kidney function, electrolytes and plasma magnesium (0.65-1 mmol/l), without any treatment or particular diet, were recruited after they had given informed consent to the protocol.

Study design

The study of 28Mg bioavailability allows us to measure digestive absorption of Mg from the liquid (or any nutrient) in which this isotope has been mixed because the isotope acts as a tracer. It has been demonstrated that solubility of magnesium chloride in water is high (solubility product 55 g MgCl2 anhydrous/100 g water at 20°C) and thus one can assume that 28Mg in the form of MgCl2 is a reliable tracer. In two different sessions (minimum washout period of 1 week), each volunteer received respectively 28Mg orally in 300 ml water and 28Mg intravenously. The counting rates of both these procedures were compared, their ratio giving the magnesium digestive absorption (Van den Berg et al, 1995). The counting rates measured on the forearm were expressed as the percentage of the orally or intravenously administered dose. The actual size of the administered dose parameter in counts per minute was measured by the counting rate of the whole body volunteer 1 h after isotope administration and before any urinary or faecal excretion. All counting rates were corrected for radioactive decay.


On their first session day, volunteers came to the laboratory at 8:00 after a 12 h fast and were allocated at random to an intravenous or an oral administration of 28Mg. On their second session day, the treatments were reversed. In the meantime, subjects were instructed not to modify their dietary habits. Before tracer administration, background activity was determined by whole body counting and forearm measurements. For the intravenous administration, a 20-gauge catheter (Critikon 4220) with stylet was inserted in the antecubital vein of the right forearm, the left forearm being used for the measurement of radioactivity.

After recording background activity, one volunteer received an oral load of approximately 185 kBq (5 µCi) of 28Mg in 300 ml of mineral water, and the other volunteer was injected i.v. with 74 kBq (2 µCi) 28Mg in NaCl 9permil. The radionuclide 28Mg (T1/2=21.2 h, 400 keV) was obtained by irradiation of 27Al (Probst et al, 1976) and from the reaction 27Al (a.3p) 28Mg. The irradiation was carried out at the Cyclotron Unit of the Vrije Universiteit Brussel (VUB). Approximately 370 kBq (10 µCi) of 28Mg were produced in each run and were devoid of contaminants after chemical purification.

One hour after administration of 28Mg and in the absence of faecal and urinary loss, a whole body count was taken as 100% of the administered dose for expression of whole body retention. Forearm counts (five measurements of 1 min) with identical geometry were carried out every hour for 6 h and 8 and 24 h after administration of 28Mg. A decay correction was applied after periodic standard measurements of an aliquot containing 37 kBq in a volume of 12 l.

Figure 1 represents the schedule for a typical test day. The mineral water was provided by SEV (France) and contained: Mg, 3.9 mmol/l; Ca, 4.6 mmol/l; Na, 26.8 mmol/l; Cl, 8.7 mmol/l; bicarbonates, 37.8 mmol/l.

Daily dietary intake was assessed in the month preceding the first test day from quantitative 1 week food records. The subjects were instructed by a dietician to note in detail, over 7 consecutive days, their daily food and drink intake and the portion size, either weighed or expressed in usual household measures. The record was then reviewed and validated by the dietician who translated the quantities expressed as household measures into grams using a computerised dietary system. The software converted the collected data into daily energy and nutrient intakes using a food composition database containing the energy (kJ) and 41 food constituents of more than 1500 food items (compilation of SOUCI, Southgate and CIQUAL databases). Calcium and magnesium in mineral water were determined by atomic absorption spectrophotometry. Serum magnesium was determined using the calamagite method (Wong, 1975).

Statistical analysis

The significance of the relationship between magnesium absorption rate and nutrients intake was evaluated using the non-parametric Wilcoxon rank matched-pairs signed-ranks test (Siegel, 1956). Values are means and standard deviation (s.d.).

Body counting

Radioactivity was counted using a whole body counter (Nuclear Enterprise Ltd, Edinburgh, UK) with four NaI (T1) detectors (diameter 15 cm, height 10 cm) equipped with cylindrical collimators. Their vertical separation and orientation were adjustable, allowing measurement of either the forearm or the whole body.

For the whole body measurements, the subject was supine in a standardised manner. The measurements were made at a scan rate of 40 cm/min over a total length of 220 cm. The four detectors were placed in opposite pairs (D1, D2 up and D3, D4 down).

For the forearm activity measurements, the subject was positioned outside the whole body counter chamber in a standardised manner; the left arm was stretched out through a window into the wall of the chamber with the forearm lying on a splint inside the chamber. Two detectors (D1 up, D3 down) were used in opposition. Forearm background activity was recorded in the same way. To obtain an identical counting geometry for the measurement of the administered dose and the activity of the forearm, the administered dose was measured with the disposable syringe lying on the same forearm splint inside the whole body chamber.

All measurements (geometric mean of detectors in opposition divided by the scanning time) were corrected for background and decay. The results were expressed as a percentage of administered dose of 28Mg (whole body count 1 h after administration). The ratio between oral and i.v. forearm retention at 24 h gave the bioavailability of magnesium:



D1=detector 1 counting rate (counts/min or cpm) corrected for background and decay; D3=detector 2 counting rate (counts/min or cpm) corrected for background and decay.

Ethical considerations

The study was approved by the Human Research Ethical Board of our Institution, and was in agreement with the Helsinki statement on human rights (1975, reviewed in 1983).


All 10 subjects completed both sessions. Periodic standard measurements showed no electronic shift in the whole body chamber throughout the study. Because of the long time needed for the 28Mg production (10 h), failures of cyclotron productions were numerous, explaining the average mean time between two sessions (oral and intravenous) of 54 days, ranging from 35 to 84 days. Anthropometric data of the subjects and mean daily nutrient intakes are detailed in Table 1. Mean bioavailability of Mg at 24 h with the forearm method (Table 2) was 59% (s.d. 13.6). The plot of mean absorption curve as a function of time (until 24 h) is shown in Figure 2. Age was inversely related (r=-0.68; P=0.035) to Mg absorption rate (Figure 3).


The objective of this study was to assess the bioavailability of a magnesium-rich carbonated water in humans. Magnesium net absorption is the result of enteral absorption and excretion. Endogenous faecal loss represents 20-50 mg/day. Absorption occurs all along the digestive tract, but mainly in the distal duodenum and the ileon. Magnesium absorption takes place by two mechanisms: a main non-saturable passive mechanism, and a regulable saturable active mechanism involved in the case of low magnesium intake (Rayssiguier et al, 2000). Because of technical problems, the interval between two sessions reached 84 days in two cases; in the meantime, subjects declared that they had not modified their usual diet. It is noticeable that in their study, producing 28Mg once every 4 weeks, Schwartz et al (1984) also had variable time lapse of i.v. to oral (30-90 days).

In humans, a wide range (10-75%) of magnesium absorption rates has been reported (Danielson et al, 1979; Schwartz et al, 1984; Schuette et al, 1990; Fine et al, 1991; Brink & Beynen 1992; Lönnerdal et al, 1993; Lönnerdal, 1995; Benech & Grognet, 1995; Serfaty-Lacrosniere et al, 1995; Knudsen et al, 1996; Fairweather-Tait & Hurrell, 1996; Andon et al, 1996; Abrams et al, 1997; Sojka et al, 1997) and have been recently reviewed (Ekmekcioglu, 2000); higher values have been observed in animal studies (Lönnerdal et al, 1993; Lazichi Lakshmanan et al, 1984; Kikunega et al, 1995). In these previous studies, various protocols have been applied, including balance studies (Lönnerdal et al, 1993; Lönnerdal, 1995; Serfaty-Lacrosniere et al, 1995; Andon et al, 1996), true bioavailability studies with stable (Schwartz et al, 1984; Schuette et al, 1990; Knudsen et al, 1996; Sojka et al, 1997) and have been recently reviewed (Ekmekcioglu, 2000); higher values have been observed in animal studies (Lönnerdal et al, 1993; Lazichi Lakshmana et al, 1984; Kikynega et al, 1995). In these previous studies, various protocols have been applied, including balance studies (Lönnerdal et al, 1993; Lönnerdal, 1995; Sarfaty-Lacrosniere et al, 1995; Andon et al, 1996), true bioavailability studies with stable (Schwartz et al, 1984; Schuette et al, 1990; Knudsen et al, 1996; Sojka et al, 1997) and/or radio-isotopic techniques (Danielson et al, 1979; Mountokalakis et al, 1980; Schwartz et al, 1984), and gastrointestinal washing (Fine et al, 1991); these may account, at least in part, for the wide range of observed results. Other elements such as the magnesium formulation may also be of concern as pharmaceutical (Schuette et al, 1990; Fine et al, 1991) or nutritional origin of preparations that have been administered (Schwartz et al, 1984; Abrams et al, 1997). Furthermore, the magnesium load administered tested varies widely among studies (from 35 to 960 mg), notwithstanding the age of subjects (Schwartz et al, 1984; Schuette et al, 1990; Andon et al, 1996; Abrams et al, 1997; Sojka et al, 1997), their physical condition and the proximity to meals for administration (Andon et al, 1996). The 59% bioavailability rate observed in the present study lies in the upper range reported in the literature for adolescents and adults. Considering the low level of Mg administered (1.2 mmol), our observations are consistent with the 65% bioavailability of 1.5 mmol Mg reported by Fine et al (1991). This latter study underlines that fractional absorption falls progressively with each increment of intake. Ekmekcioglu (2000) has plotted fractional absorption rates of Mg from liquids, from various studies. He has shown that the fractional absorption rate is high at low Mg load and decreases exponentially with increasing carrier amounts. Several authors have outlined that higher bioavailability is observed when a given amount of Mg is distributed over a day rather than being consumed in a single bolus (Schuette et al, 1990; Fine et al, 1991; Lönnerdal, 1995); consequently a regular water intake distributed throughout the day would be expected to lead to a higher absorption.

In line with recent results (Abrams et al, 1997), we did not find any relationship between the daily Mg intake assessed with the 7 day questionnaire and the absorption rate of the Mg load tested.

An age-related decline in the capacity of the intestine to absorb dietary Mg has been suggested but is not well documented (Mountokalakis et al, 1980; Durlach et al, 1993). Although our subjects had only a two-decade age range, we also made this observation in our study. Nevertheless, the age-dependence Mg bioavailability found in our study is a preliminary result due to the restricted sample size.

In conclusion, our objective was to investigate water Mg bioavailability from a carbonated water (containing 3.9 mmol/l Mg), and not to compare it with another source. The mean rate found, 59% (s.d. 13.6), lies in the upper reported range for solid foods. This could be favoured by Mg solubility. Further questions should be explored in the future: what would be the effects of a long term consumption of such water on Mg bioavailability? The main anions accompanying Mg in natural waters are bicarbonate and sulphate. Knowing their respective alkalinizing or acidifying properties, could these anions influence Mg metabolism and excretion? The observation of decreased Mg absorption with age also deserves further investigation.


We are grateful to Drs D Willems and P Bergmann for the clinical biochemistry and to C Steenwinckel for analysing the diets of the subjects.


Abrams SA, Grusak MA, Stuff J, O'Brien KO. (1997). Calcium and magnesium balance in 9-14-y-old children. Am. J. Clin. Nutr., 66: 1172-1177. MEDLINE

Andon MB, Illich JZ, Tzagournis MA, Matkovic V. (1996). Magnesium balance in adolescent females consuming a low- or high-calcium diet. Am. J. Clin. Nutr., 63: 950-953. MEDLINE

Benech H, Grognet JM. (1995). Recent data on the evaluation of magnesium bioavailability in humans. Magnesium Res., 8: 277-284.

Brink EJ, Beynen AC. (1992). Nutrition and magnesium absorption: a review. Prog. Food Nutr. Sci., 16: 125-162. MEDLINE

Cleveland LE, Goldman JD, Borrud LG. (1996). Data Tables: Results from USDA's 1994 Continuing Survey of Food Intakes by Individuals and 1994 Diet and Health Knowledge Survey. Beltsville, MD: Agricultural Research Service, US Department of Agriculture.

Danielson BG, Johansson G, Jung B, Ljunghall S, Lundqvist H, Malmborg P. (1979). Gastrointestinal magnesium absorption, kinetic studies with 28Mg and a simple method for determination of fractional absorption. Mineral Electrolyte Metab., 2: 116-123.

Durlach J, Durlach V, Bac P, Rayssiguier Y, Bara M, Guiet-Bara A. (1993). Magnesium and ageing. II. Clinical data: aetiological mechanism and pathophysiological consequences of magnesium deficit in the elderly. Magnesium Res., 6: 379-394.

Durlach J, Bac P, Durlach V, Bara M, Guiet-Bara A. (1997). Neurotic, neuromuscular and autonomic nervous form of magnesium imbalance. Magnesium Res., 10: 169-195.

Ekmekcioglu C. (2000). Intestinal bioavailability of minerals and trace elements from milk and beverages in humans. Nahnung, 44: 390, 397.

Fairweather-Tait S, Hurrell RF. (1996). Bioavailability of minerals and trace elements. Nutr. Res. Rev., 9: 295-324.

Fine KD, Santa A, na C, Porter JL, Fordtran JS. (1991). Intestinal absorption of Mg from food and supplements. J. Clin. Investi., 88: 396-402.

Galan P, Preziosi P, Durlach V, Valeix P, Ribos L, Bouzid D, Favier A, Hercberg S. (1997). Dietary magnesium intake in a French adult population. Magnesium Res., 10: 321-328.

Kikunega S, Ishii H, Takahashi M. (1995). The bioavailability of spinach and the effect of oxalic acid on magnesium utilization examined in diets of magnesium-deficients rats. J. Nutr. Sci. Vitaminol., 41: 671-685. MEDLINE

Knudsen E, Sandstrom B, Solgaard P. (1996). Zinc, copper an magnesium absorption from a rich-fiber diet. J. Trace Elements Med. Biol., 10: 68-76.

Lazichi L, akshmanan F, Rao RB, Kim WW, Kelsay JL. (1984). Magnesium intakes, balances and blood levels of adults consuming self-selected diets. Am. J. Clin. Nutr., 40: 1380-1389. MEDLINE

Lönnerdal B. (1995). Magnesium nutrition in infants. Magnesium Res., 8: 99-105.

Lönnerdal B, Yuen M, Glazier C, Litov RE. (1993). Magnesium bioavailability from human milk, cow milk and infant formula in suckling rat pups. Am. J. Clin. Nutr., 58: 392-397. MEDLINE

Mountokalakis T, Singhellakis P, Aleviaki C, Virvidakis K, Ikkos D. (1980). Relationship between degree of renal failure and impairment of intestinal magnesium absorption, In Magnesium in Health and Disease, eds. Cantin & Seelig,. 453-458, New York: SP Medicine/Science Books.

Probst HJ, Qaim SM, Weinreich R. (1976). Excitation functions of high-energy particle induced nuclear reaction on aluminium and magnesium: production of 28Mg. Int. J. Appl. Radiat. Isoto., 27: 431-441.

Rayssiguier Y, Durlach J, Boirie Y. (2000). Apports Nutritionnels Conseillés pour la population française. A Martin coordonnateur. Paris: TEC DOC Lavoisier.

Rubenowitz E, Axelsson G, Rylander R. (1998). Magnesium in drinking water and body magnesium status measured using an oral loading test. Scan. J. Clin. Lab. Invest., 58: 423-428.

Schuette SA, Ziegler EE, Nelson SE, Janghorbani M. (1990). Feasibility of using the stable isotope 25Mg to study Mg metabolism in infants. Pediatr. Res., 27: 36-40. MEDLINE

Schwartz R. (1982). 26Mg as a probe in research on the role of magnesium in nutrition and metabolism. Fed. Proc., 41: 2709-2713. MEDLINE

Schwartz R, Spencer H, Welsh JEJ. (1984). Magnesium absorption in human subjects from leafy vegetables intrinsically labeled with stable 26Mg. Am. J. Clin. Nutr., 39: 571-576. MEDLINE

Serfaty-Lacrosniere C, Wood RJ, Voytko D. (1995). Hypochlorhydria from short-term Omeprazole does not inhibit intestinal absorption of calcium, phosphorus, magnesium or zinc from food in humans. J. Am. College Nutr., 14: 364-368.

Shils ME. (1988)). Magnesium in health and disease. A. Rev. Nutr., 8: 429-460.

Siegel S. (1956). Non Parametric Statistics for the Behavioral Sciences. Tokyo: McGraw-Hill Kogakusha.

Sojka J, Wastney M, Abrams S, Lewis SF, Martin B, Weaver C, Peacock M. (1997). Magnesium kinetics in adolescent girls determined using stable isotopes: effects of high and low calcium intake. Am. J. Physiol., 273: (Regulatory Integrative Comp. Physiol 42), R710-R715. MEDLINE

Van den Berg GJ, Wolterbeek HTH, De Goeij JJM, Beynen AC. (1995). Absorption and retention studies of trace elements and minerals in rats using radio-tracers and whole body counting. Lab. Anim., 29: 66-77. MEDLINE

Wong HK. (1975). Fully automated procedure for serum magnesium. Clin. Chem., 21: 169-. MEDLINE


Figure 1 Schedule of a 24 h day-test. Hour is indicated as the day time. Eight counting rates were made on the forearm of each volunteer (n=10) and were expressed as the percentage of the orally or intravenously administered dose. Electronic shift of the whole body chamber was checked and decay corrected by periodic standard measurements under identical conditions. The first whole body count (WBC) of 28Mg gave the administered dose for the whole body 28Mg retention measurement at 24 h expressed as percentage of administered dose.

Figure 2 Mean bioavailability of magnesium against time studied by the forearm counting method on 10 subjects. Results in percentage of administered dose (%AD) after intravenous (IV), oral administration (OR) and the resulting absorption (ABS).

Figure 3 Correlation of 28Mg bioavailability and age (n=10 adult subjects); r=-0.68 (P=0.035).


Table 1 Anthropometric and daily dietary intake. The daily magnesium, calcium, protein, lipid, carbohydrates and water intake were calculated after a computerised analysis taking into account the daily food and drink intake and the portion size

Table 2 Individual plasma magnesium and bioavailability. Magnesium bioavailability is obtained by the ratio of forearm activity after oral and i.v. administration of 28Mg. Data collected are expressed as percentage of administered dose (%AD) for 10 volunteers

Received 30 March 2001; revised 10 August 2001; accepted 6 September 2001
May 2002, Volume 56, Number 5, Pages 442-447
Table of contents    Previous  Article  Next    [PDF]