Factors contributing to the degree of polyuria in a patient with poorly controlled diabetes mellitus

Factors Contributing to the Degree of Polyuria in a Patient With Poorly Controlled Diabetes Mellitus Avrum Spira, MDCM, Manjula Gowrishankar,

MD, and Mitchell L. Halperin, MDCM

0 Polyuria due to a glucose-induced osmotic diuresis is common in patients with hyperglycemia. This diuresis usually abates when the plasma glucose level approaches its renal threshold; the usual time course is less than 8 hours after commencing therapy. A 69-year-old man with non-insulin-dependent diabetes mellitus maintained hyperglycemia (540 mg/dL) and polyuria (4.7 L/24 hr) for 40 hours. Because there was no external supply of glucose, a balance study was conducted between the third and 40th hour after commencing treatment. In this interval, the overall concentration of glucose in the urine was less than 100 mmol/L and the urine osmolality was 378 mOsm/kg H20. To evaluate the expected composition of the urine during a glucose-induced osmotic diuresis, urine was analyzed in normal rats infused with glucose plus urea and in untreated BB diabetic rats (plasma glucose and urea similar to that in our patient) as well as in 29 patients with hyperglycemia and polyuria. Glucose accounted for 60% of the urinary osmoles in rats and humans. Two subgroups of patients had a much lower urine glucose: one had an impaired concentrating ability (n = 6) and the other had an increased rate of renal glucose reabsorption (n = 5). In conclusion, in polyuria caused by hyperglycemia, the urine glucose should be 300 to 400 mmol/L with normal renal function. In the case we report, both the concentration of glucose and its excretion rate were much lower than expected with steady-state hyperglycemia (540 mg/dL) due to the high rate of excretion of NaCI, a concentrating defect, and excessive renal reabsorption of glucose. 0 1997 by the National Kidney Foundation, Inc. INDEX

WORDS:

Gentamicin;

glucose;

insulin;

osmotic

diuresis;

T

HE HORMONAL setting that permits hyperglycemia to develop is one with a relatively low level of insulin and/or high levels of hormones with actions that oppose those of insulin.’ When a patient has extracellular fluid volume contraction and an infection, the levels of adrenergic hormones should be elevated, which could cause a further inhibition of the release of insulin by an a-adrenergic effect.2 Once the hormonal setting permits hyperglycemia to develop, how severe it will be should be considered. Two questions become important: how much glucose was ingested and/or produced and what is the rate of renal excretion of glucose? The ingestion of a large quantity of fruit juices or sweetened soft drinks that have a high content of sugar3 may lead to a more severe degree of hyperglycemia in a patient with a relative lack of insulin. Because poor gastric emptying is a common finding in patients with diabetes mellitus in poor glycemic control, a high intake of sugar can prolong the duration of hyperglycemia once treatment is instituted if glucose was retained in a large dilated stomach. In this setting, one might observe a somewhat longer duration of polyuria with a urine composition typical of glucose-induced osmotic diuresis (urine glucose close to 54 g/L, sodium [Na+] close to 50 mmol/ L, and potassium [KC] close to 30 mm01/L).~ Glucose can be made endogenously as well. To be quantitatively important when the duration American

Journal

of Kidney

Diseases,

Vol 30, No 6 (December),

polyuria;

renal

glucose

reabsorption.

of the illness exceeds 24 hours (time to deplete liver glycogen), the source of endogenous glucose should be muscle glycogen or protein (true gluconeogenesis [GNG]). For the former, one typically finds excessive muscular activity (seizure, tremors, exercise); to establish the presence of the latter, the rate of urea appearance is measured using a stoichiometry of 1.72 urea/glucose formed in true GNG.4 Glucose can be removed by metabolic means (oxidation or conversion to a storage compound such as glycogen or triglycerides) or by renal excretion.4 A low rate of excretion of glucose can be anticipated if there is either a low glomemlar filtration rate (GFR) and/or an enhanced renal reabsorption of glucose. 6-SThe focus of the present report is to consider aspects of glucose excretion and polyuria in quantitative terms. The patient to be reported had an unexpectedly prolonged, severe degree of hyperglycemia (540 mg/dL); it was sustained for 37 hours despite

From the Renal Division, St Michael’s Hospital, University of Toronto, Toronto, Canada. Received March 10, 1997; accepted in revised form June 20, 1997. Address reprint requests to Mitchell L. Halperin, MDCM, Division of Nephrology, St Michael’s Hospital Annex, 38 Shuter St, Toronto, Ontario, M5B lA6. E-mail:mitchell. [email protected] 0 1997 by the National Kidney Foundation, Inc. 0272-6386/97/3006-0015$3.00/O 1997:

pp 829-835

829

830

SPIRA, Table

1. Laboratory in the

Index

Data on Admission Case Plasma

Glucose (mg/dL) Creatinine (mg/dL) Na+ (mmol/L) K+ (mmol/L) Cl- (mmol/L) HC03(mmol/L) Anion gap (mEq/L) PH PcoZ (mm Hg) Albumin (g/L) BUN (urea) (mg/dL) Lactate (mmol/L)

Table

882 2.2

-

128 5.5 95 17 16 7.42 27 31 78 1.7

43 22 55 0 10 -

REPORT

A 69-year-old man was found in a confused state. He did not recall recent events other than that he felt weak and thirsty for several weeks before admission. He did not take any medication, smoke, use alcohol, or take drugs. On physical examination, he was rousable, but confused and uncooperative; he was oriented only to person. His blood pressure was 125/75 mm Hg; postural changes in blood pressure and pulse rate were not assessed because of noncompliance. Of note, his blood pressure was 150/80 mm Hg after recovery. His pulse rate on admission was 110 beats/min and regular, respiratory rate was 20 breaths/min, and temperature was 36.7”C. His jugular venous column height was observed at the level of the sternal angle. His heart sounds were normal, as was an abdominal examination. There were no other positive findings noted.

Laboratory

Data

The blood sugar was very high (880 mg/dL), plasma (HCO,-) was modestly reduced, plasma anion gap was somewhat elevated (16 mEq/L, plasma albumin 31 g/L), and the blood pH was normal (Table 1). The serum ketones were only modestly positive in an undiluted sample and the concentration of creatinine in plasma was 2.2 mg/dL.

Course in Hospital Treatment in the first 3.5 hours consisted of isotonic saline (1.5 L) and insulin (5 U bolus intravenously and 5 U/hr for 3 hours; total, 20 U). Because white blood cells and bacteria were found in his urine, he received 1 g ampicillin and 80

AND

Data for the 37-Hour in the Index Case

Urine

polyuria and the absence of a large input of glucose. Further analysis revealed a much lower than anticipated concentration of glucose in his urine, and hence a low rate of glucosuria. The basis for the lower than expected glucosuria and an analysis of the pathophysiology of his polyuria will be considered in the Discussion section. CASE

2. Plasma Course

GOWRISHANKAR,

HALPERIN Hospital

Time (hr) 6.5

Plasma Glucose (mg/dL) Creatinine (mg/dL) BUN (urea) (mg/dL) Na+ (mmol/L) K’ (mmol/L) Cl- (mmol/L) HC03(mmol/L) Anion gap (mEq/L)

32.5

20.5

522

504

522

1.7 59 140 5.5 llrl 18 12

1.5 48 141 4.4 113 16 12

1.5 39 141 4.3 110 17 14

NOTE. For details, see text. The point to emphasize is that there was no appreciable decline in glycemia over this 37-hour period. For simplicity, the beginning of this 37-hour period is considered to be 0 hours; this represents 3.5 hours after initiation of treatment.

mg gentamicin intravenously. His urine output was 0.7 L in the first 3.5 hours. The feature that dominated management over the next 37 hours was persistent hyperglycemia (540 mg/dL), despite continuing therapy with insulin (Table 2). Total fluid input in this period was 8.5 L of the equivalent of isotonic saline, 1.5 L of water, and 320 mmol KCl. Urine output was 5.5 L; no glucose was administered. The concentration of glucose in his urine was only 10 g/L in the first collection period and 24 g/L in the second collection period (Table 3). There was no evidence of diabetes insipidus in this patient.

Data From Other Patients With Hyperglycemia Data were analyzed from a retrospective analysis of 29 patients who presented with a severe degree of hyperglycemia

Table

3. Urine Composition After Initiating

for the Therapy

37 Hours

Collection O-18 hr

Volume (L) Na+ (mmol) K+ (mmol) Cl- (mmol) Urea (mmol) Glucose (mmol) Glucose (g) Osmolality (mOsm/kg Urea clearance (U24

H,O) hr)

2.7 178 (53) 46 (14) 213 (64) 260 (79) 146 (54) 26 (10) Not done 19

Period 18-37

hr

2.8 156 (47) 57 (17) 195 (60) 298 (138) 378 (135) 68 (24) 377 27

NOTE. Urine was collected over two time periods. The clearance of urea was calculated because the plasma urea concentration did not vary markedly over the period of observation. Concentrations are shown in parentheses.

HYPERGLYCEMIA

AND

OSMOTIC

831

DIURESIS

(plasma glucose, 650 mg/dL; blood urea nitrogen, 3 1 mg/dL) in whom the urine osmolality and glucose concentration were measured (data from Halperin, unpublished observations, n = 20; and Seldin and Tarail,’ n = 9). There was a linear relationship between the urine glucose concentration and osmolality. A lower urinary glucose concentration was found in patients who reabsorbed a larger than anticipated amount of glucose (contained in the oval in Fig 1) or in those with a lower urine osmolality, including our index case (contained in the square in Fig I).

800

1

Studies in Rats Two groups of rats were studied. First, normal rats (n = 6) were anesthetized with inactin (100 mg/kg). A femoral artery canula was inserted for blood sampling, a femoral vein cannula for infusion of glucose and urea, and a urinary bladder cannula to obtain urine samples. Rats were given 1.2 pg desmopressin acetate (DDAVP) intravenously before the infusion. Following an initial bolus of 2.5 mL of 50% dextrose in water, the infusion was changed to one containing 6 g/dL urea and 10 g/dL glucose; its rate of administration was adjusted to match urine output. Urine was collected every 20 minutes, and blood samples were drawn at times 0, 70, and 140 minutes. Second, BB diabetic rats (n = 6) had insulin withheld for 24 hours before commencement of the study. These rats were housed in individual metabolic cages. Blood samples and timed urine collections were obtained after the rats developed glucosuria.

h0

600

Urine

Osmolality

1000

(mOsm/kg

H20

Fig 1. Composition of urine in a glucose-induced osmotic diuresis. The urine osmolality is shown on the x-axis and the concentration of glucose (mmol/L) is shown on the y-axis. The solid circles represent the data in rats; the open circles represent the hyperglycemic patients. The line of best fit y = 0.60x - 55, r2 = 0.575. The square denotes patients with a lower urinary glucose concentration due to a concentrating defect; the oval denotes patients with an enhanced renal reabsorption of glucose.

RESULTS

Hyperglycemia (468 ? 50 mg/dL and 648 t 81 mg/dL) and elevated levels of urea in plasma (76 + 16 mg/dL and 53 t 16 mg/dL) were present in the normal and BB rats, respectively. The relationship between their urine osmolality and the concentration of glucose in the urine is depicted in Fig 1. Glucose accounted for approximately 60% of the osmoles excreted. DISCUSSION

The central issue to explain is how a relatively severe degree of hyperglycemia (522 mg/dL) was sustained for 40 hours in the face of such a large urine output. The presentation will be organized along four lines: first, how glucose balance could be maintained for 40 hours with this degree of hyperglycemia and polyuria; second, the expected values for the urine osmolality during a glucose-induced osmotic diuresis; third, why so little glucose was excreted; and fourth, the specific cause for polyuria in this patient. Balance for Glucose To sustain a marked degree of hyperglycemia for 40 hours, there must be a high input and/or

a markedly decreased rate of removal of glucose. When the degree of hyperglycemia is severe, excretion of glucose becomes a quantitatively important consideration.4 Two calculations help to illustrate this point. First, 1 L of a glucoseinduced osmotic diuresis typically contains 55 g of glucose (Fig 1).4,6-8With 5.5 L of urine, close to 300 g of glucose should be excreted. Second, the amount of glucose in our patient’s body is only 104 g because glucose distributes in the extracellular fluid plus the intracellular fluid (ICF) of organs such as the liver, in which insulin does not influence glucose transport (20 L in our patient”) (blood sugar of 522 mg/dL or 5.22 g/ L X 20 L). Because there was no decrease in glycemia and his glucose pool size did not change over this time (Table 2), the 300 g of glucosuria would require a production or intake of that much glucose. Our patient would have needed 5 L of 5% dextrose in water to receive 300 g of glucose, but no glucose was given. It is unlikely that 300 g of glucose was in his stomach on admission. There also was no obvious alternate endogenous

832

source of glucose because there should be little glycogen in his liver (normally 90 g in the fed state), there was little glucose precursor, lactate, in his extracellular fluid (plasma lactate 1.7 mmol/L; Table l), and there was no apparent stimulus for excessive breakdown of glycogen in his muscle (a seizure). Because GNG and ureagenesis share a common metabolic pathway, the syntheses of new glucose and urea from amino acids are obligatorily linked.5,” His rate of appearance of urea over 37 hours is the amount of urea excreted (558 mmol; Table 3) minus the decline in his urea pool size (21 mmol/L - 14 mmol/L [Table 21 X 40 L total body water), or 278 mmol. With a stoichiometry of 1.72 ureas per glucose produced by GNG,’ only 29 g rather than 300 g of glucose could be produced via GNG. Glucose removal by oxidation could be low if there was oxidation of fat-derived fuels.” Given the additional “stress” of a urinary tract infection and the moderate plasma test for ketoacids, it is possible that little glucose was oxidized. The source of glucose for oxidation could have been the small amount of glucose production by GNG plus any glucose present in his stomach due to poor gastric motility on admission. To summarize, if our patient indeed did have the “expected” amount of glucosuria, as judged from his hyperglycemia and polyuria, glucose balance could not be maintained in this 40-hour time interval without a large and unidentified input of glucose. Hence, we suspected that little glucose was excreted. Urine Osmolality During a Glucose-Induced Osmotic Diuresis To determine how high the urine osmolality is in hyperglycemic patients with an osmotic diuresis, the composition of the urine was examined in 29 patients presenting with a severe degree of hyperglycemia (Fig 1). To be certain that these values represented the osmotic diuresis per se and not possible variations due to underlying renal damage or a lack of antidiuretic hormone (ADH), studies also were performed in two groups of rats, normal rats infused with glucose and urea to match values in our patient, and in a group of BB diabetic rats in poor glycemic control. In all three groups, the urine osmolality was much lower than the corresponding values

SPIRA,

GOWRISHANKAR,

AND

HALPERIN

during water deprivation (1,200 mOsrr&g Hz0 in humans and 3,000 mOsm/kg Hz0 in rats; for review, see Knepper and Rector13). Moreover, glucose accounted for almost 60% of the osmoles excreted (Fig 1). The urine osmolality of 378 mOsm/kg HZ0 (Table 3) in our patient was of interest for two reasons. First, it made diabetes insipidus an extremely unlikely basis for the polyuria. Second, this value is considerably lower than the mean value observed in the 29 patients (553 t 30 mOsm/kg H,O; Fig 1). A lower than expected urine osmolality when ADH acts implies either a failure of water to approach osmotic equilibrium between the medullary collecting duct (MCD) and the corresponding interstitial fluid and/or a lower value for this interstitial fluid osmolality (discussed below). To raise the osmolality in the medullary interstitial fluid, “solute (NaCl) without water” must be transported from the water-impermeable nephron segment, the thick ascending limb of the loop of Henle (LOH) (the single effect).13’14 The osmolality of interstitial fluid in any given horizontal plane in the medulla will decrease when “solute-free water” is reabsorbed from the thin descending limb of the LOH and the MCD. We shall now consider quantitative aspects of these additions during an osmotic diuresis. Less reabsorption of NaCl in the thick ascending limb of the LOH in an osmotic diuresis. There are two points to consider. First, during an osmotic diuresis, the concentration of Na+ is lower in fluid delivered from the proximal convoluted tubule (PCT). In more detail, because PCT fluid is isosmolar to plasma,15 the higher its concentration of glucose, the lower its Na+ concentration (same total osmolality). In a robust osmotic diuresis in the dog, the concentration of Na+ in end-PCT fluid declined by one third and the absolute delivery of Na’ to the LOH was not changed.15 Second, there is a lower limit for the concentration of Na+ in the lumen of fluid exiting the thick ascending limb of the LOH.13,14 For illustrative purposes, we shall select 50 mmol/L for this minimum concentration of Nat. If 300 mmol of Na+ exited from the PCT in a given time, the volume would be 2 L in normal subjects (150 mmol/L NaCl) and 3 L during a glucoseinduced osmotic diuresis with an identical delivery of Na+ and 100 mmol/L glucose (100 mmol/

HYPERGLYCEMIA

AND

OSMOTIC

DIURESIS

L NaS, 100 mmol/L Cl-, and 100 mmol/L glucose). Now, if the LOH lowers its luminal Nat to 50 mmol/L in each case, NaC reabsorption would be 200 mmol in normals, but only 150 mmol during hyperglycemia. Additional “electrolyte-free water” to the medullary interstitium. During a robust osmotic diuresis, the volume delivered to the MCD is much larger than normal. The rationale is that when ADH acts, the osmolality of luminal and peritubular fluid should be equal because ADH causes aquaporin-2 water channels to be present in the luminal membrane of the cortical collecting ductI Hence, for every 300 mOsm excreted, 1 L should exit from the cortical collecting duct (plasma osmolality assumed to be 300 mOsmIkg H,O for ease of calculation). If normal subjects delivered 900 mOsm to the MCD, then 3 L of fluid would enter the MCD with an osmolality of 300 mOsrn/kg H,O. During an osmotic diuresis, if osmole delivery increased twofold to 1,800 mOsm/d, the volume delivered to the MCD would be 6 L. Now, if the osmolality in the urine doubled to 600 mOsm/kg Hz0 due to water abstraction as fluid traversed the MCD in both examples, 1.5 L of “solute-free water” would be reabsorbed in normals and 3 L with hyperglycemia. As discussed above, because a larger volume of fluid is delivered to the water-permeable thin descending limb of the LOH during an osmotic diuresis,15 more “solute-free water” will be reabsorbed from this nephron segment as fluid descends into the hyperosmolar medulla at any given osmolality here. Hence, both the lower addition of “water-free NaCl” from the thick ascending limb of the LOH and the higher reabsorption of “solute-free water” from the thin descending limb of the LOH and the MCD act in concert to cause the medullary interstitial and thereby urine osmolality to be lower during an osmotic diuresis. Excretion of a Smaller Than Expected Amount of Glucose With a mild increase in glycemia (up to 180 mg/dL), virtually all the filtered glucose is reabsorbed in the PCT via a cotransport system driven by active reabsorption of Na+. This transporter, which is a low-affinity, high-capacity Na+/glucase cotransporter, has a coupling ratio of 1 :l

633 Table

4. Expected Per

Rate of Glucose 24 Hours

Excretion

Glucose GFR (Ud)

Filtered

180 60 30

5,400 1,800 900

Reabsorbed

2,000 667 333

Excreted

3,400 1,133 567

NOTE. The assumptions are that the plasma glucose concentration is 30 mmol/L, or that either 2,000 mmol (normal), 667 mmol (expected for a GFR of 60 L/d), or 333 mmol (expected for a GFR 30 L/d) of glucose was reabsorbed. The patient excreted 363 mmol glucose when data were extrapolated over this 24-hour period.

and has been recently cloned and characterized.r6 When the quantity of glucose filtered exceeds 180 mg/dL, glucosuria becomes an important route for glucose disposal.5-7.9 Nevertheless, we found surprisingly little glucosuria (26 g over the first 18 hours; Table 3) despite our patient’s hyperglycemia, which means either a decreased rate of filtration (low GFR) and/or increased reabsorption of glucose by his PCT. If the mechanism was simply a low GFR, the GFR would have to be extremely low to explain the very small amount of glucosuria. With a normal rate of renal reabsorption of glucose, our patient should have excreted only 3.42 g glucose/ L of GFR (5.22 - 1.80 g/L). With only 26 g of glucosuria in the first 18 hours (Table 3), our patient’s GFR would have to be approximately 8 L/l8 hr or 11 L/24 hr, which is much less than 10% of normal. Two facts make this estimate surprisingly low. First, our patient’s plasma creatinine remained stable and close to 1.6 mg/dL (Table 2). Second, although it is difficult to obtain a reliable value for the GFR in our patient because there were no special tests performed (an infusion of inulin) and creatinine was not measured in his urine, an approximation of his GFR can be made from the data provided in Table 3. The amount of urea excreted over 37 hours was 558 mmol (260 + 298 mmol; Table 3). Given a mean plasma urea level of 17 mmol/L, his urea clearance was 33 L over 37 hours or 21 L/24 hr. If his fractional excretion of urea was 50%,r7 his GFR would be close to 40 L/24 hr. Nevertheless, a better estimate is a somewhat

834

lower value because the fractional excretion of urea is somewhat higher in polyuric states.17 The amount of glucose reabsorbed with the usual degree of glucose reabsorption per liter of GFR is provided in Table 4. We chose two values for the GFR, 60 L/d and 30 L/d, to bracket his expected value. For the lower GFR of 30 L/d, the rate of glucose excretion should have been 567 mmo1/24 hr, almost double the rate he excreted (363 mmo1/24 hr). Thus, our patient appears to have an enhanced rate of renal glucose reabsorption. We suggest two possible explanations for this: first, prior extracellular fluid volume contraction could have led to Na+-conserving measures, which might include enhanced proximal reabsorption of glucose”; and second, an enhanced delivery of NaC and Cl- to the macula densa could have led to a lower GFR due to a tubuloglomerular feedback mechanism.lg If true, PCT function (including glucose reabsorption) may be better preserved than was his GFR, leading to enhanced glucose reabsorption per liter GFR. In this case, the enhanced delivery of Na+ and Cl- to the macula densa could be an expected result from either the NaCl load, hypoxic injury, and/or the effect of gentamicin to block NaCl reabsorption in the thick ascending limb of the LOH (to be described below). Specific Causes for the Polyuria The polyuria seen in our patient was not simply a glucose-induced osmotic diuresis. The most abundant osmoles in his urine were Na+ and Cl-, which could be due to overaggressive administration of NaCl (total volume of saline infused, 8.5 L) and/or renal effects of the cationic aminoglycoside antibiotic, gentamicin. Cationic substances like neomycin bind to the ionized Ca2+ receptor on the basolateral aspect of the thick ascending limb cells of the LOH.20,21 Once activated, this receptor leads to inhibition of the luminal K+ channel via a G-protein-mediated signal system.‘l Failure of K+ to enter the lumen of the LOH leads to less reabsorption of Naf and Cl- via the Na+, Kf, 2 Cl--cotransporter and thereby to both a natriuresis and a diminished capacity to concentrate the urine. Thus, two factors contribute to a lower medullary interstitial osmolality: less “water-free NaCl” reabsorption in the LOH and more “solute-free water” reabsorption in the MCD for a given increase in os-

SPIRA,

GOWRISHANKAR,

AND

HALPERIN

molality because more solute (NaCl) is delivered there. The patient, however, received only a single dose of gentamicin. CONCLUSION

The sustained hyperglycemia reflects both metabolic events (not enough insulin was given) and a lower than expected excretion of glucose. The much lower concentration of glucose in the urine during an osmotic diuresis was the result of three factors: a diminished maximum urine osmolality, a larger excretion of nonglucose osmoles (NaCl), and possibly an enhanced proximal reabsorption of glucose. A potential role for the volume of saline administered and gentamitin were considered in the pathophysiology. REFERENCES 1. Taylor SI, Accili D, Imai Y: Perspectives in diabetes: Insulin resistance or insulin deficiency; Which is the primary cause of NIDDM? Diabetes 43:735-740, 1994 2. Porte DJ: Sympathetic regulation of insulin secretion. Arch Intern Med 123:252-260, 1969 3. Kamel KS: Challenging consults: Application of principles of physiology and biochemistry to the bedside. A case with marked hyperglycemia. Clin Invest Med 15:544-554, 1992 4. Halperin ML, Goguen JM, Scheich AM, Kamel KS: Clinical consequences of hyperglycemia and its correction, in Seldin DW, Giebisch G (eds): Clinical Disturbances of Water Metabolism. New York, NY, Raven, 1993, pp 249272 5. Halperin ML, Rolleston FS: Clinical Detective Stories: A Problem-Based Approach to Clinical Cases in Energy and Acid-Base Metabolism. London, UK, England, Portland Press, 1993 6. Deetjen P, Baeyer HV, Drexel H: Renal glucose transport, in Seldin DE, Gieveisch G (eds): The Kidney: Physiology and Pathophysiology (ed 3). New York, NY, Raven, 1992, pp 2873-2888 7. Baeyer HV, Haeberle DA, Van Liew JB, Hare D: Glomerular tubular balance of renal D-glucose transport during hyperglycemia. Pflugers Arch 384:39-47, 1980 8. Halperin ML, Goldstein MB, Richardson RMAR, Robson WLM: Quantitative aspects of hyperglycemia in the diabetic: A theoretical approach. Clin Invest Med 2:127-130, 1980 9. Seldin DW, Tarail R: The metabolism of glucose and electrolytes in diabetic acidosis, J Clin Invest 29:552-565, 1950 10. Roscoe JM, Halperin ML, Rolleston FS, Goldstein MB: Hyperglycemia-induced hyponatremia: Metabolic considerations in calculation of serum sodium depression. Can Med Assoc J 112:452-453, 1975 11. Jungas RL, Halperin ML, Brosnan JT: Lessons learnt from a quantitative analysis of amino acid oxidation and related gluconeogenesis in man. Physiol Rev 72:419-448, 1992

HYPERGLYCEMIA

AND

OSMOTIC

DIURESIS

12. Randle PJ: Fuel selection in animals. Biochem Sot Trans 14:799-806, 1986 13. Knepper MA, Rector FCJ: Urinary concentration and dilution, in Brenner BM (ed): Brenner and Rector’s, The Kidney. Philadelphia, PA, Saunders, 1996, pp 532-570 14. Jamison RL, Roy DR, Layton HE: Countercurrent mechanisms and its regulation, in Seldin DW, Giebisch GH (eds): Clinical Disturbances of Water Metabolism. New York, NY, Raven, 1993, pp 119-156 15. Seely JF, Dirks JF: Micropuncture study of hypertonic mannitol diuresis in the proximal and distal tubule of the dog kidney. J Clin Invest 48:2330-2340, 1969 16. Kanai Y, Lee W-S, You G, Brown D, Hediger MA: The human kidney low affinity Na+/glucose cotransporter SGLT2: Delineation of the major renal reabsorptive mechanism for D-glucose. .I Clin Invest 93:397-404, 1994

835 17. Bankir L: Urea and the kidney, in Brenner BM (ed): Brenner and Rector’s, The Kidney. Philadelphia, PA, Saunders, 1996, pp 571-606 18. Kurtzman NA, White MG, Rogers RW, Flynn JJ III: Relationship of sodium reabsorption and glomerular filtration rate to renal glucose reabsorption. J Clin Invest 51:127-133, 1972 19. Seney FD Jr, Persson EG, Wright FS: Modification of tubuloglomerular feedback signal by dietary protein. Am J Physiol 252:F83-F90, 1987 20. Hebert SC: Extracellular calcium-sensing receptor: Implications for calcium and magnesium handling in the kidney. Kidney Int 50:2129-2139, 1996 21. Brown EM: Extracellular Ca’+-sensing, regulation of parathyroid cell function, and role of Ca” and other ions as extracellular (first) messengers. Physiol Rev 71:371-411, 1991