hyperglycemia during childhood diarrhea

Hyperglycemia during childhood diarrhea Anne Ronan, MRCP, FRACP, Abul K. Azad, MB,BS, Omar Rahman, MD, Rodney E. Phillips, MD, FRACP, and Michael L. Bennish, MD

From the Royal Children's Hospital, Melbourne, Victoria, Australia; the International Centre for Diarrhoeal Disease Research, Dhaka, Bangladesh; the Rand Corporation, Santa Mon- ica, California; the Nuffield Department of Clinical Medicine, John Radcliffe Hospital, Ox- ford, United Kingdom; and the Departments of Pediatrics and Medicine, New England Medical Center, Tufts University School of Medicine, Boston, Massachuselts

Objective: To determine the cause of hyperglycemia in childhood diarrhea. Methods: During an 8-month period, patients admitted to a diarrhea treatment center in Bangladesh had their blood glucose concentrations determined. Sixteen patients aged 2 to 10 years with hyperglycemia (blood glucose concentration >10.0 mmol/L) and 20 patients in the same age group with a normal blood glucose concentration (3.3 to 9.0 mmol/L) had blood samples obtained on admission and 4 and 24 hours later for determination of glucoregulatory hormones and gluconeogenic substrates. Results: Prevalence of hyperglycemia among patients aged 2 to 10 years was 9.4%. Compared with the normoglycemic patients, hyperglycemic patients more offen had severe dehydration (100% versus 10%, p <0.001), infection with Vibrio cholerae 01 or toxigenic Escherichia coli (94% vs 25%, p <0.001), and had similar duration of fasting (16 vs 14 hours, p = 0.677). Concentrations of epinephrine (7.15 vs 2.00 pmol/L), norepinephrine (10.35 vs 3.50 IJmol/L), cortisol (1.38 vs 0.82 pmol/L), glucagon (36 vs 14 pmol/L), and C-peptide (1.22 vs 0.35 nmol/L) were all significantly (p -<0.014) higher in patients with hyperglycemia than in normoglycemic patients. Conclusions: The development of hyperglycemia in diarrhea is caused by a stress response to hypovolemia. (J Pediatr 1997; 130:45-5 I)

Supported by the International Centre for Diarrhoeal Disease Re- search, Bangladesh (ICDDR,B). The ICDDR, B is currenfly supported by the Governments of Australia, Bangladesh, Belgium, Canada, Denmark, France, Japan, The Netherlands, Norway, Swe- den, Switzefland, the United Kingdom, and the United States; the United Nations Development Program, the United Nations Chil- dren's Fund, the World Health Organization; and the Ford and Säsakawa Foundations.

Submitted for publication Mar. 26, 1996; accepted July 22, 1996.

Reprint requests: Anne Ronan, MRCP, FRACP, Department of Emergency Medicine, John Hunter Hospital, Locked Bag 1, Hunter Region Mail Centre, Newcastle, New South Wales 2310, Austra- ]Ja.

Copyright © 1997 by Mosby-Year Book, Inc. 0022-3476/97/$5.00 + 0 9/21/76845

Hyperglycemia during childhood diarrhea has been reported from both developed I-6 and developing countries. 7' 8 Despite

these and other smdies on glucose metabolism and storage in children with diarrhea, 9, 10 the canse of the hyperglycemia

ICDDR,B International Centre for Diarrhoeal Disease Research, Bangladesh

remains uncertain. This smdy explores the physiologic basis of hyperglycemia duri'ng diarrhea in children in Bangladesh.

M E T H O D S

Recruitment of patients. The study was condu¢ted at the diarrhea treatment center of the International Centre for Di- arrhoeal Disease Research, Bangladesh, in Dhaka. This la-

45

4 6 Ronan et al. The Journal of Pediatrics Januaß 1997

Tab le I. Admission characteristics of hyperglycemie and normoglycemic children with diarrhea

Hyperglycemic Normoglycemic (n = 16) (n = 20) p

Age (mo) 66 (26-108) 45 (24-120) 0.312 Male: n (%) 10 (62.5%) 16 (80.0%) 0.285 Diarrhea duration (hr) 9 (4-48) 84 (3-288) <0.001 History of vomiting: n (%) 14 (87.5%) 12 (60.0%) 0.133 Hours since last feeding 16 (1-22) 14 (2-96) 0.677 Weight for age (mean, SD)* 66.9%, 11.1% 58.5%, 10.1% 0.026 Severe malnutrition-~: n (%) 5 (31.3%) 12 (60.0%) 0.055 Severe dehydration: n (%) 16 (100.0%) 2 (10.0%) <0.001 Liver palpable: n (%) 4 (25.0%) 11 (55.0%) 0.141 Stool watery: n (%) 16 (100.0%) 9 (45.0%) <0.001 Stool bl0ody/mucoid: n (%) 0 11 (55.0%) <0.001 Enteric pathogens isolated

Vibrio cholerae 14 (87.5%) 5 (25.0%) <0.001 Enterotoxigenic 1 (6.3%) 0 0.444 Escherichia coli Shigella (alone or mixed) 1 (6.3%) 6 (30.0%) 0.084 Other pathogens 0 4 (20.0%):) 0.113 No pathogen isolated 1 (6.3%) 5 (25.0%) 0.196

Values expressed as median (range) unless otherwise stated. *Expressed as a percentage of the National Center for Health Statistics (NCHS) median. ?Weight for age less than 60% of the NCHS median or presence of kwashiorkor. $Campylobacter jejuni, Entamoeba histolytica, Vibrio non-01 serogroup.

cility provides treatment free of charge to 75,000 patients per year, approximately 70% of whom are younger than 10 years of age. Most are cared for as outpatients, but 7% of the patients, mosüy children, are admitted for inpatient care.

During a study to investigate the physiologic basis of dis- turbed glucose metabolism in childhood diarrhea, 9 all inpa- tients had their blood glucose concentration determined by

using a reflectometer (G!ucochek SC, Medistron, Horsham, England) and glucose oxidase-peroxidase test strips (Haemo- Glukotest 20-800R, Boehringer-Mannheim, Mannheim,

Germany). Hyperglycemia was defined as a blood glucose concentration of more than 10.0 mmol/L (180 mg/dl), hypoglycemia as a blood glucose level less than 2.2 mmol/L (40 mg/dl), and normoglycemia as a blood glucose level of between 3.3 and 9.0 mmol/L (60 to 162 mg/dl). ~1

The first 500 consecutive inpätients of all ages admitted during daytime hours over a 4-month period were screened for hyperglycemia or hypoglycemia. During the ensuing 4-month period, only patients with altered mental status were screened for hypoglycemia or hyperglycemia. Only children aged 2 to 10 years were included in the smdy to determine the cause of hyperglycemia, because ethical con- cerns about the volume of blood required precluded enrolling patients younger than 2 years of age. Normoglycemic patients were selected by enrolling the first age-matched normoglycemic patient with diarrhea to be admitted to the inpatient unit after detection of hypoglycemia in a patient. Twenty normoglycemic patients aged 2 to 10 years were in-

cluded in this study for comparison with hyperglycemic pa-

tients. Study protocol. When hyperglycemia was diagnosed or

a suitable normoglycemic contro! subject was found, in- formed consent was obtained from the patent or guardian. It" consent was given, blood was obtained for the determination of the plasma concentration of glucose, C-peptide, [3-hydroxybutyrate, lactate, alanine, cortisol, epinephrine, norepinephrine, growth hormone, pancreatic glucagon, electrolytes, and creatinine, and for the determination of the serum

albumin, bilirubin, and aspartate transaminase concentrations. A history was taken and a full physical examination performed by a study physician. Severely dehydrated patients were then rapidly rehydrated with an intravenous glucose-free polyelectrolyte solution, and patients with mild or moderate dehydration received an oral rehydration formula containing 111 mmol/L glucose. All patients received food as soon as they were able to eat. Blood sampling was repeated at 4 hours and 24 hours if the patient remained in the hospital at that time. Many of these children responded quickly to treatment, and no child was kept in hospital for longer than was medically indicated.

Nutritional status was assessed by using percentage weight-for-age, defined as weight after rehydration expressed as a percentage of the National Center for Health Statistics median weight-for-äge. 12 Weight-for-age of less than 60% of this standard or the presence of nutritional edema defined severe malnutrition. Dehydration level was

The Joumal of Pediatrics Ronan et al. 4 7 VoIume 130, Number 1

l ' o b l e Il. Admission biochemical and hormonal findings in hyperglycemic and normoglycemic children with diarrhea

Hyperglycemic Normoglycemic (n = 16) (n = 20) Reference range* p

Glucose (mmol/L [ m g / d l ] ) 14.83,10.20-24.72 5.32,3.20-9.00 3.30-5.50 - - [267.08,183.70-445.21] [95.81,57.63-162.09] [60.00-100.00]

C peptide (lamol/L) 1.22,0.24-2.50 0.35,0.00-1.47 0.25-0.50 0.001 Cortisol (pmol/L) 1.38,0.90-1.55 0.82,0.25-1.38 0.14-0.64 <0.001 Growth hormone (pg/L) 26.50,2,00-61.00 11.50,2.00-61.00 <5.00 0.166 Glucagon (pmol/L) 36.00,17.00-98.00 14.00,2.00-64.00 4.00-8.00 0.014 Epinephrine (pmol/L) 7.15,1.50-44.00 2.00,0.10-13.00 <0.38 0.012 Norepinephrine (pmol/L) 10.35,1.10-44.00 3.50,0.10-16.20 0.59-2.36 0.001 B-Hydroxybutyrate, (mmol /L) 0.30,0.06-2.06 0.66,0.06-3.50 <0.47 0.267 Lactate (lamaol]L) 3.90,0.80-5.90 2.20,1.00-11.30 0.90-1.70 0.048 Alanine (mmol/L) 0.61,0.15-0.93 0.20,0.09-0.77 0.24-0.40 0.001 Bilirubin (pmol/L [mg/dl]) 3.00,1.00-38.00 2.00,1.00-6.00 3.40-17.10 0.289

[0.18,0.06-2.22] [0.11,0.06-0.35] [0.20-1.00] Aspartate transaminase ( U / L ) 65.50,27.00-106.00 72.00,57.00-127.00 8.00-45.00 0.062 Albumin (g/L) 38.00,25.00-56.00 28.00,14.00-43.00 38.00-50.00 0.005 Blood leukocytes (xl09/L) 23(5-49) 23(15-43) 5-15 0.834 Hematocrit 0.38,0.35-0.54 0.35,0.25-0.44 0.30-0.45 0.015 Carbon dioxide (mmol/L) 12(6-21) 14(8-18) 18-24 0.361 Sodium (mmol/L) 134(124-141) 130(114-141) 135-146 0.311 Potassium (mmol/L) 4.5(3.2-5.6) 3.8(1.8-8.2) 3.5-5.8) 0.054

Values are expressed as median, range in SI units with equivalent in brackets where appropriate. *Refer to reference 9 for sources of reference values.

defined according to the World Health Organization clinical definitions.13

Labora tory methods. Plasma was separated within 15 minutes of blood sampling and then stored at $-70 ° C until it was transported, packed in dry ice, to collaborating labo- ratories in England. Hemolysis in some samples precluded the determination of insulin concentration; therefore plasma C-peptide concentrations, which are not affected by hemolysis, were determined instead in all samples. All biochemical measurements were done by using previously described methods, 9 All measurements were done in England, except for determination of concentrations of electrolytes and cre-

atinine, which were measured at the ICDDR,B. A single stool sample was obtained from each patient for the identi- fication of Shigella, Salmonella, Campylobacter jejuni, Vibrio cholerae, or Escherichia coli producing heat-stable or heat-labile toxin, and rotavirus.

Statisties. Data were entered into a personal computer by using Statpac Gold version 3.0 (Walonick Associates, Min- neapolis, Minn.) and analyzed by using Statistical Package for Social Sciences version 6.0 for Windows (SPSS Inc., Chicago, Ill.). The significance of differences in proportäons was tested by using the chi-square test with a continuity cor- rection, or the Fisher Exact Test if the predicted size of any cell was five or less. Differences between independent groups of normally distfibuted variables were tested by using the Student t test or the Wflcoxon Rank Sum test if the observations were not normally distributed. Pearson corre-

lation coefficients were calculated if the variables examined were normally distributed, or Spearman tank correlation was used if the variables, as in most cases, were not normally distributed. Least squares multiple regression was performed on predetermined variables. Tests of significance were two sided. Figures were generated by using Stata Statistical Software release 4.0 (Stata Corp., College Station, Tex.).

This study was approved by the ethical review committee of the ICDDR,B.

R E S U L T S

In the initial screening of 500 patients with diarrhea, 430 of whom were younger than 10 years, 30 patients (6.0%) were found to have blood glucose levels of more than 10.0 mmol/L. Orte hundred sixty patients screened were between the ages of 2 and 10 years, and of these, 15 (9.4%) had hyperglycemia. During the total 8-month period of the study, 16 patients with hyperglycemia between the ages of 2 and 10 years consented to participate, and 20 age-matched normoglycemic patients were obtained, as described, for comparison. Ninety-one children with blood glucose levels less than 2.2 mmol/L were identified, and the detailed study of 46 of these patients was previously described. 9

Clinieal findings. All 16 patients with hyperglycemia had watery diarrhea and severe dehydration on admission, compared with only 2 (10.0%) severely dehydrated normoglycemic patients (p <0.001), the majority of whom had mucoid or bloody mucoid diarrhea (Table I). All severely dehydrated

4 8 Ronan et al. The Journal of Pediatrics January 1997

!!l O 10

5

O -

Hyperglycemic

o

Normoglycemic

t

2.5-

0 4 24 0 4 24

Hours after admission

Fig. 1. Box-and-whisker plots* of plasma glucose and C-peptide concentrations in the first 24 hours after admission for hyperglycemic and normoglycemic groups.

patients had feeble or absent radial pulses and extreme leth- argy. Duration of diarrhea was significantly shorter in the hyperglycemic children (median, 9 vs 84 hours; p <0.001) than in the normoglycemic children. No significant difference was seen between the two groups in duration of vomiting or time from last feeding. None of the 36 patients had received oral rehydration therapy before admission.

Nutritional status was better in the hyperglycemic than in the normoglycemic group. Only five pafients with hyperglycemia (31.3%) had evidence of severe malnutrition (two with kwashiorkor), and median weight-for-age was 66.9% in this group. Among the 20 normoglycemic patients, 12

*The boxes in Figs. 1, 2, and 3 depict the 25th, 50th, and 75th percenfiles for each group, with the whiskers extending to the "adjacent values," defined as three halves of the interquartile range. Values beyond this range are depicted as open circles.

0

0

ùu

. m ~«

.=. s.., 0

Z;

45

40

30

2 0 ¸

t 0 -

O-

Hyperglycemic

o

o

Normoglycemie

o o

4 5 -

~ßt 40- 0

O 30 --

¢t 2O . -

o~ t0

o

o

i - - -T- - I

o o

, ~ ~ ~ 0 4 24 0 4 24


Fig. 2. Box-and-whisker plots of plasma catecholamine concentrations in the first 24 hours after admission for hyperglycemic and normoglycemic groups.

patients (60.0%) had severe malnutrition (p = 0.055 compared with hyperglycemic patients), of whom two had kwashiorkor, and median weight-for-age was 58.5% (p = 0.026, compared with patients with hyperglycemia).

Vibrio cholerae was isolated from the stools of patients with hyperglycemia (14 [87.5%] of 16) more offen than from normoglycemic patients (5 [25.0%] of 20). Enterotoxigenic

E. coli was isolated from the stool of one other patient with hyperglycemia. All patients were examjned for evidence of concomitant infections. SaImonella typhi was isolated from a culture of a blood specimen from 1 of 18 normoglycemic patients tested, but cultures of blood from four patients with hyperglycemia were sterile.

Glueose metabolism and hormonal findings. The median plasma glucose level in the hyperglycemic group was 14.8 mmol/L, with a range of 10.2 to 24.7 mmol/L (267.1 mg/dl; range, 183.7 to 445.2 mg/dl). On repeated testing 4

The Journal of Pediatrics Ronan et al. 4 9 Volume 130, Number 1

hours after admission, no patient had hyperglycemia, and all patients continued to be euglycemic during the 24 hours of the study (Fig. 1).

Marked differences in glucoregulatory horrnones and gluconeogenic substrates between the hyperglycemic and normoglycemic groups were found on admission (Table II). Median plasma C-peptide concentration was significantly (p = 0.001) higher in the hyperglycemic than the normoglycemic patients. The median concentrations of counterregu- latory hormones cortisol, glucagon, epinephrine, and norepinephrine were also significantly (p <0.015) higher on admission in hyperglycemic than in normoglycemic patients. Concentrations of all these hormones remained elevated up to 24 hours in many cases (Figs. 2 and 3). Glu- cagon in pm~icular remained elevated at 24 hours in all eight hyperglycemic and four normoglycemic patients tested.

Glucose precursors alanine and lactate were also significantly (/9 <:0.05) higher on admission in patients with hyperglycemia than in normoglycemic paUents. Ketones, measured as [3-hydroxybutyrate, were lower in the hyperglycemic than in the normoglycemic group, but this difference was not significant.

Admission concentrations of the three glucoregulatory hormones enpected best to predict plasma glucose in the hyperglycemie patients, norepinephrine, glucagon, and cortisol, I4 were fitted to a multiple regression model (Table III). In the final model, plasma norepinephrine and glucagon were independent predictors of glucose concentration (adjusted R 2 = 0.68, p = <0.001).

Biochemical and hematologic findings. Mean blood leukocyte counts (measured in 9 of the 16 hyperglycemic and 15 of the 20 nolxnoglycemic patients) were elevated in both groups. There was no significant difference in serum total CO2, sodium, potassium, chloride, creatinine, aspartate transaminase, or bilirubin concentrations between the two groups on admission. No significant (p <0.05) correlation was found between plasma glucose and serum sodium or between plasma glucose and acidosis as measured by total serum CO2 on admission. No hyperglycemic child was hypernatremic (serum sodium concentration >146 mmol/L), and the lowest potassium eoncentrafion in this group was 3.2 mmol/L.

Effect of malnutrition. Five (31.3%) of the hyperglycemic patients had severe malnutrition, including 2 with kwashiorkor. Except for alanine (median concentration, 0.34 mmolA~ in severely malnourished patients versus 0.64 mmol/L in the other patients p = 0.010), clinical and biochemical findings in these 5 patients were similar to those in the 11 hyperglycemic patients who were not severely malnourished.

Outcome. One (6.3%) of the hyperglycemic patients and 3 (15.0%) of the normoglycemic patients died in hospital

0

E 0

O

Hyperglycemic

.~ [-

0

Normoglycemic

100

BO

O E~ 60

t= o I~o 40

=

~ 20

o

o

4 24 0 4 24


Fig. 3. Box-and-whisker plots of plasma cortisol and glucagon concentrations in the first 24 hours after admission for hyperglycemic and normoglycemic groups.

(p = 0.767), 15 in each group were discharged improved, and the other 2 normoglycemic patients left hospital against medical advice in spite of worsening clinical condition. No autopsies were permitted by families involved, making it impossible to ascertain exact canses of death, but all 4 chip dreh who died were severely malnourished.

D I S C U S S I O N

Plasma glucose concentrations are often not determined in patients with diarrhea, and thus hyperglycemia (and hypoglycemia) is orten not detected. The actual prevalence of hyperglycemia dufing diarrhea has remaiued uncertain. The incidence of hyperglycemia in this study was 6.0% among patients of all ages and 9.4% in children aged 2 to 10 years. In previous reports, the prevalence of hyperglycemia has ranged from 14% to 20%. 2, 3, 7 Variations in the reported prevalence of hyperglycemia dufing diarrhea are probably attfibutable to a number of factors, including

5 0 Ronan et al. The Journal of Pediatrics January 1997

T a b l e III. Glucoregulatory hormones independently asso¢iated with plasma glucose concentration in patients with hyperglycemia in a multiple regression analysis

95% Confidence Coefficient Standard error p interval

Norepinephrine (gmol/L) 0.432 0.090 0.001 0.234, 0.631 Glucagon (pmol/L) 0.107 0.029 0.004 0.042, 0.172 Constant 6.608 1.899 0.005 2.428, 10.788

Number of observations = 14. Probability >F = 0.0008. Adjusted R 2 = 0.6746. Root Mean Square Error = 2.1032.

the way that patients were selected for screening, and the prevalence of factors, such as severe dehy&ation, that have been associated with the development of hyperglyce-

mia. Hyperglycemia during diarrhea has been attributed to a

number of different causes. These include hypernatremia,2,4,5,15 excessive ingestion of sucrose, 1 acidosis, 3 hy-

pokalemia or viremia, 8 and a stress response caused by ei- ther infection 8 or marked hypovolemia. 6 The findings of this

study are consistent with a stress response to marked hypovolemia as the cause of the hyperglycemia. Compared with normoglycemic patients, patients with hyperglycemia had significantly increased concentrations of the hormones epinephrine, norepinephrine, cortisol, growth hormone, and glucagon, all of which have an antiinsulin effect. 16 All the hyperglycemic patients in this study were severely dehy-

drated. A decrease in intravascular volume is one of the most potent stimuli for the release of catecholamines and corti- so1.17 Severe dehydration in these patients probably resulted

in a sustained release of catecholamines, cortisol, and gluca-

gon, which overwhelmed the normal glucoregulatory mech- anism (the release of insulin) and resulted in hyperglycemia. In experimental studies in human beings, combination (as opposed to isolated) infusions of catecholamines, glucagon, and cortisol, with concentrations equivalent to those seen in patients in this study, are necessary to produce hyperglycemia. 14 It is unlikely that infection per se contributed to this stress response, because most of the children with hyperglycemia were infected with V. cholerae, an organism that is confined to the intestinal lumen and that does not itself elicit a marked systemic inflammatory or stress response. There was no clinical evidence of viral infection in any of these patients.

Unlike previous researchers, we found no evidence that hypernatremia is a cause of hyperglycemia in these patients. Part of the reason may be that hypernatremia is relatively

18 uncommon among patients with diarrhea in Bangladesh, or that hypernatremia itself is not responsible for hyperglycemia. Of 68 children previously described in the literature as having hyperglycemJa during diarrhea, all were dehydrated, 18 including all 28 patients with hypernatremia. 2' 4, 5, 15 In

all studies reported, including this one, restoration of intravascular volume resulted in return of glucose concentrations to normal, even if the serum sodium concentration remained elevated. Potassium concentrations were the same in hyperglycemic and normoglycemic patients in this study, sug-

gesting that hypokalemia was not a cause of the hyperglycemia.

Unlike hypoglycemia, which was associated with a marked increase in the risk of death, 9 hyperglycemia was not

associated with a poor outcome. There is no evidence to suggest that hyperglycemia contributed to dehydration in

these patients, because in severely dehydrated patients the glomerular filtration rate is very low and osmotic diuresis usually does not occur) 9 However, potential danger exists

where severe, dehydrating dimThea is not as prevalent as in Bangladesh, and where the patients with diarrhea and hyperglycemia may be mistakenly given insulin. 1

R E F E R E N C E S

1. Burman D, Glaspole D. Alimentary hyperglycaemia simulating diabetic ketosis [letter]. Lancet 1964;2:208-9.

2. Levin S, Geiler-Bernstein C. Alimentary hyperglycaemia simulating diabetic ketosis [letter]. Lancet 1964;2:595.

3. Spirer Z, Bogair N. Blood sugars in infants with diarrhoea. Lancet 1970;2:1311-2.

4. Stevenson RE, Bowyer FP. Hyperglycaemia with hyperosmo- lal dehydration in nondiabetic infants. J Pediatr 1970;77:818- 23.

5. Mandell F, Fellers FX. Hyperglycemia in hypernatremic dehydration. Clin Pediatr 1974;13:367-9.

6. Perkins KC. Hyperglycaemia without ketosis or hyper- natraemia. Am J Dis Child 1974;128:885.

7. Hirschorn H, Lindenbaum J, Greenhough WB III, Alam SM. Hypoglycaemia in children with acute diarrhoea. Lancet 1966; 2:128-33.

8. Rabinowitz R, Joffe BI, Abkiewicz C, Shires R, Greef MC, Seftel HC. Hyperglycaemia in infantile gastroenteritis. Arch Dis Child 1984;59:771-5.

9. Bennish ML, Azad AK, Rahman O, Phillips RE. Hypoglyce- mia during diarrhea in childhood: prevalence, pathophysiology, and outcome. N Engl J Med 1990;322:1357-63.

10. Butler T, Amold M, Islam M. Depletion of hepatic glycogen in the hypoglycaemia of fatal childhood diarrhoeal illnesses. Trans R Soc Trop Med Hyg 1989;83:839-43.

11. Behrman R, Kliegman R, Nelson W, Vaughan V, editors.

The Joumal of Pediatrics Ronan et al. 5 1 Volume 130, Number 1

Textbook of pediatrics. 14th ed. Philadelphia: WB Saunders, 1992.

12. Anonymous. Measuring change in nutritional status. Geneva: World Health Organization, 1983.

13. Anonymous. The treatment and prevention of acute diarrhoea: practical guidelines. Geneva: World Health Organization, 1985.

14. Shamoon H, Hendler R, Sherwin RS. Synergistic interactions among antiinsulin hormones in the pathogenesis of stress hyperglycemia in human s. J Clin Endocrinol Metab 1981 ;52:1235- 41.

15. Roberts W. Alimentary hyperglycaemia simulating diabetic ketosis [letter]. Lancet 1964;2:754.

16. Rolih CA, Ober KP. The endocfine response to critical illness. Med Clin North Am 1995;79:211-24.

17. Woolf PD. Endocrinology of shock. Ann Emerg Med 1986; 15:1401-5.

18. Samadi AR, Wahed MA, Islam MR, Ahmed SM. Conse- quences of hyponatraemia and hypematraemia in children with acute diarrhoea in Bangladesh. Br Med J Clin Res Ed 1983; 286:671-3.

19. Rahman O, Bennish ML, Alam AN, Salam MA. Rapid intravenous rehydration by means of a single polyelecuolyte solution with or without dextrose. J Pediatr 1988;113:654- 60.

FELLOWSHIPS

Fellowships available in pediatric subspecialties and those for general academic pediatric training

are listed once a year, in J anua ry , in The Journal ofPediatrics. Each June, forms for listing fel-

lowships available for the academic year beginning 18 months after publication are sent to the

Chairman of the Department of Pediatrics at major hospitals in the United States and Canada. In

addition, a copy of the application form appears in the July, August, and September issues of

The Journal (please use the current form). Should you desire to list fellowships, a separate appli-

cation must be made each year for each position. All applications taust be returned to Mosby-Year

Book, Inc., by Oc tober 15 preceding the listing year to ensure publication. Additional forms will

be supplied on request from the Journal Editing Depamnent, Mosby-Year Book, Inc., 11830 West-

line Industrial Drive, St. Louis, MO 63146-3318/800-325-4177, ext. 4317, or 314,453-4317.

hyperglycemia during childhood diarrhea

Documents