Hypoketonemia and age-related fasting hypoglycemia in growth hormone deficiency

Hypoketonemia

and Age-Related Fasting Hypoglycemia Growth Hormone Deficiency

Joseph I. Wolfsdorf,

Abdollah

Sadeghi-Nejad,

in

and Boris Senior

Body fuels were measured in 45 normal children and 17 growth hormone-deficient patients after 24 hours of fasting. After three months of therapy with human Growth Hormone (hGH1 16 of the patients were restudied. In all groups. @-hydroxybutyrate (BOHB) concentrations correlated inversely with age and with glucose concentrations. When adjusted for these factors, the concentrations of BOHB were significantly lower in the growth hormone-deficient patients than in the control children, before (P < 0.01) as well as after therapy (P < 0.01). Only the five youngest patients became hypoglycemic. During fasting, ketones, which serve as an alternative fuel for the brain, spare glucose. Thus, a shortage of ketones would compromise the ability of the patient to conserve glucose and predispose the pateint to fasting hypoglycemia. Accordingly, we propose that hypoketonemia is a critical factor in the genesis of fasting hypoglycemia in growth hormone deficiency.

W

HATEVER THE CAUSE, fasting hypoglycemia can only result from decreased production of glucose, increased utilization of glucose, or some combination of the two. Several reports implicate growth hormone deficiency as a cause of fasting hypoglycemia,’ -a but how a lack of growth hormone affects either the production or the utilization of glucose remains unclear. Although fasting concentrations of glucose are reduced in adults with growth hormone deficiency’ concentrations are lower still in children with growth hormone deficiency and particularly so in younger chlldren.h,7 Why the adverse effects of a growth hormone deficiency on fasting glucose levels should be most strikingly expressed in younger children is also unclear. Ketones increase more rapidly in fasting children than in adults’~” and by serving as an alternative fuel, they spare glucose.” We therefore questioned whether a decreased availability of ketones might cause fasting growth hormone-deficient children to become hypoglycemic. We measureed body fuels in 45 normal children and 17 growth hormone-deficient patients under fasting conditions; 16 of the growth hormone deficient patients were restudied after three months of therapy with human Growth Hormone (hGH). We found that ketones were indeed less available in children who lacked growth hormone. MATERIALS

AND

METHODS

Subjects We studied 17 patients consecutively diagnosed as having defictency of growth hormone. The clinical features of the patients are shown in Table 1. Each had responded inadequately to two standard tests of growth hormone secretion, infusion of arginine, and insulininduced hypoglycemia. One patient (patient 4 in Table l), aged 4 years 7 months, became hypoglycemic after 18 hours during the pretreatment fast but did not undergo a second post-therapy fast. Accordingly, her data have not been included in the paired analyses. SIX of these patients also lacked ACTH and TSH. They received replacement therapy with cortisone acetate and L-thyroxine in standard doses beginning at least three months before the initial fast and throughout the duration of treatment with hGH. Metabolism,

Vol. 32, No. 5 IMay). 1983

The control group consisted of 45 children; 2 I boys and 24 girls, aged 10 months to 16.7 years, who had been referred to us for evaluation of suspected hypoglycemia. Of these, free fatty acids were measured in 21 and @-hydroxybutyrate (BOHB) was measured in 23. No endocrine or metabolic abnormalities were found in any of these children.

Studies All subjects but one fasted for 24 hours. She (patient 4 in Table 1) became drowsy at I8 hours and the fast was terminated. The fast began after lunch; water was allowed ad libitum and each subject was under close observation throughout. At the conclusion of the fast, blood was drawn for assay of glucose,‘* lactate,‘j pyruvate,rJ free fatty acids,r5 BOHB,16 acetoacetate,16 glycerol,” alanine,” and insulin.‘9 Blood was collected in a tube containing fluoride for assay of glucose and in a tared, iced tube containing an equal volume of 7% perchloric acid for assay of acetoacetate. Acetoacetate was measured in the neutralized supernatant. The sample for assay of serum @-hydroxybutyrate, free fatty acids, and insulin was promptly transported on ice to the laboratory. To prevent spurious elevation of FFA resulting from in vitro hydrolysis of triglyceride, FFA was extracted within 1 hour of obtaining the blood sample. The growth hormone-deficient patients were treated with intramuscular injections of hGH three times weekly. The total weekly dose was 5.4 i- 1.9 (mean f SD) units/m*. After three months of treatment the fasting study was repeated. The hGH was administered up to and on the day of the fast. The six patients who required replacement therapy with cortisone acetate and L-thyroxine continued these medications during the fast, The study was approved by the Human Investigation Review Committee of the New England Medical Center and informed consent was obtained in all cases,

Statistical Methods We used the paired Student r-test to compare the mean concentrations of the metabolites and of insulin after the 24-hour fast in the growth hormone-deficient patients before and after treatment with hGH.

From the Department of Pediatrics, Tujis University School of Medicine, and the Pediatric Endocrine-Metabolic Service. New England Medical Center Hospital (Boston Floating Hospital for Infants and Children), Boston, MA. Received for publication November 30, 1982. Address reprint requests to: Boris Senior, M.D., New England Medical Center. 171 Harrison Avenue, Boston, MA 021 II. Q I983 by Grune & Stratton, Inc. 0026-0495/83/3205-0007$01.00/0 457

WOLFSDORF, SADEGHI-NEJAD, AND SENIOR

458

Table 1. Clinical Data, Peak Concentrations of Growth Hormone in Response to Arginine-Insulin Provocation Tests, Pretreatment Growth Rates, and Growth Response to hGH Therapy of 17 Growth Hormone-Deficient Patients Growth Rate

Patient

Sex M

1

Etiology

Other

Pre-

Peak GH

Hormone

treatment

(nglml)

Deficiencies

(cm/v)

4.4

Idiopathic

ACTH

DOS.? Age (years) BOlle

Chronologic

4.3

Height

1.3

2.9

Weight 1.1

1.1

of hGH

Growth Rate

per Week

on hGH

KJlm2)

(cm/v)

3.1

10.1

TSH IH-FSH 2

M

Idiopathic

2.3

None

6.1

3.0

1.7

1.2

1.0

6.3

9.3

3

M

Idiopathic

2.4

N0lle

4.3

3.4

1.3

1.4

0.6

3.4

6.5

4

F

Septo-optlc

2.9

ACTH

2.4

4.7

1.5

1.3

2.1

2.7

8.1

0.5

5.2

4.0

2.5

2.0

10.7

11.1

Dysplasia

TSH ?LH-FSH

5

M

1.4

SeptlYxmc Dysplasia

ACTH TSH ?LH-FSH

6

M

ldiopathac

1.2

N0lle

2.6

11.3

7.0

7.0

6.5

7.0

8.2

7

F

Craniopharyngioma

0.2

None

2.1

12.2

8.5

8.5

9.5

5.8

9.0

B

M

lrradiataon

6.1

None

4.9

13.3

9.0

8.0

8.0

6.5

8.1

9

F

Idiopathic

2.7

N0lle

3.4

13.3

10.0

12.0

17.0

4.0

5.5

10

F

Irradiation

3.0

NOlll?

2.8

14.5

14.0

10.2

11.3

4.8

0

11

F

Head Injury

0.4

ACTH

2.0

14.6

7.5

6.8

6.4

7.3

6.0

-TSH TLH-FSH 12

F

Craniopharyngioma

0.3

N0lle

0.6

14.7

11.0

10.0

12.3

4.6

7.5

13

F

Idiopathic

0.8

NolIe

1.4

14.8

10.0

9.2

14.7

4.2

7.5

14

M

Idiopathic

1.4

None

3.9

15.1

11.0

9.8

12.5

4.6

8.5

15

M

Idiopathic

5.6

N0lle

2.8

15.2

11.3

10.5

10.5

5.8

8.4

16

M

Optic Glioma

0.9

ACTH

5.2

15.6

13.5

12.5

13.8

4.0

5.9

1.0

17.2

12.3

7.8

7.5

6.7

6.2

THS LH-FSH 17

F

1.1

Idiopathic

ACTH LH-FSH

Mean

k

3.0

SD

+ 1.6

Using the method of least squares, we performed linear regression analyses to determine the relationship between BOHB and glucose and between BOHB and age. By analysis of covariance we compared the concentrations of BOHB, adjusted for the glucose concentrations and for the ages of the subjects, in the growth hormone-deficient patients, before and after hGH therapy, with the concentrations in the control children2’ RESULTS

11.2

+ 5.1

7.9

? 4.4

7.0

t 4.0

8.1

+ 5.2

5.451.9

7.4

f 2.4

deficient patients had fasting concentrations of glucose indistinguishable from those of normal children of similar ages. Only the five youngest growth hormonedeficient patients became hypoglycemic (see Fig. 1). Three months of therapy with hGH did not increase the fasting concentrations of glucose in the 12 older patients nor in the four younger patients who completed the study.

Growth y= 0.075~ +2.73 r = 0.483, p
Human growth hormone increased growth velocity in all but one of the patients. She (patient No. 10 in Table 1) had a bone age of 14 years and had been menstruating regularly when treatment was started. The rate of growth before treatment was 3.0 f 1.6 (mean + SD) cm/year, which increased to 7.4 k 2.4 cm/year during the first year of treatment (P < 0.001) (see Table 1). Glucose The fasting plasma concentrations of glucose in the normal children correlated significantly with the ages of the subjects; the younger the child the lower the concentration of glucose: r = -0.48; P -C 0.001 (Fig. 1). Before treatment, the 12 older growth hormone-

I

2

3

4

5

6

7

8

9

IO II I2 I3 I4 15 I6 I7 IS

AGE (years) Fig. 1. The 24-hour fasting concentrations of glucose relative to age. Concentrations of the 17 growth hormone-deficient patients before treatment superimposed on the mean regression line, + 2 SD, of the concentrations of the 45 control children: y = 0.075x + 2.73, r - 0.453. P < 0.001.

KETONES AND GLUCOSE IN GH DEFICIENCY

459

The concentration of glycerol before treatment was 0.180 t 0.95 mM (mean +- SD) and was virtually unchanged (0.18 rt 0.101 mM) after treatment (Fig.

t&on+ SEM Beforetreatment After treatment

3).

GLUCOSE

LACTATE

ALANINE

INSULIN

Fig. 2. The concentrations (mean ? SEMI of glucose, lactate, and alanine (mM) and of insulin (@/ml) in the 16 growth hormone-deficient patients, before (open bars) and after (speckled bars) treatment with hGH.

In the normal children, the serum concentrations of BOHB correlated inversely both with age (r = -0.84; P < 0.001) (Fig. 4) and with the concentrations of glucose (r = -0.63; P < 0.01) (Fig. 5). The younger the child and the lower the concentration of glucose, the greater the concentration of BOHB. Before hGH therapy, the concentrations of BOHB in the growth hormone deficient patients also correy =4.8-0.25x r =-09837, p < 0.001

In the 16 growth hormone-deficient patients, the fasting concentration of glucose before therapy was 3.4 i 1.3 mM (mean + SD). It was virtually identical after treatment; 3.3 + 1.3 mM (Fig. 2). Glxoneogenic

Substrates

The concentrations of the gluconeogenic substrates-lactate, pyruvate, and alanine-in the growth hormone-deficient patients did not differ from those in the control children (data not shown) nor did they change significantly after hGH treatment (see Fig. 2). Before treatment, the concentration of lactate in the growth hormone-deficient patients was 0.97 t 0.42 mM (mean + SD), pyruvate was 0.04 * 0.01 mM, and alanine 0.229 t 0.122 mM. After hGH therapy the concentrations were 0.99 + 0.63 mM, 0.05 +- 0.03 mM, and 0.230 t 0.109 mM, respectively. Fai-Derived

;

7-

B. y=3.9-0.22x

Z6

r=-0.830,p<0.001

Fuels

The concentration of free fatty acids in the growth hormone-deficient patients before treatment was I .81 + 0.56 mM (mean + SD) and 1.93 + 0.79 mM after treatment; P > 0.1 (Fig. 3).

y= 4.7-0.25x

r =-O-760, p
2

4

6

8

IO 12 14 16 18

AGE (years)

FFA

GLYCEROL

B-OH6

AcAc

TOTAL KETONES

The concentrations (mean * SEMI of free fatty acids, Fig. 3. glycerol, @-hydroxybutyrate (BOHB), acetoacetate (AcAc) and total ketones in 16 growth hormone-deficient patients before and after treatment with hGH. lP < 0.01.

Fig. 4. Relationship between the concentrations of serum &hydroxybutyrate and age at the end of the fasts. (A) Concentrations of fl-hydroxybutyrate and mean regression line in 23 normal children, y = 4.8-0.25x. r = -0.837; P i 0.001. (B) Concentrations of fl-hydroxybutyrate and mean regression line in the 16 growth hormone-deficient patients before treatment, y = 3.8-0.22~. r = -0.83; P < 0.001. (Cl Concentrations of & hydroxybutyrate and mean regression line in the 16 growth hormone-deficient patients after treatment, y = 4.7-0.25x. r = 0.76; P < 0.001.

WOLFSDORF, SADEGHCNEJAD,

460

~~6.48~1.16x

AND SENIOR

during the fast and were not significantly changed by the treatment; 10.5 f 7.6 pU/ml (mean -t SD) before treatment versus 6.4 + 7.1 r*U/ml after treatment (P > 0.05) (see Fig. 2).

r =-0.630,pcOx)I

.

DISCUSSION

y=4.02-0.76x

y=4.65-0.05x

r =-0.604, p
r =-0.684, pcO.005

c.

I23456

l

123456 PLASMA

GLUCOSE mM

Fig. 5. The relationship between the concentrations of serum &hydroxybutyrate and plasma glucose at the end of the fasts. (A) Concentrations and regression line in the 23 normal children, y = 6.48- 1.16x. r = 0.630, P < 0.01. (6) Concentrations and regression line (broken line) in the 16 untreated growth hormonedeficient patients, y = 4.02-0.78~. r = 0.804; P c 0.001. The solid line is the regression line of the control children shown in A. (C) Concentrations and regression line (broken line) of the 16 growth hormone-deficient children after treatment, y = 4.65-0.85x. r = -0.684; P < 0.005. The solid line is the regression line of the control children shown in A.

lated inversely with age (r = -0.83; P < 0.001) (see Fig. 4) and with the concentration of glucose (r = -0.80; P -c 0.001) (see Fig. 5). The inverse correlation of BOHB with age (r = -0.76; P c 0.001) (see Fig. 4) and with the concentration of glucose (r = -0.68; P < 0.005) (see Fig. 5) persisted after the hGH therapy. Therapy with hGH increased the concentrations of BOHB from 1.38 + 1.30 mM to 1.89 f 1.63 mM (P -c 0.01; see Fig. 3). It also caused a small but insignificant increase in acetoacetate concentration from 0.24 + 0.19 mM to 0.29 + 0.17 mM (P > 0.1; see Fig. 3). Adjusted for both age and for the concentrations of glucose by analysis of covariance, the mean concentration of BOHB was significantly lower in the growth hormone-deficient patients than in the control children before (P < 0.01) as well as after therapy (P -c 0.01). Insulin The serum concentrations hormone-deficient patients

of insulin in the growth were appropriately low

As in other studies,637 only the youngest growth hormone-deficient patients became hypoglycemic. Any hypothesis attempting to explain how a lack of growth hormone predisposes a patient to fasting hypoglycemia should account for this finding. In order for normoglycemia to be sustained during fasting, the production and utilization of glucose must be in equilibrium. Production entails the release of glucose from liver glycogen, glycogenolysis, as well as the synthesis of new glucose, gluconeogenesis. Past studies indicate that neither glycogenolysis nor gluconeogenesis is impaired in growth hormone deficienof glucose during fasting cy. ‘x+v’ The utilization largely reflects consumption by the brain.** Because ketones are an alternative fuel for the brain they spare glucose during fasting. Within the physiologic range, utilization of ketones, particularly by the brain, is linearly related to the concentration; the greater the concentration, the more ketones are oxidized.” Thus, as the concentration of ketones increases, the consumption of glucose diminishes.” The availability of ketones, in turn, depends upon an adequate supply of free fatty acids being released from adipose stores and their synthesis into ketones by the liver. We found that the fasting concentrations of BOHB, the major ketone, varied over a 20-fold range and correlated inversely with age and with the concentrations of glucose in both groups. A comparison of the concentrations of ketones between individuals or groups can only be meaningful when adjusted for these factors. By analysis of covariance, the concentrations of BOHB in the growth hormone-deficient patients were significantly lower than in the normal children, before as well as after hGH therapy. The therapy did evoke a significant, but small, increase in the concentration of ketones. It did not restore ketone concentrations to normal. We have emphasized that in the fasting state, ketones constitute a substantial portion of the fuel oxidized by the brain and thereby decrease the oxidation of glucose.” Because the utilization of ketones relates directly to their concentration,” the youngest children with the highest concentrations of ketones oxidize the most ketones with a proportionate diminution in their oxidation of glucose. Thus, by conserving glucose, they sustain normoglycemia. By the same token, this younger age group would be particularly

KETONES AND GLUCOSE IN GH DEFICIENCY

461

vulnerable to a shortage of ketones. This would compromise their ability to conserve glucose and predispose the child to hypoglycemia. Accordingly, we propose that hypoketonemia is a principal cause of hypoglycemia in the younger growth hormone-deficient patients, but it may not be the sole factor. During fasting, the brain is the major consumer of glucose. In the fasting young child the equilibrium between the production and utilization of glucose is more precarious because of the greater size of the brain in relation to body size as well as the greater consumption of glucose per unit mass of the brain.23 This is evident in normal children; the younger the child, the lower the fasting level of glucose (see Fig. 1). In children with a growth hormone deficiency, the body is disproportionately small relative to the size of the brainz4 a disproportion that is most marked in the younger patients. We therefore suggest that the exaggerated brain:body disproportion also plays a part by amplifying the normal tendency of all fasting children to become hypoglycemic. The failure of growth hormone therapy to fully restore ketogenesis and thereby elevate the fasting concentrations of glucose could be ascribed to the dose and timing of therapy. Although the precise rate of growth hormone secretion in fasting children is not known, it is likely that the amount of growth hormone administered was quantitatively different from that secreted by normal fasting subjects. Also, the current method of administering growth hormone does not

duplicate endogenous secretion, which is diurnal and pulsatile. Use of an infusion pump to deliver growth hormone produced marked elevations of free fatty acids25 whereas in our patients free fatty acid levels at the end of the fast, 24 hours after administering growth hormone, were not significantly different than before treatment. Growth hormone therapy was biologically effective in our patients in terms of promoting growth; but, because of the method of administration, the full effects on body fuels may not have been achieved. Because ketones, as an alternative fuel, aid in conserving glucose, a disruption of the supply of ketones, however brought about, should predispose the patient to fasting hypoglycemia. Accordingly, a broader implication of the present study is that hypoketonemia should merit consideration as a factor in the pathogenesis of a wide variety of fasting hypoglycemias. An important caveat derives from our study. Young children who lack growth hormone, particularly those under five years of age, remain susceptible to fasting hypoglycemia despite the administration of hGH in the currently recommended manner. Thus, treatment can be expected to improve the rate of growth of these children, but the tendency to fasting hypoglycemia will still be present. ACKNOWLEDGEMENT We gratefully acknowledge the donation of human Growth Hormone by the National Pituitary Agency and the help of Dr. Leslie Lipworth with the statistical analyses.

REFERENCES 1. Brasel JA. Wright JC, Wilkins L, Blizzard RM: An evaluation of 75 patients with hypopituitarism beginning in childhood. Am J Med 38:484-494, 1965 2. Wilber JF. Odell WD: Hypoglycemia and dwarfism associated with the isolated deficiency of growth hormone. Metabolism 14:590597, 1965 3. Goodman HG, Grumbach MM, Kaplan SL: Growth and growrh hormone 11. A comparison of isolated growth hormone deficrency and multiple pituitary-hormone deficiencies in 35 patients with idiopathic hypopituitary dwarfism. N Engl J Med 278:57-68, 1968 4. Roe TF, Kogut MD: Hypopituitarism and ketotic hypoglycemia. 9m J Dis Child 121:296-299. 1971 5. Haymond MW, Karl 1, Weldon VV, Pagliara AS: The role of growth hormone and cortisone on glucose and gluconeogenic substrate regulation in fasted hypopituitary children. J Clin Endocrinol Metab 42:846-856, 1976 6. Hopwood, NJ, Forsman PJ, Kenny FM, Drash AL: Hypoglycemia in hypopituitary children. Am J Dis Child 129:918-926. 1975 7. Zuppinger KA: Hypoglycemia in childhood: evaluation of diagnostic procedures. Monogr Paediatr 4:53-66, 1975 8. Merimee TJ, Felig P, Marliss E, Fineberg SE, Cahill GF Jr: Glucose and lipid homeostasis in the absence of human growth hormone. J Clrn Invest 50:574-582, 1971 9. Wolfsdorf JI, Sadeghi-Nejad A, Senior B: Fat-derived fuels

during

24-hour

starvation

in children.

Em J Pediatr

138:141--144,

1982 IO. Saudubray JM, Marsac C, Lima1 JM, Dumurgier E, Charpentier C, Ogier H. Coude FX: Variation in plasma ketone bodies during a 24-hour fast in normal and in hypoglycemic children: relationship to age. J Pediatr 98:904-908. I98 1. 11. Robinson AM, Williamson DH: Physiological roles of ketone bodies as substrates and signals in mammalian tissues. Physiol Rev 60:143-187, 1980 12. Kadish AH, Little, RL, Sternberg JC: A new and rapid method for the determination of glucose by measurement of rate of oxygen consumption. Clin Chem 14: I 16-l 3 1, 1968 13. Hohorst HJ: L-(+)-lactate. Determination with lactic dehydrogenase and DPN, in Bergmeyer HU (ed): Methods of Enzymatic Analysis. New York, Academic Press, 1963, p 266 14. Butcher T, Czok R, Lamprecht W, Latzko E: Pyruvate. in Bergmeyer HU (ed): Methods of Enzymatic Analysis. New York. Academic Press, 1963. p 253 15. Dole VP, Meinertz H: Microdetermination fatty acids in plasma and tissues. J Biol Chem 1960

of long-chain 285:259552599.

16. Williamson DH, Mellanby J. Krebs HA: Enzymic determination of D( -)-P-hydroxybutyric acid and acetoacetic acid in blood. Biochem J 82:90-96. 1962 17. Wieland

0: Glycerol

UV-method,

in Bergmeyer

HU (ed):

462

Methods of Enzymatic Analysis. New York, Academic Press, 1974, vol. 3, pp 1404-1408 18. Spackman DH: Model 120B Amino Acid Analyzer Instruction Manual. Palo Alto, CA, Beckman Instruments Inc., Spinco Division, 1962 19. Soeldner JS, Slone D: Critical variables in the radioimmunoassay of serum insulin using the double antibody technic. Diabetes 14:771-779,1965 20. Snedecor GW, Cochran WC: Statistical Methods (ed 7). Iowa State University Press, Ames, Iowa, 1980 21. Underwood LE, Van den Brande JL, Antony GJ, Voina SJ, Van Wyk JJ: Islet cell function and glucose hemeostasis in hypopituitary dwarfism: synergism between growth hormone and cortisone. J Pediatr 82:28-37, 1973

WOLFSDORF, SADEGHI-NEJAD,

AND SENIOR

22. Cahill GF Jr: Starvation in man. Clin Endocrinol Metab 5:397415,1976 23. Sokoloff L: Circulation and energy metabolism of the brain, in Albers RW, Siegel GJ, Katzman R, et al (eds): Basic Neurochemistry. Boston, Little Brown, 1972, pp 299-325 24. Zachmann M, Fernandez F, Tassinari D, Thakker R, Prader A: Anthropometric measurements in patients with growth hormone deficiency before treatment with human growth hormone. Eur J Pediatr 133:277-282, 1980 25. Gertner J, Page S, Tamborlane W: Continuous subcutaneous infusion of growth hormone in growth hormone deficiency: feasibility and short-term metabolic effects. (Abstr) Pediatr Res 15:508, 1981