|Year : 2015 | Volume
| Issue : 2 | Page : 69-74
Serum and intracellular levels of ionized sodium, potassium, and magnesium in type 2 diabetic subjects
Md Erfan Reza1, Md Abdur Rashid1, Mahmuda Haque2, Farzana Pervin1, Liaquat Ali1
1 Department of Biochemistry and Cell Biology, Biomedical Research Group, Bangladesh Institute of Research and Rehabilitation for Diabetes, Endocrine and Metabolic Disorders (BIRDEM), Dhaka, Bangladesh
2 Department of Pharmacy, Southeast University, Dhaka, Bangladesh
|Date of Submission||08-Dec-2014|
|Date of Acceptance||05-Feb-2015|
|Date of Web Publication||23-Mar-2015|
Md Abdur Rashid
Labarotory of Molecular Signaling, National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institutes of Health (NIH), 5625 Fisher Lane, Room # 3S-O2, Maryland - 20852, USA
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Background and Aims: Alterations of ionized sodium, potassium, and magnesium in the serum or within erythrocytes have been reported in diabetes mellitus (DM) subjects, both as causes and consequences. There is also increasing evidence that electrolyte imbalances are early biochemical events responsible for long-term diabetic complications. Considerable variations in the electrolyte metabolism may exist in populations depending on the genetic constitution, nutritional status, and environmental situation. The present study was undertaken to investigate the serum and erythrocyte levels of Na + , K + , and Mg 2+ , and also to explore their relationship to glycemic status in a group of Bangladeshi type 2 diabetic patients without any complications. Materials and Methods: There were 30 newly-diagnosed type 2 diabetic subjects [age in years 45.17 ± 1.66; body mass index (BMI) 27.26 ± 1.59, M ± SD] who were studied with 30 age- and BMI-matched control subjects (age 46.30 ± 2.41; BMI 26.50 ± 1.78). Serum and intracellular concentration of Na + , K + , and Mg 2+ were estimated by the ion sensitive electrode (ISE) method using an Auto Analyzer (Nova Biomedical Corporation, 200 Prospect Street, Waltham, MA 02254-9141, USA). Serum glucose and lipid profile were measured by enzymatic colorimetric method. Results: The serum levels of ions (mmol/L, M ± SD) in the control subjects were as follows: Na + - 145 ± 1, K + - 3.78 ± 0.25, and Mg 2+ - 0.47 ± 0.02. In diabetic subjects, significantly lower value of Na + (143 ± 2, P < 0.0001) and Mg 2+ (0.44 ± 0.03, P < 0.0001) and higher value of K + (4.19 ± 0.41, P < 0.0001) were observed. Serum Na + and Mg 2+ showed negative correlation (P < 0.0001 for both ions) and serum K + showed positive correlation (P < 0.0001) with serum glucose only in the diabetic group. In these diabetic subjects, erythrocyte Na + was higher [(values in mmol/L, M ± SD): 11.20 ± 5.40 in diabetic subjects vs. 10.40 ± 4.77 in control subjects, P < 0.0001] and erythrocyte K + [(values in mmol/L, M ± SD): 139 ± 4.8 in control subjects vs. 133.06 ± 3.71 in diabetic subjects, P < 0.0001] and Mg 2+ [(values in mmol/L, M ± SD): 1.82 ± 0.22 in control subjects vs. 1.75 ± 0.20 in diabetic subjects, P < 0.0001] were found to be significantly lower as compared to control subjects. A significantly positive correlation between erythrocyte Na + (P < 0.0001) and a negative correlation between erythrocyte K + and Mg 2+ (P < 0.0001) were observed with serum glucose. Conclusions: These data confirm the existence of hyponatremia, hyperkalemia, and hypomagnesemia, paralleled by a reverse change of Na + and K + in erythrocytes of type 2 diabetic subjects. The study also demonstrates that hyperglycemia-induced effects on cellular transport process play a major role in mediating the electrolyte imbalances in diabetes.
Keywords: Hyperglycemia, hyperkalemia, hypomagnesemia, hyponatremia
|How to cite this article:|
Reza ME, Rashid MA, Haque M, Pervin F, Ali L. Serum and intracellular levels of ionized sodium, potassium, and magnesium in type 2 diabetic subjects. Int J Nutr Pharmacol Neurol Dis 2015;5:69-74
|How to cite this URL:|
Reza ME, Rashid MA, Haque M, Pervin F, Ali L. Serum and intracellular levels of ionized sodium, potassium, and magnesium in type 2 diabetic subjects. Int J Nutr Pharmacol Neurol Dis [serial online] 2015 [cited 2020 Jan 27];5:69-74. Available from: http://www.ijnpnd.com/text.asp?2015/5/2/69/153796
| Introduction|| |
Electrolytes are ionized particles, which carry either a positive or a negative charge. They are present in the body fluids as positively-charged ions (cations) and negatively-charged ions (anions), which migrate toward the cathode and the anode, respectively, when they are placed in an electrical field.  Alterations of ionized sodium, potassium, and magnesium in the serum or within the erythrocytes have been reported in diabetes mellitus (DM) subjects as both causes and consequences. There is also increasing evidence that electrolyte imbalances are early biochemical events responsible for long-term diabetic complications. Considerable variations in the electrolyte metabolism may exist in populations depending on the genetic constitution, nutritional status, and environmental situation.
In recent studies, it has been shown that ionized Na + plays an important role in β-cell function and modulation of insulin release. From different experiments and observations it has been shown that glucose augments insulin release by inhibiting the adenosine triphosphate (ATP)-sensitive K channels. These result in ionized Ca 2+ entry through voltage-gated channels. The increase in ionized Ca 2+ leads to exocytosis of the insulin secretory granule. It was found that ionized Na + plays an important role in the binding of agonist or antagonist to the α2 -receptor and in suppressing insulin release.  The role of sodium stimulus in the pancreatic β-cell secretion has remained a controversial issue. There is also preliminary evidence to suggest that the plasma membrane-bound enzyme Na + -K + -ATPase is nonenzymatically glycosylated in DM subjects.  The Na + - Na + -K + -ATPase and Mg + -ATPase activity of erythrocyte membrane in DM subjects were found to be significantly reduced.  It has been suggested that alterations in Na + , K + , and Ca 2+ and other biologically relevant elements might occur due to malfunction of Na + -K + pumps. There is increasing evidence that these alterations of electrolytes across the cell may play a vital role in the mechanism of cellular injury leading to retinopathy, nephropathy, and neuropathy in DM subjects.  Cellular damage, regardless of initial insult, is accompanied and may be mediated by disruption of subcellular element (Na + , K + , and Ca 2+ ) compartmentalization. 
Different studies have suggested that cellular potassium uptake, which is mostly dependent on Na + -K + ATPase activity, may be reduced in diabetic patients. Chronic hyperglycemia or insulin deficiency may be implicated in the pathogenesis of reduced enzyme activities, but the exact cause has not been elucidated yet.  Due to an altered transport system, diabetic subjects show increased K + level in the serum and decreased level within the erythrocyte.
Magnesium plays a unique role in carbohydrate metabolism. There is close association between magnesium deficiency and insulin resistance.  Intracellular Mg 2+ has a good correlation with intracellular K + , and Mg 2+ depletion could cause atherogenesis and hypercoagulability.  It has been shown that type 2 diabetic subjects have intracellular Mg 2+ deficiency, which may be a key factor leading to enhanced platelet reactivity in type 2 DM and can enhance the risk of vascular disease.  A recent study shows that diabetic ketoacidosis (DKA) subjects have higher potassium and lower sodium levels, but magnesium level is the same in subjects with and without DKA diagnosed on the basis of ketonemia.  Another study shows that β-cell secretory capacity was lower in impaired fasting glucose (IFG) subjects and significantly higher in impaired glucose tolerance (IGT) subjects, but insulin sensitivity was significantly decreased in IFG subjects and IGT subjects, suggesting that it may be a cause of electrolyte imbalance, which leads to development of type 2 diabetes. , Some studies suggest that total body Mg 2+ may be significantly depleted in some diabetic patients, especially with advanced retinopathy, and that Mg 2+ level may have some effect on the onset and development of diabetic retinopathy. The present study was undertaken to investigate the serum and erythrocyte levels of Na + , K + , and Mg 2+ , and also to explore their relationship to glycemic status in a group of Bangladeshi type 2 diabetic patients without any complications.
| Materials and methods|| |
The Ethical Committee of the Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders (BIRDEM) approved of the protocol for this study. There were 30 newly diagnosed type 2 diabetic subjects [age in years 45.17 ± 1.66; BMI 27.26 ± 1.59; (M ± SD)] who were studied with 30 age- and body mass index (BMI)-matched control subjects [age in years 46.30 ± 2.41; BMI 26.50 ± 1.78 (M ± SD)]. Serum glucose was measured by the glucose oxidase method. Serum cholesterol, serum triglyceride, and serum high-density lipoprotein (HDL) and low-density lipoprotein LDL were measured by enzymatic colorimetric. Serum erythrocytes Na + , K + , and Mg 2+ were estimated by the ion sensitive electrode method by using an ion sensitive electrode (ISE)-based Autoanalyzer.
Collection of blood samples
Collection of blood sample to obtain serum
On the day of appointment, a fasting sample (4 mL of blood) was drawn from the antecubital vein with all aseptic precautions in a plain glass test tube. Blood was allowed to clot for 20-30 min and centrifuged for 10 min at a rate of 3000 revolutions per minute (rpm). The serum was separated and quickly aliquoted into microcentrifuge tubes and stored in a freezer at −70°C for future analysis. The electrolytes were estimated on the same date of collection of the samples.
Collection of blood for determination of intracellular electrolytes
On the day of appointment, overnight fasting blood (at least 12 h) blood samples were collected from the antecubital vein in aseptic condition. Six mL of blood were taken in a heparin (10 μL/mL of blood)-containing glass test tube and was sealed with parafilm, and blood was mixed thoroughly by a gentle inversion. The test tube was kept in ice until the washing procedure started.
Washing of red blood cells (RBCs)
The heparinized blood was centrifuged at 4°C within 2 h of collection. The plasma and buffy coat were discarded. Then the packed cells were washed with ice-cold sucrose (300 mM) solution and centrifuged at 3000 rpm for 15 min. The washing procedure was repeated three times.
Preparation of homogenized RBC suspension
The washed RBCs were taken in a microcentrifuge tube and the same volume of Triton X-100 was mixed to break down the cells. Vortex was performed properly to ensure complete lysis of the cells. An appropriate buffer and deionized water were added for the parallel measurement of the electrolytes.
| Results|| |
Clinical and anthropometric parameters of the study subjects [Table 1]
Mean age (±SD) of the study subjects were 45.17 ± 1.66 years in control subjects and 46.30 ± 2.41 in diabetic subjects. Mean BMI (±SD) of the control subjects and diabetic subjects were 26.50 ± 1.78 and 27.26 ± 1.59, respectively. They were well matched. Systolic blood pressure (SBP) of the study subjects was 122.83 ± 5.20 mm Hg in control subjects and 127.17 ± 5.97 in diabetic subjects, and diastolic pressure was 79.33 ± 4.30 in control subjects and 79.33 ± 4.50 in diabetic subjects. SBP of the two groups differed significantly (P < 0.004), but diastolic blood pressure was similar in both the groups. The subscapular-triceps ratio (STR) was 1.86 ± 0.30 and 1.94 ± 0.36 in the control subjects and the diabetic subjects, respectively. They were well matched. Their waist-hip ratio (WHR) was 0.93 ± 0.02 and 0.94 ± 0.03 in the control subjects and the diabetic subjects, respectively. They were well matched too.
Fasting serum glucose and lipid profile of the study subjects [Table 2]
The fasting serum glucose (mmol/L, M ± SD) was 4.72 ± 0.50 in control subjects and 10.14 ± 1.74 in diabetic subjects. Glucose levels of the two groups differed significantly (t = -16.402, P = 0.0001). Except in case of total cholesterol, the diabetic subjects had a similar lipid profile with that of the control subjects (lipids in mg/dL: Tg 166.73 ± 44.33, HDL 32.67 ± 5.19, LDL 114.59 ± 23.54 in control subjects vs Tg 150 ± 37, HDL 33.67 ± 5.23, LDL 126.33 ± 13.86 in diabetic subjects, M ± SD). Total serum cholesterol was significantly higher in the diabetic subjects as compared to the control subjects (T-CHOL, in mg/dL: 180.60 ± 23.61 in control subjects vs. 196 ± 15.97 in diabetic subjects, M ± SD, P < 0.004).
Serum electrolytes of the study subjects [Table 3]
Serum Na + (mmol/L, M ± SD) level of the study subjects was 145 ± 2 in control subjects and 143 ± 2 in diabetic subjects. The diabetic subjects showed a significantly lower value when compared to the control subjects (t = 4.01, P = 0.0001). Serum K + (mmol/L, M ± SD) level of the study subjects was 3.78 ± 0.25 in control subjects and 4.19 ± 0.41 in diabetic subjects. The diabetic subjects showed a significantly higher level of serum K + compared to the control subjects (t = -4.673, P = 0.0001). Serum Mg 2+ (mmol/L, M ± SD) level of the study subjects was 0.48 ± 0.02 in control subjects and 0.44 ± 0.02 in diabetic subjects. The diabetic subjects showed a significantly lower value compared to the control subjects (t = 4.04, P = 0.0001).
Intraerythrocyte electrolyte concentration of the study subjects [Table 4]
Erythrocyte Na + concentration (mmol/L, M ± SD) levels of the study subjects were 10.40 ± 4.77 in control subjects and 11.20 ± 5.40 in diabetic subjects. The two groups of subjects did not show any significant difference (t = 0.608, P = 0.54). Erythrocyte K + concentration (mmol/L, M ± SD) level of the study subjects was 139 ± 4.81 in control subjects and 133 ± 3.71 in diabetic subjects. The diabetic subjects showed a significantly lower value when compared to the control subjects (t = 5.06, P = 0.0001). Erythrocyte Mg 2+ concentration (mmol/L, M ± SD) levels of the study subjects were 1.82 ± 0.22 in control subjects and 1.72 ± 0.20 in diabetic subjects. The two groups of subjects did not show any significant difference (t = 1.25, P = 0.213).
Correlation analysis between fasting serum glucose and serum electrolytes [Table 5]
A negative correlation was observed in case of diabetic subjects between serum fasting glucose and Na + and Mg 2+ levels (Na + , r = -0.87, P = 0.0001; Mg 2+ , r = 0.88, P = 0.0001), but no such correlation was found in the control subjects. The diabetic subjects also showed a positive correlation in the case of K + levels (r = 0.83, P = 0.0001), which was not observed in the control subjects.
|Table 5: Correlation study between fasting serum glucose levels and serum electrolytes of the study subjects|
Click here to view
Correlation analysis between fasting serum glucose and erythrocytes Na + , K + , and Mg 2+ [Table 6]
A significant positive correlation was observed in case of diabetic subjects between serum fasting glucose and erythrocyte Na + levels (Na + , r = 0.70, P = 0.0001), and significant negative correlation was found with K + and Mg 2+ levels (K + , r = -0.70, P = 0.0001; Mg 2+ , r = -0.88, P = 0.0001). No correlation was observed in the case of control subjects.
|Table 6: Correlation study between fasting serum glucose and RBC electrolytes|
Click here to view
Correlation analysis between fasting serum glucose and lipid [Table 7]
A significant positive correlation was observed between serum fasting glucose and total cholesterol (r = 0.54, P = 0.002) and LDL cholesterol (r = 0.51, P = 0,003). No correlation was observed in the case of HDL cholesterol and triglyceride in diabetic subjects; also, no correlation was observed in the case of control subjects.
|Table 7: Correlation study between fasting serum glucose and lipid level of the study subjects|
Click here to view
| Discussion|| |
One of the major objectives of the present study was to develop a technique for the measurement of some physiologically important ions, i.e., Na + , K + , and Mg 2+ , in the serum as well as in the erythrocytes. So far, such a technique has not been reported in our country and no study has been done specifically to investigate the relationship between serum and erythrocyte ionized sodium, potassium, and magnesium in DM subjects. Thus, the physiological role of the ions in a real life situation has remained subject to speculation to a certain extent.
It is satisfying that such a technique could be initiated in this study and it is even more rewarding that measurement has been standardized by using a commercially available ISE-based Autoanalyzer which will facilitate reliability on the experiment by other groups. Although the present investigation covers only a section of healthy subjects and diabetic subjects in our population, it will give important methodological and technical standardization for further studies in this field. It may also have a beneficial impact on clinical practice.
The ionized forms of Na, K, and Mg are biologically the most relevant fractions of these electrolytes and their levels are critically maintained in the body fluids. The present study is the first attempt for a parallel estimation of these ions in the serum as well as in the RBCs of the Bangladeshi population using an advanced analytical technique, which can reliably measure only the biologically active, ionic forms of the physiologically important cations. In healthy control subjects, the serum levels (mmol/L, M ± SD) of the electrolytes have been found as follows: Na + : 145 ± 1, K + : 3.78 ± 0.26, and Mg 2+ : 0.48 ± 0.02. In comparison to healthy European subjects, it may be seen that the value of Na + is marginally higher, and that K + is at the lowest level of the reference ranges, but the level of Mg 2+ is substantially below the corresponding reference range (European range: mmol/L, Na + : 132-144, K + : 3.6-5.1, Mg 2+ : 0.75-1).  On the other hand, intraerythrocytic levels, M ± SD in mmol/L, Na + , K + , and Mg 2+ were 10.40 ± 4.77, 138.70 ± 4.8, and 1.82 ± 0.22, respectively. In comparison with the reference range of healthy European people, it was shown that the value of intraerythrocytic Na + was higher but the levels of K + and Mg 2+ were below the range (Na + = 10 mmol/L, K + = 160 mmol/L and Mg 2+ = 2.5 mmol/L).  Although a larger population is required to establish the reference range for these ions, the results of the present study may at least indicate the importance of establishing our own reference values. Apart from ethnic variations, cultural habits like high table salt intake may influence and modulate the internal milieu of our body fluids and these aspects should be explored carefully.
Electrolyte imbalance may complicate both acute and chronic metabolic abnormalities of diabetes. Thus, it is very important to have a baseline data for diabetic patients regarding their plasma as well as intracellular electrolytes, particularly RBC electrolytes. Again, the present study is the first attempt in this line. In the present study RBC electrolyte level of diabetic subjects were, in mmol/L, M ± SD, Na + : 11.20 ± 5.4, K + : 133 ± 3.71, and Mg 2+ : 1.75 ± 0.20. In control subjects, the levels were Na + : 10.40 ± 4.77, K + : 139 ± 4.80 and Mg 2+ : 1.82 ± 0.22, respectively. The RBC electrolyte status of the diabetic subjects as compared to that of the control subjects show that the intracellular level of Na + and Mg 2+ do not differ significantly between the two groups, whereas the level of K + is significantly lower (P < 0.0001) in diabetic subjects.
The data of the present study show a lower serum level of Na + and a tendency for higher intracellular Na + in the diabetic subjects in comparison to the control subjects. Glucose level shows inverse correlation with serum Na + and positive correlation with intracellular Na + . Thus, the lower level of serum Na + in diabetic patients is possibly due to a hyperglycemia-induced intracellular sequestration of Na + (reflected by a tendency of higher Na + in the RBCs, [Table 4]) combined with increased Na + excretion as a result of osmoregulatory response.  It is now necessary to conduct parallel measurements of serum and intracellular and urinary Na + (along with serum glucose) in the diabetic patient to explore the relative role of sequestration versus excretion of Na + in the pathogenesis of hyponatremia.
The data regarding the K + level show higher serum levels and lower intracellular levels in the diabetic subjects compared to the control subjects. Glucose levels show positive correlation with serum and inverse correlation with intracellular K + levels. A simple explanation based on electrical neutrality to balance the entry of Na + in the cell may be the K + efflux in response to hyperglycemia.
Both serum as well as intracellular levels of Mg 2+ are lower in diabetic subjects compared to control subjects. Both serum and intracellular Mg 2+ show an inverse correlation with glucose levels. As a substantial decrease in serum Mg 2+ is not paralleled by the rise in intracellular Mg 2+ , it seems that the theory of Mg 2+ sequestration in response to hyperglycemia is implausible and there might be a possibility that Mg 2+ has been excreted, which supports our previous study. 
The reason for hyponatremia and hyperkalemia in the plasma as well as the tendency for decreased value of potassium within erythrocyte in the diabetic subjects compared to the control subjects are still not definitely known. However, depression of the function of Na-K-ATPase may be one of the important causes as postulated by others.  Apart from this, change in membrane fluidity may be another cause, which complicates the normal transport of ions across the cell membrane. In a study, it was shown that the mean erythrocyte Na + and K + content of diabetic patients and control subjects were very similar. The ATPase activity of the erythrocyte membrane was, on contrary, significantly reduced in diabetic patients compared to the control subjects (P < 0.005).  Decreased magnesium levels in serum had been reported earlier in the Bangladeshi diabetic population.  There is a remote possibility that these electrolyte changes may influence the chemical events responsible for long-term diabetic complications. We have to consider further that intraerythrocytic electrolyte levels may have therapeutic effects, especially in the case of potassium levels. Moreover, as assumed by others,  the intracellular electrolyte levels that reflect the insulin effect should be taken into consideration in the management of patients by insulin.
It may be concluded that serum levels of Na + and Mg + are significantly lower while the serum level of K + is higher in diabetic subjects. In the erythrocytes of diabetic subjects, there is a tendency of higher Na + , lower K + , and lower Mg 2+ in comparison to control subjects. These data confirm the existence of hyponatremia, hyperkalemia, and hypomagnesemia, paralleled by a reverse change of Na + and K + in erythrocytes of type 2 diabetic subjects. These data suggest that hyperglycemia-induced effects on cellular transport process play a major role in mediating the electrolyte imbalances in diabetic patients.
| References|| |
Black D. General properties of body fluid. In: Black D, editor. Essentials of Fluid Balance. 2 nd
ed. Oxford: Thomas, Springfield; 1960. p. 11.
Tian WN, Duzie E, Lanier SM, Deth RC. Determinants of alpha 2-adrenergic receptor activation of G proteins: evidence for a precoupled receptor/G protein state. Mol Pharmacol 1994; 45:524-31.
Graner MH, Spector A. Glucose-6-phosphate modification of bovine renal Na,K-ATPase: A model for changes occurring in the human renal medulla in diabetes. Biochem Biophys Res Commun 1985;131:1206-11.
Finotti P, Palatini P. Reduction of erytherocyte (Na+-K+) ATPase activity in type 1 (insulin-dependent) diabetic subjects and its activation by homologous plasma. Diabetologia 1986;29:623-8.
Lowery JM, Eichberg J, Saubermann AJ, LoPachin RM Jr. Distribution of elements and water in peripheral nerve of streptozocin-induced diabetic rats. Diabetes 1990;39:1498-503.
LoPachin RM Jr, LoPachin VR, Sanbermann AJ. Effects of axotomy on distribution and concentration of elements in rat sciatic nerve. J Neurochem 1990;54:320-32.
Mimura M, Makino H, Kanatsuka A, Yoshida S. Reduction of erythrocyte (Na(+)-K +
) ATPase activities in non-insulin-dependent diabetic patients with hyperkalemia. Metabolism 1992;41:426-30.
Paolisso G, Scheen A, D′Onofrio F, Lefèbvre P. Magnesium and glucose homeostasis. Diabetologia 1990;33:511-4.
Fujii S, Takemura T, Wada M, Akai T, Okuda K. Magnisium levels of plasma, erythrocyte and urine in patients with diabetes mellitus. Horm Metabol Res 1982;14:161-2.
Nadler JL, Malayan S, Luong H, Shaw S, Natarajan RD, Rude RK. Intracellular free magnesium deficiency plays a key role in increased platelet reactivity in type II diabetes mellitus. Diabetes Care 1992;15:835-41.
Rashid MA, Chowdhury HS, Haque M, Faruque MO, Chowdhury MR, Liaquat A. Role of measurement of blood ketone bodies in the management of diabetic ketoacidosis. Int J Nutr Pharmacol Neurol Dis 2013;3:335-40.
Rashid MA, Faruque MO, Haque M, Karim MR. Association of amylin in the development of impaired fasting glucose and impaired glucose tolerance. Int J Nutr Pharmacol Neurol Dis 2013;3:347-51.
Islam MN, Hossain M, Hafizur RM, Khan I, Rashid MA, Shefin MS, et al
. Ratio of fasting glucose to adiponectin is an important predictor for the development of type 2 diabetes. J Diabetol 2011; 3:1-7.
Haslett C, Chilvers ER, Hunter JA, Boon NA. Biochemical and haematological values, Appendices. In: Haslett C, Chilvers ER, Hunter JA, Boon NA, editors. Davidsons Principle and Practice of Medicine. 18 th
ed. UK: Churchill Living Stone; 1999. p. 1135.
Swaminathan R. Renal handeling of magnesium. In: Davison A, Cameron S, Grunfeld J, Ponticelli C, editors. Oxford Text Book of Clinical Nephrology. 2 nd
ed. Vol. 1. UK: Oxford University Press; 1998. p. 271-94.
McNair P, Madsbad S, Christainsen C, Christensen MS, Transbøl I. Hyponatremia and hyperkalemia in relation to hyperglycemia in insulin related diabetic out-patients. Clin Chem Acta 1982;120:243-50.
Khan LA, Alam AM, Ali L, Goswami A, Hassan Z, Sattar S, et al
. Serum and urinary magnesium in young diabetic subjects in Bangladesh. Am J Clin Nutr 1999;69:70-3.
Román F, Toldi Z, Kürti K, Pataki L. Red blood cell electrolyte changes in patients with juvenile diabetes mellitus. Acta Paediatr Hung 1990;30:233-9.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7]