Bone 44 (2009) 120124

Contents lists available at ScienceDirect

Bone
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b o n e

Alkaline mineral water lowers bone resorption even in calcium sufciency: Alkaline mineral water and bone metabolism
Emma Wynn a,, Marc-Antoine Krieg a, Jean-Marc Aeschlimann b, Peter Burckhardt c
a b c

Centre of Bone Diseases, Lausanne University Hospital, Switzerland Nestl Research Centre, Switzerland Clinic Bois-Cerf, Osteoporosis Consultation, Lausanne, Switzerland

a r t i c l e

i n f o

a b s t r a c t
Background: Dietary acid charge enhances bone loss. Bicarbonate or alkali diet decreases bone resorption in humans. We compared the effect of an alkaline mineral water, rich in bicarbonate, with that of an acid one, rich in calcium only, on bone markers, in young women with a normal calcium intake. Methods: This study compared water A (per litre: 520 mg Ca, 291 mg HCO 3, 1160 mg SO4, Potential Renal Acid load (PRAL) +9.2 mEq) with water B (per litre: 547 mg Ca, 2172 mg HCO 3, 9 mg SO4, PRAL 11.2 mEq). 30 female dieticians aged 26.3 yrs (SD 7.3) were randomized into two groups, followed an identical weighed, balanced diet (965 mg Ca) and drank 1.5 l/d of the assigned water. Changes in blood and urine electrolytes, Ctelopeptides (CTX), urinary pH and bicarbonate, and serum PTH were measured after 2 and 4 weeks. Results: The two groups were not different at baseline, and showed a similar increase in urinary calcium excretion. Urinary pH and bicarbonate excretion increased with water B, but not with water A. PTH (p = 0.022) and S-CTX (p = 0.023) decreased with water B but not with water A. Conclusion: In calcium sufciency, the acid calcium-rich water had no effect on bone resorption, while the alkaline water rich in bicarbonate led to a signicant decrease of PTH and of S-CTX. 2008 Elsevier Inc. All rights reserved.

Article history: Received 14 January 2008 Revised 22 August 2008 Accepted 10 September 2008 Available online 26 September 2008 Edited by: R. Baron Keywords: Bicarbonate Mineral water Nutrition PRAL PTH Resorption markers

Introduction The benecial effects of numerous mineral waters on bone metabolism are mainly attributed to their calcium content [16]. Calcium-rich mineral waters are an alternative to dairy products as their bioavailability is similar or even better [7]. They have decreased bone resorption, but only in short-term trials or in calcium and oestrogen deciency [3]. Minerals other than calcium, especially bicarbonate, also have a positive effect on bone metabolism [8,9]. Nutritional acid load contributes to the age related bone loss [10]. To keep pH constant, the organism uses buffer systems such as bone, expiration of CO2 and renal excretion of acids [11]. The endogenous production of acid, consequent to a normal Western diet, is about 1 mEq/kg body weight, per day [12]. To buffer this, 2 mEq of calcium is necessary [12]. The more acid precursors a diet contains, the greater the need for buffering [13] and the enhancement of bone resorption. In addition, with advanced age, the ability to excrete acids by the kidney declines [14].

In vitro, an efux of calcium from bone tissue is already noted after 3 h incubation in an acidic medium [15]. Indeed, bone has been shown to contribute to the maintenance of the extra-cellular uid pH [16]. A low acid diet could contribute to the maintenance of bone mineral density [17]. In healthy subjects, supplements of potassium bicarbonate, potassium citrate, and even mineral water rich in bicarbonates decreased the calciuria and bone resorption markers [1821]. We investigated whether a high bicarbonate and calcium intake from an alkaline mineral water decreases bone resorption more than a high calcium intake from an acid water, in healthy subjects on a balanced diet adapted to their energetic needs. Subjects and methods For this open, randomized, controlled study, thirty female dieticians or student dieticians were recruited. The inclusion criteria were good health, 1845 years and BMI 18.525 kg/m2. The exclusion criteria were medication interfering with calcium metabolism (diuretics, corticoids), pregnancy, and inclusion in other protocols. They were declared healthy after completion of a medical questionnaire. Out of the 30 participants, 23 were on the contraceptive pill. The subjects were randomized into 2 groups using the method of randomly permuted blocks and assigned to mineral water A or B. 4 of the 7 subjects not taking a contraceptive pill were randomly attributed to Water A and 3 to Water B.

This study was funded by the Foundation for Research on Osteoporosis and Bone Diseases, Lausanne, Switzerland. The manufacturers generously provided the mineral water. * Corresponding author. DAL-Hpital orthopdique, Centre des Maladies Osseuses, Avenue Pierre-Decker 4, 1011 Lausanne, Switzerland. Fax: +41 021 314 47 37. E-mail address: [email protected] (E. Wynn). 8756-3282/$ see front matter 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2008.09.007

E. Wynn et al. / Bone 44 (2009) 120124

121

Each subject had to drink 1.5 l a day of one of the two mineral waters during 28 days. The composition of the waters is shown in Table 1. One is very rich in calcium (Water A, AdelbodnerR, Switzerland) and the other is equally rich in calcium but also rich in bicarbonate (Water B, KryniczankaR, Poland). In addition, water A was also rich in sulphate which increases the acid load. For the assessment of the acid or alkaline load of food, a calculation model was developed by Remer and Manz [22], the potential renal acid load (PRAL). PRAL can be accurately calculated when the nutrient data for protein, P, Cl, K, Mg, Ca and Na is known. Phosphate content of mineral water is negligible and for this reason not indicated. On request, both manufacturers conrmed that the phosphate content of their waters is non signicant. Therefore, phosphate was not included in the PRAL algorithm. PRAL takes into account the average intestinal absorption rate of the respective components. Its algorithm is (0.00049 protein[mg]) + (0.027 Cl [mg]) + (0.037 P [mg]) (0.021 K [mg]) (0.026 Mg [mg]) (0.0413 Na [mg]) (0.013 Ca [mg]) [22]. This algorithm takes into account methionine and cysteine as nutritional sources of SO4 coming from protein. However in mineral waters, SO4 is in solution. This justies a correction of the formula by using the molecular weight of SO4 instead of the two amino acids which is 96, and an absorption rate of 70k, resulting in a conversion factor of 0.0146 instead of 0.00049 (Remer T, personal communication, 2007). Therefore, the PRAL is +9.2 mEq/l for water A (acid) and 11.2 mEq/l for water B (alkaline). The subjects were informed of the aims of the study and gave written consent. Our protocol was accepted by the Ethic Committee of the Faculty of Medicine of the University of Lausanne, Switzerland. The study took place at the University Hospital. Prior to the rst of the four visits at the hospital, the subjects were given full explanations by phone about the study and were questioned on their health status to identify inclusion and exclusion criteria. 35 volunteers applied and 30 could be included. Five were not included as they lived too far away to be on time for the early morning appointments. At the rst visit the subjects got the bottles of mineral water and they met a qualied dietician (EW) who explained the imperatives of the study, i.e. mainly the dietary recommendations and delivered the same weighed and balanced diet plan for 28 days to all subjects. The diet plan followed the recommendations of the daily consumption frequencies of 5 portions of fruit and vegetables, 3 of starch, 3 of dairy products and 1 of meat, sh or egg. The diet plan included 1875 kcal, 237 g of carbohydrate, 67 g of lipids, 75 g of protein and 965 mg of calcium. Subjects were allowed to eat more or less starch depending on their appetite and physical activity. The other food groups had to be consumed in the established quantities. All participants had to drink 1.5 l a day of the assigned mineral water. They were advised not to drink more than one cup of coffee or tea a

day, not to drink alcohol and to drink a little tap water if they were really too thirsty. A glass of fruit juice had to be considered as a portion of fruit, and a glass of milk as a portion of dairy product. Subjects were asked to avoid salty processed food and to moderately salt their food. The three following visits occurred early morning of the rst day, at 2 weeks and at 4 weeks of the study. They enabled reception of the 24 h urine, collection of the fasting morning urine and blood sampling. Blood tests The following analyses were undertaken at baseline and at 2 and 4 weeks: sodium, potassium, ionized and total calcium, phosphorus, alkaline phosphatase, albumin, creatinine, 25-OH vitamin D (DiaSorin, Stillwater, Minnesota, 55082-0285 USA, (RIA+ extraction), intra- and interassay coefcients 8.612.5k and 6.113.2k, detection limit 2 g/l), blood and urinary C-telopeptides (CTX), PTH (Intact PTH, Nichols Institute, San Clemente, CA 92673, USA, (CIA), intra- and interassay coefcients 5.76.2k at and 7.39.7k, detection limit 2 ng/l), glucose (baseline only) and blood cell count (baseline only). The electrolytes, albumin and alkaline phosphatase were analysed at the routine Laboratory of the hospital using standard methods. CTX (C-terminal fragment of the type I collagen) were assayed with the Beta-Cross Laps Elecsys, Roche Diagnostics GmbH, D-68293 Mannheim, (ECLIA), intraand interassay coefcients 1.64.7k and 3.94.3k, detection limit 0.010 g/l) at the Endocrinology Laboratory of the hospital. All the samples were hermetically closed at reception and transported to the lab in less than 15 min. The measurement of ionised calcium was introduced during the trial. For this reason it was performed only in 5 subjects on water A and 7 subjects on water B. A multiple specic electrode automate was used (ABL 800, Radiometer, Copenhagen, RSCH). Urinary tests Calcium, sodium, potassium, creatinine, bicarbonate and pH were measured in the 24 h urine and in the 2 h fasting morning urine (second miction). In addition, CTX were measured (Nordic Bioscience Diagnostics A/S, Herlev-Hovedgade 207, 2730 Herlev, Denmark, (ELISA), intra- and interassay coefcients 4.79.4k and 6.616.3k, detection limit 0.050 mg/l). Urine gas analysis was done with a gazometer (Bayer Diagnostics 248) at the Paediatric Research Laboratory. Statistical analysis The results were analysed using ExcelR functions, StatistiX 8 (Analytical Software, Tallahassee, FL, USA) and SPSS version 14.0 (SPSS Inc., Chicago, IL, USA). Differences were considered signicant at p b 0.05. Descriptive statistics were undertaken for all anthropometric parameters and analytical variables. For each variable, groups' homogeneity was checked at baseline by Student t-tests. Groups were compared using paired changes from baseline by classical t-tests (Mean and SD), robust ones (Median and robust SD), as well as by Wilcoxon rank-sum tests. Results from these 3 tests being comparable, we only present the classical t-tests (Mean and SD). Results The two groups were comparable in age (group A: 26.3 yrs 7.6; group B: 26.3 yrs 7.3) and BMI (group A: 20.6 1.8 kg/m2; group B: 21.3 2.3 kg/m2). The blood formula and the glycaemia were all in the reference limits. There were no signicant differences among the two groups in any of the laboratory values, as well as in age, height, weight and BMI. Being dieticians, the participants showed excellent compliance and perfect understanding of the diet plan. All 30 volunteers completed the study.

Table 1 Composition of the mineral waters Water A mg/l Calcium Magnesium Sodium Potassium Chloride Sulphate PRALa Fluoride Bicarbonate 520 35 5.5 1.55 5.6 1160 0.2 291 mmol/l 13 1.4 0.2 0.04 0.2 12.1 0.011 4.8 mEq/l 26 2.9 0.2 0.04 0.2 24.2 +9.2 0.011 4.8 Water B mg/l 548 65 67 6.5 10.6 9 0.1 2172 mmol/l 13.7 2.7 2.91 0.2 0.3 0.09 0.005 35.6 mEq/l 27.4 5.4 2.91 0.2 0.3 0.18 11.2 0.005 35.6

Manufacturers always indicate the composition of mineral waters in mg/l. Results of this trial are indicated in mmol/l for comparison with the literature. PRAL in mEq/l allows to assess the acid or the alkaline load. a PRAL according to Remer and Manz ([22], Remer T, personal communication, 2007).

122 Table 2 Biochemical results (mean SD) Units Serum PTH CTX Vitamin D Phosphatase alkaline Phosphorus Water A Water B Water A Water B Water A Water B Water A Water B Water A Water B ng/l g/l g/l U/l

E. Wynn et al. / Bone 44 (2009) 120124

Baseline value (n = 15) 28.6 (11.6) 33.3 (11) 0.380 (0.115) 0.479 (0.261) 23.1 (6.3) 25.5 (10) 53.0 (14.2) 49.1 (11.8) 1.26 (0.16) 1.28 (0.18)

2 weeks (n = 15) 27.0 (9.4) 26.1 (10) 0.385 (0.137) 0.413 (0.198) 22.4 (7.7) 24.5 (10.3) 53.8 (16.8) 49.2 (11.2) 1.28 (0.16) 1.31 (0.2)

4 weeks (n = 15) 29.1 27.7 0.389 0.401 23.6 25.8 51.3 46.7 1.27 1.29 (15.9) (8.1) (0.110) (0.234) (7.3) (9.6) (13.9) (11.8) (0.12) (0.14)

mmol/l

24 h urinary excretion pH Bicarbonate Ca Na K Creatinine CTX

Water A Water B Water A Water B Water A Water B Water A Water B Water A Water B Water A Water B Water A Water B

mmol/24 h mmol/24 h mmol/24 h mmol/24 h mol/24 h mg/24 h

6.3 (0.46) 6.0 (0.50) 4.9 (5.52) 2.7 (2.36) 4.7 (1.80) 3.8 (1.90) 123 (35) 120 (25) 69.4 (26) 60.1 (18.1) 9779 (2362) 9619 (1781) 2.8 (1.02) 3.1 (2.04)

5.9 (0.34)* 6.7 (0.29) 1.3 (0.72)* 13.0 (7.84) 5.6 (1.48) 4.5 (1.41)* 110 (30) 143 (57) 67.4 (8.6) 70.5 (20.1) 9975 (1639) 10143 (1415) 2.8 (1.42) 3.2 (1.41)

6.0 6.6 2.1 13.8 5.5 4.5 119 146 73.4 78.7 10515 10542 2.7 3.1

(0.35) (0.43)# (1.92) (6.67)# (1.16) (1.58)# (41) (47) (18.5) (21.3)* (1635) (1823) (0.71) (1.46)

2 h fasting urine pH Bic/Creat. Calcium/Creat. CTX/Creat.

Water A Water B Water A Water B Water A Water B Water A Water B

mmol/mmol mmol/mmol mg/mmol

5.7 (0.52) 5.6 (0.5) 0.14 (0.20) 0.11 (0.13) 0.25 (0.15) 0.21 (0.13) 0.066 (0.040) 0.086 (0.066)

5.6 6.0 0.08 0.32 0.29 0.20 0.060 0.060

(0.36) (0.7) (0.06) (0.45) (0.16) (0.11) (0.041) (0.039)

5.5 6.3 0.08 0.75 0.22 0.17 0.048 0.079

(0.47) (0.6)# (0.09) (1.09)*# (0.13) (0.09) (0.026) (0.098)

Differences from baseline: *p b 0.05, p b 0.01 p b 0.001. Differences between Water A and B: #p b 0.05.

Laboratory tests Table 2 shows the analysed parameters in absolute values. The 4 gures present the changes between baseline, 2 weeks and 4 weeks in the urine and serum parameters except for pH which was expressed in absolute values. Blood tests The following parameters remained stable during the 4 weeks of the trial and showed no signicant differences between the two

groups: phosphorus, potassium, sodium, calcium, ionised calcium, alkaline phosphatase, albumin, creatinine and 25-OH vitamin D. In group B, mean serum CTX changes from baseline to week 4 were signicant (Fig. 1), as well as mean PTH changes (Fig. 2), while CTX and PTH stayed stable in group A. At week 4, the changes in serum CTX and PTH were signicantly different between the two groups. Urinary tests 24 h urine The urinary volume increased regularly in both groups between baseline and week 4 with no signicant difference between the two groups.

Fig. 1. Average changes (+/ SEM) of blood C-telopeptides according to water and study week.

Fig. 2. Average changes (+/ SEM) of blood PTH according to water and study week.

E. Wynn et al. / Bone 44 (2009) 120124

123

concentrations of uoride, chloride and phosphates content was negligible. Therefore, none of these ions can inuence the results of this study. Bicarbonate Bicarbonate and urine pH increased signicantly in Group B, showing the expected impact of Water B on the acid-base metabolism. The precision of the measurement of bicarbonate can be questioned as the urine must be collected immediately after emission and transported in an airless container. However, the results are concomitant with the change in bicarbonate intake, as already observed in a previous trial [20]. Potassium The increase of potassium excretion on Water B cannot be explained by the K contents of the waters which were low and similar, neither by the food intake which was controlled. It can be explained by the enhanced delivery of bicarbonate to the renal tubules which increases K excretion, the physiologic rationale being the possibility to eliminate more K [25]. Sulphate In general, calcium and sulphate contents in mineral water are positively correlated [26]. One can argue that the sulphate content of the two waters may play a role, since it was high in Water A. But in Water B, calcium is associated with bicarbonate but not with sulphate. In animals, sulphate supplementation increases calcium excretion [18,27,28]. But studies in humans show contradictory results: sulphates either have no effect [29] or increase calciuria due to their acidogenic action [30]. Indeed, acid load increases urine calcium excretion [30]. The acid load produced by the sulphate is included in the PRAL formula which shows the total acid load. When applied to the two mineral waters, it was estimated +9.2 mEq/l for water A (acid) and 11.2 mEq/l for water B (alkaline). Therefore, the effects we observed were enhanced by the acidity of water A and the alkalinity of water B and were not exclusively due to the difference in bicarbonate content. PTH Serum PTH decreased on water B but not on water A. This was surprising because total calcium, as well as ionized calcium remained stable. The same phenomenon was observed in a previous human study [8]. In dogs [31], metabolic alkalosis decreased PTH secretion also without any change in blood calcium, supposedly by modifying the calcium-sensors on the parathyroid glands. In our study, the calcium-rich water A did not lower the PTH level, perhaps because the nutritional calcium intake was already relatively high at baseline. This does not exclude a short-term effect [23], which however would not inuence the blood parameters of the next morning. Therefore, the

Fig. 3. Average (+/ SEM) of pH in 24 h urinary excretion according to water and study week.

pH increased in group B between baseline and weeks 2 and 4 and was signicantly different between both groups at 4 weeks (Fig. 3). Mean changes in urinary bicarbonate excretion between baseline and weeks 2 and 4 were signicant in group B and also signicant between both groups at week 4 (Fig. 4). Creatinine and sodium excretion stayed stable and showed no signicant differences between the 2 groups. Calcium excretion increased signicantly between baseline and week 4 in both groups (by 34.5 mg, SD 25.7 in group A p = 0.020, and by 27.7 mg, SD 12.8 in group B p = 0.019), with no signicant difference between the groups. When expressed in percent of the absorbed calcium (calculated according to reference 24), the urinary calcium excretion was 62k on Water A and 50k on Water B. Urinary CTX decreased slightly in both groups, but not signicantly, and no signicant differences between both groups were observed. Potassium excretion was signicantly different between both groups at week 4 (p = 0.033). It remained stable in group A, but increased signicantly between baseline and week 4 in group B (p = 0.0015). Fasting morning urine Expressed as ratio with creatinine, calcium, sodium, and CTX stayed stable and showed no signicant differences between the two groups. Bicarbonate increased in group B throughout the 4 weeks and was signicantly higher than in group A (p = 0.014). pH increased in group B between baseline and week 4 ( p = 0.0006) and was signicantly different between both groups at 4 weeks (p = 0.0001). Potassium showed an increase in group B at week 2 (p = 0.014) and week 4 (p = 0.013). Discussion A 4-week dietetic intervention with 1.5 daily litres of calcium-rich mineral waters showed that additional bicarbonate enhances inhibition of bone resorption. The effect of calcium-rich mineral water on bone resorption has already been shown, but in postmenopausal, calcium-decient women [3,23]. Our study was performed in calciumand oestrogen-sufcient women. Since all subjects followed the same diet plan, it can be assumed that all differences observed were not inuenced by changes in nutrition, but by the mineral waters only. Calcium The increase in urinary calcium excretion by approximately 30 mg/ 24 h is in line with the increased calcium intake by 800 mg/d from the mineral water. This is compatible with a fractional absorption rate of about 20k at the given intake of calcium. When expressed in percent of the absorbed calcium [24], the urinary calcium excretion was signicantly lower (p b 0.001) in Group B, perhaps an indirect sign of a decrease in bone resorption. Both waters had very low and similar

Fig. 4. Average changes (+/ SEM) of bicarbonate in 24 h urinary excretion according to water and study week.

124

E. Wynn et al. / Bone 44 (2009) 120124 [5] Cepollaro C, Orlandi G, Gonnellli S, Ferrucci G, Arditti J, Borracelli D, et al. Effect of calcium supplementation as a high-calcium mineral water on bone loss in early postmenopausal women. Calcif Tissue Int 1996;59:2389. [6] Costi D, Calcaterra P, Iori N, Vourna S, Nappi G, Passeri M. Importance of bioavailable calcium drinking water for the maintenance of bone mass in postmenopausal women. J Endocrinol Invest 1999;22(11):8526. [7] Bhmer H, Mller H, Resch KL. Calcium supplementation with calcium-rich mineral waters: a systemic review and meta-analysis of its bioavailability. Osteoporos Int 2000;11:93843. [8] Roux S, Baudoin C, Boute D, Brazier M, De la Gueronnire V, De Vernejoul M. Biological effects of drinking-water mineral composition on calcium balance and bone remodeling markers. J Nutr Health Aging 2004;8(5):3804. [9] Buclin T, Cosma M, Appenzeller M, Jacquet A, Dcosterd L, Boillaz J, et al. Diet acids and alkalis inuence calcium retention in bone. Osteoporos Int 2001;12:4939. [10] Krieger N, Frick K, Bushinsky D. Mechanism of acid-induced bone resorption. Curr Opin Nephrol Hypertens 2004;13:42336. [11] New S. The role of the skeleton in acid-base homeostasis. Proc Nutr Soc 2002;61:15164. [12] Kurtz I, Maher T, Hulter H, Schambelan M, Sebastien A. Effect of diet on plasma acidbase composition in normal humans. Kidney Int 1983;24:67080. [13] Bushinsky D. Acidbase imbalance and the skeleton. In: Burckhardt P, DawsonHughes B, Heaney R, editors. Nutritional Aspects of Osteoporosis. Proceedings of the 3rd International Symposium on Nutritional Aspects of Osteoporosis, Switzerland, 1997. Italy: Ares-Serono Symposia Publications; 1998. [14] Frassetto L, Curtis Morris R, Sebastian A. Effect of age on blood acidbase composition in adult humans: role of age-related renal functional decline. Am J Physiol 1996;271:F111422. [15] Bushinsky D. Acidosis and bone. Miner Electrolyte Metab 1994;20:4052. [16] Arnett T, Spowage M. Modulation of the resorptive activity of rat osteoclasts by small changes in extracellular pH near the physiological range. Bone 1996;18 (3):2779. [17] Tucker K, Hannan M, Chen H, Cupples A, Wilson P, Kiel D. Potassium, magnesium, and fruit and vegetable intakes are associated with greater bone mineral density in elderly men and women. Am J Clin Nutr 1999;69(4):72736. [18] Whiting S, Draper H. Effect of a chronic acid load as sulfate or sulfur amino acids on bone metabolism in adult rats. J Nutr 1981;111(10):17216. [19] Sebastian A, Harris S, Ottaway J, Todd K, Morris C. Improved mineral balance and skeletal metabolism in postmenopausal women treated with potassium bicarbonate. N Engl J Med 1994;330(25):177681. [20] Burckhardt P, Waldvogel S, Aeschlimann J, Arnaud M. Bicarbonate in mineral water inhibits bone resorption. J Bone Miner Res 2002;17(Suppl):M360S476. [21] Sellmeyer D, Schloetter M, Sebastian A. Potassium citrate prevents increased urine calcium excretion and bone resorption induced by a high sodium chloride diet. J Clin Endocrinol Metab 2002;87(5):200812. [22] Remer T, Manz F. Potential renal acid load of foods and its inuence on urine pH. J Am Diet Assoc 1995;95(7):7917. [23] Guillemant J, Huyen-Tran L, Accarie C, Tzenas du Montcel S, Delabroise AM, Arnaud M, et al. Mineral water as a source of dietary calcium: acute effects on parathyroid function and bone resorption in young men. Am J Clin Nutr 2000;71:9991002. [24] Heaney RP, Saville P, Recker R. Calcium absorption as a function of calcium intake. J Lab Clin Med 1975, June:88190. [25] Carlisle E, Donnelly S, Ethier J, Quaggin S, Kaiser U, Vasuvattakul S, et al. Modulation of the secretion of potassium by accompanying anions in humans. Kidney Int 1991;39(6):120612. [26] Wynn E, Raetz E, Burckhardt P. The composition of mineral waters sourced from Europe and North America in respect to bone health. Br J Nutr Sep 8:15 [Electronic publication ahead of print]. [27] Walser M, Browder A. The effect of sulfate infusion on calcium excretion. J Clin Invest 1959;38:140411. [28] Guguen L, Besanon P. Effect of sulfates on phosphorus and calcium metabolism. Utilization of calcium sulfate by sheep. Ann Biol Anim Biochim Biophys 1972;12 (4):58998. [29] Fardellone P, Bellony R, Texier C, Brazier M, Delabroise A, Arnaud M, et al. Comparative study of calcium bioavailability of a high-calcium mineralized water. Osteoporos Int 1996;6(Suppl 1):126. [30] Brandolini M, Guguen L, Boirie Y, Rousset P, Bertire M, Beaufrre B. Higher calcium urinary loss induced by a calcium sulphate-rich mineral water intake than by milk in young women. Br J Nutr 2005;93(2):22531. [31] Lopez I, Rodriguez M, Felsenfeld A, Estepa J, Aguilera-Tejero E. Direct suppressive effect of acute metabolic and respiratory alkalosis on parathyroid hormone secretion in the dog. J Bone Miner Res 2003;18(8):147885. [32] Jehle S, Zanetti A, Muser J, Hulter H, Krapf R. Partial neutralization of the acidogenic western diet with potassium citrate increases bone mass in postmenopausal women with osteopenia. J Am Soc Nephrol 2006;17:321322.

inhibition of PTH observed in this study, is supposed to be an effect of the alkali load, respectively of the trend of rising pH on parathyroid cells. This could also explain why bone resorption decreased. Whatever the main cause of the difference in bone resorption was, alkalinity versus acidity or high bicarbonate versus low bicarbonate, direct effect on bone or mediated through PTH, it is the composition of the mineral water which was determinant. Other studies have also shown a decrease in bone resorption through nutritional alkaline load, either by potassium bicarbonate [19], or an alkali diet with bicarbonate-rich mineral water [9] or potassium citrate [32]. The latter even increased bone mineral density. Our study allows us to presume, that even under a free diet and a high calcium intake, alkaline bicarbonate-rich mineral water inhibits bone resorption. Limitations Our study does have its limitations. The relatively free diet implied that the changes in urinary excretions showed some differences between individuals, and that only the major changes became statistically signicant. The number of subjects is low and the study might have lacked power to detect more effects. Although our subjects were dieticians and were fully qualied to follow the diet plans, we cannot exclude some mistakes. The serum and urine samples were taken to the laboratories in approximately 15 min, which could have inuenced the pH and bicarbonate data, but the differences between the two groups were consistent. Conclusion In conclusion, the present study showed that a bicarbonate- and calcium-rich alkali mineral water decreased bone resorption more than a calcium-rich acidic mineral water, in healthy pre-menopausal subjects on a calcium sufcient balanced diet. Further investigations need to be undertaken to study whether these positive effects are maintained on the long-term and if they can inuence bone mineral density. Acknowledgments EW: performed the study and wrote the manuscript. MAK: provided permanent scientic supervision. JMA: performed all statistical tests. PB: wrote the research protocol, directed and coordinated the study and supervised the manuscript. References
[1] Burckhardt P. Mineral Waters: Effects on bone and bone metabolism. In: Burckhardt P, Dawson-Hughes B, Heaney RP. editors. Nutritional Aspects of Osteoporosis. USA: Elsevier Science, 2004; 439, 447. [2] Burckhardt P. The effect of mineral waters on bone metabolism and bone health. In: New S, Bonjour JP, editors. Nutritional aspects of bone health. Cambridge: The Royal Society of Chemistry; 2003; 313, 321. [3] Meunier P, Jenvrin C, Munoz F, de la Gueronnire V, Garnero P, Menz M. Consumption of a high calcium mineral water lowers biochemical indices of bone remodelling in postmenopausal women with low calcium intake. Osteoporos Int 2005;16:12039. [4] Heaney RP. Absorbability and utility of calcium in mineral waters. Am J Clin Nutr 2006;84(2):3714.