Does bicarbonated mineral water rich in sodium change insulin sensitivity of postmenopausal women?

¿Modifica el agua mineral bicarbonatada rica en sodio la sensibilidad a la insulina de las mujeres postmenopáusicas?

S. Schoppen^*, F. J. Sánchez-Muniz^**, A. M.ª Pérez-Granados^*, J. A. Gómez-Gerique^***, B. Sarriá^*, S. Navas-Carretero^* and M.ª Pilar Vaquero^*

^*Department of Metabolism and Nutrition. Instituto del Frío. Spanish Council for Scientific Research (CSIC).
^**Department of Nutrition. Faculty of Pharmacy. Complutense University of Madrid.
^***Biochemical Service. Fundación Jiménez Díaz. Madrid. Spain.

ABSTRACT

Aim: To study the effects of drinking 0.5 L of two sodium-rich bicarbonated mineral waters (BMW-1 and 2), with a standard meal, on postprandial insulin and glucose changes. And to determine, if the effects vary depending on insulin resistance, measured by homeostasis model assessment (HOMA).
Methods: In a 3-way randomized crossover study, 18 healthy postmenopausal women consumed two sodiumrich BMWs and a low-mineral water (LMW) with a standard fat-rich meal. Fasting and postprandial blood samples were taken at 30, 60 and 120 min. Serum glucose, insulin, cholesterol and triacylglycerols were determined. Insulin resistance was estimated by HOMA and insulin sensitivity was calculated by quantitative insulin sensitivity check index (QUICKY).
Results: Glucose levels did not change. HOMA and QUICKY values were highly inversely correlated (r = -1,000; p < 0.0001). Insulin concentrations showed a significant time effect (p < 0.0001) and a significant water x time interaction (p < 0.021). At 120 min insulin levels with BMW-1 were significantly lower than with LMW (p = 0.022). Postprandial insulin concentrations showed significantly different patterns of mineral water intake depending on HOMA n-tiles (p = 0.016).
Conclusion: Results suggests an increase in insulin sensitivity after BMWs consumption. This effect is more marked in the women, who have higher HOMA values. These waters should be considered part of a healthy diet in order to prevent insulin resistance and cardiovascular disease.

Key words: Bicarbonated mineral water. Insulin. Glucose. Sodium. Fluoride.

RESUMEN

Palabras clave: Agua mineral bicarbonatada. Insulina. Glucosa. Sodio. Flúor.

Introduction

Natural mineral water intake has only been recently considered from a perspective of nutrition and health. Over the last decade several clinical studies have been directed to investigate the health benefits of mineral water and its relation to chronic diseases. Several studies have demonstrated a connection between drinking water and ischemic and cardiovascular diseases.^{1, 2} Other studies demonstrated that salt rich mineral waters are able to reduce serum total cholesterol^{3, 4} and postprandial triacylglycerol (TAG) - lipaemia.⁵ High fasting TAGs are associated to coronary heart disease risk and are one of the main characteristics of insulin resistance (IR),⁶ that is associated to diabetes and a number of other diseases, including obesity, hypertension, dyslipidemia, and coronary artery disease.^{7, 8} There is some evidence suggesting that postprandial TAG increase may also be related to insulin resistance.⁹

The metabolic syndrome (MBS) is a cluster of interrelated common clinical disorders, including obesity, insulin resistance, glucose intolerance, hypertension and dyslipemia. The risk of cardiovascular disease (CVD), attributed to MBS, appears to be especially high in postmenopausal women,¹⁰ as a result of the reduction of the positive effects of estrogens on lipid metabolism. 17-β-estradiol modulates the concentration of lipids in plasma by regulating lipid metabolism in adipocytes and hepatocytes.¹¹ This protective effect disappears after menopause.

The mechanism through which body fat contributes to IR is unclear, but previous studies have shown a correlation of body mass index (BMI) with insulin in postmenopausal women.¹²

Subjects and methods

Subjects

Eighteen postmenopausal women from the Menopause Program of the Madrid City Council Food and Health Department were recruited for the study. Postmenopause was defined as ammeorrheia status for at least one year. Subjects were 51 to 59 years old. Women in this prevention program periodically undergo clinical evaluation by means of anthropometric measurements, blood tests, bone mineral density determination and mammography. The women gave written informed consent to a protocol approved by the Ethics Committee of the Spanish Council for Scientific Research. No women suffered from any digestive or metabolic disease, as verified by medical history and fasting blood indices. Volunteers selected for the study could not be obese (BMI < 30 kg/m²), could not be receiving oestrogen replacement therapy or any other medication known to affect bone and lipid metabolism, or be taking vitamin, mineral or phytoestrogen supplements. None of the women smoked. Study participants were instructed not to deviate from their regular habits, and to maintain their normal diet and body weight, alcohol consumption and exercise levels.

Postprandial study

The subjects visited the laboratory facilities 3 times at 2-wk intervals. On each occasion the subjects fasted overnight for ≥ 12 h. In order to unify food intake, the women followed written instructions with regard to dinner composition (lettuce and tomato with olive oil, vinegar and salt; grilled chicken filet; bread and fruit) the evening before the study. On the morning of each intervention blood pressure, weight, height, and waist-to-hip ratio were measured and compliance with dinner instructions was verified with a questionnaire. After a cannula (ABOCATH 20G, Abbott Laboratories, Abbott Park, Illinois, USA) was inserted into a vein for blood sampling, baseline samples were obtained and the volunteers received the standard meal and 0.5 L of one of the study mineral waters. Three different mineral waters were used in the study: bicarbonated mineral water 1 (BMW 1), bicarbonated mineral water 2 (BMW 2), and low mineral water (LMW). The three waters were provided in 0.5 L bottles (Vichy Catalán, S.A.).

In a triple crossover design, volunteers were randomly assigned to an individual sequence of drinking the three mineral waters, until every woman had completed the postprandial study with all three waters. The two BMWs were both rich in bicarbonate, sodium and chloride, while the LMW was low in minerals (table I). BMW 1 contained 5.7 times more fluoride than BMW 2, and 37.5 times more than the LMW. The study meal provided 4,552 kJ, and contained 75.3 g of fat, 21.5 g of protein, 86.5 g of carbohydrates, and 289 mg of cholesterol. The energy profile was: protein 7.9% energy, lipids 62.3% energy and carbohydrates 29.8% energy, and the lipid profile was: SFA 11.8% energy, MUFA 39.7% energy, and PUFA 6.6% energy.

Postprandial blood samples were taken 30, 60 and 120 min after finishing the study meal. On the average, the meal was eaten in 31 min. To maintain hydration throughout the postprandial time, women drank 100 ml of demineralised water after 240 min and 390 min.

Blood samples were collected in Venoject tubes with Gel + Clot Activator for the serum separation.

Blood samples were chilled, and centrifuged immediately for 15 min at 1,500 x g and 4 ºC, and subsequently stored at -80 ºC.

Glucose was measured by an automatic analyzer (RA-XT Technicon, Tarrytown, NY). Insulin was determined in serum by means of an immunometric assay (Immulite^® 2000 Insulin, Diagnostic Products Corporation (DPC), CA, USA) with an IMMULITE 2000 autoanalyzer.

Serum cholesterol and TAGs (S-chol, S-TAG) were determined utilising automated enzymatic methods (CHOD-PAP and GPO-PAP; Boehringer Mannheim, Germany; RA-XT Technicon, Tarrytown, NY).

Assessment of insulin status and classification in HOMA and QUICKY n-tiles

No areas under the curve for glucose and insulin were calculated because adjusting four time points only does not result in valuable affirmations.

Insulin resistance was estimated using the homeostasis model assessment (HOMA; fasting insulin [μU/ml] x fasting glucose [mmol/L] / 22.5).¹³ Insulin sensitivity was estimated using the quantitative insulin sensitivity check index (QUICKY; 1/[log(fasting insulin mU/ml) + log(fasting glucose mg/dL)].¹⁴ HOMA values were classified in three n-tiles, one central n-tile (P25 to P75), and two n-tiles for both extreme ends (< P25 and > P75). HOMA n-tile 1 was defined as all HOMA values < 1.18 (< P25), n-tile 2 was defined as all HOMA values > 1.18 and < 1.89 (> P25 and < P75), and n-tile 3 was defined as all HOMA values > 1.89 (> P75).

Statistical analyses

Statistical analyses of the results were performed using SPSS 13.0 for Windows XP. Data are presented as mean and standard deviation. TAG data were log normalised for statistical analysis.

A two-factor repeated measures analysis of variance (ANOVA) was carried out for serum glucose and insulin for water, time and water × time interaction effects for all three waters. When significant time x water interactions were found, post hoc analysis using the Bonferroni test was performed to study the effects of water and time separately.

A one-factor repeated measures ANOVA followed by Bonferroni test was performed to study peak concentrations of serum glucose and insulin concentrations, and to determine the influence of HOMA and QUICKY n-tiles on baseline lipoprotein and anthropometric parameters.

The water effect on time to peak was calculated and analysed by the Friedman non parametric test.

Relation between HOMA and QUICKY values was analysed by Spearman correlation.

P values < 0.05 were considered statistically significant.

Results

All the 18 volunteers completed the study, having drunk the three waters. Attending to the three baseline values, mean values were calculated for the three HOMA n-tiles. Likewise, mean values were calculated for weight, BMI, waist and hip circumferences, W/H-Ratio, serum cholesterol and TAG. Table II shows the baseline characteristics of the study women. Baseline weight, BMI, and hip circumference presented significantly different values depending on the HOMA n-tiles (table III).

Figure 1 shows postprandial insulin levels depending on HOMA n-tiles and figure 2 presents the correlation between HOMA and QUICKY indexes.

Glucose levels did not change along the 2h postprandial period for any of the three waters tested (table IV). Insulin concentrations showed a significant time effect (p < 0.000) and a significant water x time interaction (p < 0.021) (table IV).

No significant differences of insulin and glucose peak concentration, and time to peak were observed for all three waters (data not shown).

Discussion

Worldwide obesity epidemia is driving an increasing burden of preventable CVD and the perspective of hyperinsulinemia and IR grows. Postprandial triglyceridemia is considered an independent risk factor for coronary heart disease, but its relationship with IR has not yet been clearly established. The studied BMWs have been shown to be hypolipemic, lowering postprandial serum and quilomicron triacylglycerols.^{4, 5} The present results indicate that postprandial insulinemia varies depending on water composition and insulin HOMA.

The fact that after 120 min insulin concentration with BMW 1 was lower than that corresponding to LMW may be due to increased insulin secretion and sensitivity after the consumption of BMW 1. Polushina et al.¹⁵ described significantly improved insulin sensitivity in patients with impaired carbohydrate tolerance and diabetes-2 patients after ingestion of a mineral water rich in sodium and bicarbonate. According to Krashenitsa and Botvineva,¹⁶ the insulinotropic effect varies with the mineralization of the water and the concentration of bicarbonate and sodium, which is in agreement to the results obtained considering the sodium (aprox. 1 g/L) bicarbonated (aprox. 2 g/L) mineral water in the present study.

Mean fasting insulin values of the study subjects were half the values considered cut-off point for IR (15 μU/ml) according to Ten and Mclaren.¹⁷ Only one woman presented an insulin value > 15 μU/ml. Thus, insulin data suggest, that in general terms, the volunteers were at very low risk for IR, and hence for MBS. According to ATPIII they were also at very low risk for CVD⁴ Results show that HOMA and QUICKY indexes are significantly correlated (fig. 2), and thus equally valid to assess differences in glucose and insulin metabolism. Therefore, all results are expressed for HOMA n-tiles. This correlation is higher than that reported by Katz et al.¹⁴

HOMA data suggest that all women presented an adequate insulin metabolism. Kuwana et al.¹⁸ calculated the reference values to be 1.77 ± 0.44 in healthy non obese individuals. Mean HOMA levels were lower (table II) than the risk cut-off point (2.67) suggested by Nasution et al.¹⁹ for elderly women. Ascaso et al.²⁰ using HOMA index P90 suggested the cut-off point of 3.8.

It is known that IR is strongly related to BMI²¹ and in the present study even in non-obese postmenopausal women we observed that BMI values were associated to HOMA and QUICKY levels (data not shown for QUICKY levels). Egan et al.²² suggest that BMI and HOMA indexes may be interchangeable, which is in agreement with the low HOMA levels found in these women. A waist-to-hip ratio cut point for a high risk of MBS is 0.85 in women and 0.9 in men, together with a BMI of 30 kg/m². In all 3 HOMA n-tiles mean waist-tohip ratios of all subjects were lower than this cut point. Our results confirm that body weight, BMI and hip circumference, but not waist circumference, were associated to HOMA n-tiles. This suggests a typical gynoid, and not a central fat distribution.

When the women were divided into 3 different groups according to HOMA n-tiles, different responses to mineral water intake were obtained. In the HOMA-3 group all insulin values at 30 min were higher than in the other two HOMA groups, especially when BMW 1 was consumed. Interestingly, when consuming BMW1 glucose levels of the HOMA-3 group were also slightly higher than with the other two waters (data not presented).

Recent changes in the pertinent European legislation have established that the fluoride content of mineral waters may not exceed 5.0 mg/l.²³ Consequently, BMW 2 substitutes for BMW 1 in the market. It has been reported that fluoride inhibits a variety of different enzymes such as esterases and phosphatases, but the influence of fluoride contained in water on insulin is very difficult to explain. Fluoride in this study seems to increase insulinemia mostly in the women with higher HOMA n-tiles (fig. 1), because insulin levels with BMW1 tend to be higher at 30 min than with BMW2. Other studies suggest that NaF shows an inhibition of insulin secretion in isolated Langerhans islets.²⁴ Nevertheless, both BMWs, which presented higher fluoride content than LMW, normalised insulinemia faster than LMW, suggesting that fluoride plays a role in insulin metabolism. De la Sota et al.²⁵ observed in subjects from a geographic area of endemic fluorosis that the area under the curve of insulin, corresponding to a standard glucose tolerance test, showed an inverse relationship with fluoremia. Differences between the later studies and the present one should be adscribed to fluoride concentration, being less than 50% of the water fluorosis level, and to the different study models used (rats and in vitro).

In addition, it has been reported that HCO₃- participates in the secretory sequence at sites distal to the identification of secretagogues in both endocrine end exocrine cells.²⁶ In the current study BMW 1 had the highest bicarbonate content; this fact could have increased the insulin secretion. This effect of mineral waters rich in bicarbonate has been described in medical hydrology literature.²⁷ Nevertheless, insulin concentrations at 120 min with BMW 1 was as low as with BMW 2, indicating that insulin degradation was faster with the two BMWs compared to the LMW, and/or that the insulinothrophic effect was not sustained during the whole postprandial 2 h. This was also reflected in the fact that comparing the two BMWs there was no significant differences observed, while BMW1 compared to LMW showed a significant water × time interaction.

Natural mineral waters rich in sodium bicarbonate have been extensively used in traditional crenotherapy as an accompanying component of diabetes therapy.²⁷ Debray and de la Tour²⁸ reported that alkalizing organic mediums favours insulinic actions. It has been reported that these mineral waters present a hypoglucemiant action and neutralize the hyperglucemiant response to adrenalin, cortisol, and aloxan, apart from the reinforcement of the insulinic action as evidences the potentiation of the "Staub effect" after glucose administration. ²⁷ In the present study there was no effect of subject's glucose response; this could be related to the composition of the test meal: low in carbohydrates (28.9%) and high in fat (62%). Carbohydrate content was mainly provided by a plain sponge cake, which presents a glycemic index of 46 ± 6.²⁹

The possible mechanisms behind these effects are not completely understood, but probably lay on the interaction of the total ionic strength and its multiple implications in physiological and pathological processes like neurosecretion, contraction, membrane stability, or enzymatic activation. There are reports which indicate the influence of high salt rich mineral waters on endocrine metabolism, pointing out various effects of these types of waters.^{30, 31} Unfortunately, these data are not easily accessible and more clinical trials have to be done in order to understand the exact mechanisms.

Conclusion

In short, bicarbonated sodium rich mineral water consumed with a fat rich meal by healthy postmenopausal women produced an insulinotropic effect, being more distinct in the women with a higher HOMA index and for the bicarbonated mineral water 1. Future studies are needed to test the effect of these bicarbonated waters on the insulinemia, glucemia and HOMA index of different obese/non obese populations consuming meals with different glycemic indexes.

We are indebted to Laura Barrios for her statistical advice and we would also like to thank the volunteers of this study. This work has been funded by Vichy Catalán S.A.

SS participated in the design and carried out the study, data and sample analysis and wrote the manuscript; FJSM participated significantly in the writing of the manuscript and provided advice and consultation; AMPG participated in the design and carried out the study, data and sample analysis and participated in the writing of the manuscript; JAGG participated in the writing of the manuscript and provided advice and consultation; BS carried out the experiment; SNC carried out the experiment; PV obtained the funding, designed and conducted the study and participated significantly in the writing of the manuscript.

All authors read and approved the final manuscript.

References

1. Nerbrand C, Agréus L, Lenner RA, Nyberg P, Svärdsudd K. The influence of calcium and magnesium in drinking water and diet on cardiovascular risk factors in individuals living in hard and soft water areas with differences in cardiovascular mortality. BMC Public Health 2003; 3:21.        [ Links ]

2. Rylander R, Arnaud MJ. Mineral water intake reduces blood pressure among subjects with low urinary magnesium and calcium levels. BMC Public Health 2004; 4:56.        [ Links ]

3. Capurso A, Solfrizzi V, Panza F et al. Increased bile acids excretion and reduction of serum cholesterol after crenotherapy with salt-rich mineral water. Aging (Milano) 1999; 11:273-6.        [ Links ]

4. Schoppen S, Pérez-Granados AM, Carbajal A et al. A sodiumrich carbonated mineral water reduces cardiovascular risk in postmenopausal women. J Nutr 2004; 134:1058-63.        [ Links ]

5. Schoppen S, Pérez-Granados AM, Carbajal A et al. Sodium bicarbonated mineral water decreases postprandial lipemia in postmenopausal women compared to a low mineral water. Br J Nutr 2005; 94:582-7.        [ Links ]

6. Austin MA. Plasma triglycerides and coronary heart disease. Arterioscl Throm Vas 1991; 11:2-14.        [ Links ]

7. Serrano Ríos M. Relationship between obesity and the increased risk of major complications in non insulin dependent diabetes mellitus. Eur J Clin Invest 1998; 28 (Supl. 2):S14-S18.        [ Links ]

8. Cruz ML, Evans K, Frayn KN. Postmenopausal lipid metabolism and insulin sensitivity in young Northern Europeans, South Asians and Latin Americans in the UK. Atheroscl 2001; 159:441-9.        [ Links ]

9. Schrezenmeir J, Fenselau S, Keppler I et al. Postprandial triglyceride high response and the metabolic syndrome. Ann NY Acad Sci 1997; 827:353-68.        [ Links ]

10. Wilsen PW, Kannel WB, Silberschatz H, D'Agostino RB. Clustering of metabolic factors and coronary heart disease. Arch Intern Med 1999; 159:1104-9.        [ Links ]

11. Masding MG, Stears AJ, Burdge GC, Wooton SA, Sandeman DD. Premenopausal advantages in postprandial lipid metabolism are lost in women with type 2 diabetes. Diabetes Care 2003; 26:3243-49.        [ Links ]

12. Chu MC, Cosper P, Orio F, Carmina E, Lobo RA. Insulin resistance in postmenopausal women with metabolic syndrome and the measurements of adiponectin, leptin, resistin, and ghrelin. Am J Obstetrics Gynecol 2006; 194:100-4.        [ Links ]

13. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: Insulin resistance and β-Cell function from fasting plasma glucose and insulin concentration in man. Diabetologia 1985; 28:412-9.        [ Links ]

14. Katz A, Nambi SS, Mather K et al. Quantitative Insulin Sensitivity Check Index: A simple accurate method for assesing insulin sensitivity in humans. J Clin Endocrin Met 2000; 85:2402-10.        [ Links ]

15. Polushina ND, Botvineva LA, Frolkov VK. Changes in tissue insulin sensitivity under the action of potable mineral waters (Clinic-Experimental Research). Vopr Kurortol Fizioter Lech Fiz Kult 1999; 6:16-9.        [ Links ]

16. Krashenitsa GM, Botvineva LA. The dynamics of glucose homeostasis in non-insulin-dependent diabetics under the influence of mineral water intake. Vopr Kurortol Fizioter Lech Fiz Kult 1992; 3:21-4.        [ Links ]

17. Ten S, MacLaren N. Insulin resistance syndrome in children. J Clin Endocrinol Metab 2004; 89:2526-39.        [ Links ]

18. Kuwana B, Urayama O, Hawai K. Reference value and cut-off value for diagnosis of insulin resistance in type 2 diabetes mellitus. Rinsho Byori 2002; 50: 398-403.        [ Links ]

19. Nasution IR, Setiati S, Trisnohadi HB, Oemardi M. Insulin resistance and metabolic syndrome in elderly women living in nursing homes. Acta Med Indones 2006; 38:17-22.        [ Links ]

20. Ascaso JF, Romero P, Real JT, Priego A, Valdecabras C, Carmena R. Insulin resistance quantification by fasting insulin plasma values and HOMA index in a non diabetic population. Med Clin 2001; 117:530-3.        [ Links ]

21. Molist-Brunet N, Jimeno-Mollet J, Franch-Nadal J. Correlation between the various measurements of obesity and the degree of resistance to insulin Aten Primaria 2006; 37:30-6.        [ Links ]

22. Egan BM, Papademetriou V, Wofford M et al. TROPHY Substudy Investigative Team: Metabolic syndrome and insulin resistance in the TROPHY sub-study: Contrasting views in patients with high-normal blood pressure. Am J Hypertens 2005; 18:3-12.        [ Links ]

23. Official Journal of the European Union. Commission Directive 2003/40/EC of 16 May 2003. ANNEX I:L126/37.        [ Links ]

24. Menoyo I, Rigalli A, Puche RC. Effect of fluoride on the secretion of insulin in the rat. Arzeimittelforschung 2005; 55:455-60.        [ Links ]

25. De la Sota M, Puche R, Rigalli A, Fernández LM, Benassati S, Boland R. Changes in bone mass and in glucose homeostasis in subjects with high spontaneous fluoride intake. Medicina (B. Aires) 1997; 57:417-20.        [ Links ]

26. Youssef R, Malaisse WJ, Courtois P, Sener A. Alteration of alpha-amylase secretion from rat parotid cells in the absence of extracellular bicarbonate. Int J M Med 2003; 12:199-200.        [ Links ]

27. Armijo Valenzuela M, San Martín Bacaicoa J. Curas balnearias y climáticas. Talasoterapia y helioterapia, Madrid: Editorial Complutense, 1994.        [ Links ]

28. Debray C, De la Tour C, Vaille C, Rozé C, Souchard M. Action de l'insuline, seule et en présence d'eau bicarbonateé sodique sur le sécrétion pancréatique et biliaire chez le rat. Therapie 1969; 14:283-95.        [ Links ]

29. Foster-Powell K, Holt SH, Brand-Miller JC. International table of glycemic index and glycemic load values: 2002. Am J Clin Nutr 2002; 76:5-56.        [ Links ]

30. Polushina ND, Agaev AA, Shchelkunov AV, Eseneev SM. Hormonal and metabolic effects of drinking mineral water and phytoaeronisation in experimental prostatitis. Vopr Kurortol Fizioter Lech Fiz Kult 2004; 6:25-7.        [ Links ]

31. Staforandova NV, Polushina ND, Istoshin NG, Nerovnia EA. Differential health resort drinking treatment of patients with chronic pancreatitis. Vorp Kurortol. Fizioter. Lech Fiz Kult 2004; 2: 8-20.        [ Links ]

Correspondence: ]]> Dra. Stefanie Schoppen.
Department of Metabolism and Nutrition.
Instituto del Frío, CSIC.
C/ José Antonio Novais, 10 - 28040 Madrid
E-mail: sschoppen@fjd.es, mpvaquero@if.csic.es

Recibido: 21-V-2007.
Aceptado: 6-VI-2007.