- Email: [email protected]
Glycemic response to lipid infusion in the premature neonate
108
January 1982 The J o u r n a l o f P E D I A T R I C S
Glycemic response to lipid infusion in the premature neonate The effect of lipid infusion on glucose homeostasis in the preterm newborn infant was evaluated. Seven infants were given a test dose of 0.25 gm/kg/hour o f lipid emulsion. Their response was characterized by: (1) a twofold increase in serum free fatty acid concentrations, (2) a small, transient rise in insulin values, and (3) a sustained increase in serum glucose eoneentration (mean increment in serum glucose was 24% over baseline). Nine infants received a test dose o f 0.5 g m / k g / h r o f lipid. Their response was similar to that in the lower infusion group, but o f a greater magnitude: an eightfoM increase in free fatty acids', sustained increase in serum insulin concentration, and a mean increment in serum glucose values of 65% over baseline. Increased lipid availabifity in the low-birth-weight newborn infant plays a significant role in promoting an increase in serum glucose concentrations.
R i t a A. Vileisis, M . D . , R i c h a r d M. Cowett, M.D., and W i l l i a m Oh, M.D.,* P r o v i d e n c e , R . L
ADEQUATE CALORIC INTAKE in the premature infant
is often difficult to achieve without intravenous nutritional supplementation. Consequently, many infants receive lipid emulsion in addition to glucose-amino acid mixtures in an attempt to maintain positive caloric and nitrogen balance. In experimental models, increased lipid availability, both in vivo and in vitro, has been reported to produce contradictory effects on carbohydrate metabolism. 1. ~ For example, both improved :' and worsened 4 glucose tolerance tests have been described, as have both increases s and decreases 6 in blood glucose concentration after lipid administration. Insulin responses to lipid ingestion and infusion have also been variable, ranging from little change to significant increases caused by high concentrations of plasma free fatty acids? In the human newborn infant, the role of lipid infusion in glucose regulation remains undefined. Thus, the specific aim of this study
From the Department o f Pediatrics, Women and Infants Hospital o f Rhode Island, Brown University Program in Medicine. Supported in part by Major Research Program, HD11343 National Institute o f Child and Human Development. *Reprint address: Women and lnfants Hospital of RL 50 Maude St., Providence, RI 02908.
VoL lO~No. 1, pp. 108-112
was to define the effect of lipid infusion on glucose homeostasis in the premature human infant. MATERIALS
AND METHODS
The study population included 16 clinically stable premature infants who were receiving glucose-amino acid Abbreviation used FFA: free fatty acid
]
mixtures parenterally. To provide additional caloric supplementation, the primary physician had prescribed intravenous lipid infusion as part of nutritional management. To evaluate the difference in response to two different doses of lipid emulsion, study infants were divided into two groups, based on the test dose of lipid administered. The clinical characteristics of each study group are listed in the Table. One infant in each group was small for gestational age. In group A (infants receiving 0.25 gm/ kg/hour) four infants were receiving theophylline for apnea of prematurity. All patients were giveh nothing by mouth except one, who had received two small feedings and was fasted for four hours prior to study. One patient was receiving supplemental oxygen. Three infants of group B (0.5 gm/kg/hour) were receiving theophylline therapy for apnea. The three patients who were begun on small oral feedings were also fasted for four to six hours
0022-3476/82/010108+05500.50/0@
1982 The C. V. Mosby Co.
Volume 1O0 Number 1
prior to initiation of the study protocol. Three patients in this group were receiving supplemental oxygen. The mean serum indirect bilirubin value in all study patients was 5.3 _+ 0.4 mg/dl (M _+ SEM); none of the infants was receiving phototherapy. The protocol was begun after informed parental consent was obtained. Patients received the parenteral nutrition solution without lipid via constant infusion pump for 12 to 18 hours prior to the initiation of the study. During the first hour of study, baseline serum samples were obtained via heelstick for the following determinations: glucose ( x 3), free fatty acid, and insulin. To qualify as being in a steady state, the maximal allowable variability in the three serum glucose values was < 10%. After the baseline period, group A patients received 0.25 gm/kg/hr for two hours of lipid emulsion (t0% Intraiipid, Cutter Laboratories, Irvine, Calif.) via a second constant infusion syringe pump. Blood samples were obtained every hour for FFA and insulin values in addition to serum glucose determinations taken every half hour. One hour after termination of the lipid infusion, while the patients were still receiving glucose and amino acids, final glucose, insulin, and FFA values were determinedl Group B patients followed the identical protocol except that they received 0.5 gm/kg/hour of lipid emulsion. Maximum blood withdrawal from each infant was 2 ml during the entire study protocol. Throughout the study period, no Patient received any feedings or theophylline medication. No episodes of apnea or hypothermia were observed. Glucose was determined by the glucose oxidase method on a YSI glucose analyzer (Yellow Springs Instrument Co., Inc., Yellow Springs, Ohio): .insulin, by radioimmunoassayT; and FFA, by a radioactive tracer technique. ~ Statistical analysis between groups A and B was performed using the unpaired t test; data between baseline and study period were analyzed by paired t test. Significance was assigned to the P < 0.05 level. RESULTS Serum FFA concentration increased twofold in group A from a baseline of 263 _ 40 /~Eq/L (M _+ SEM) to 554 _+ 43 /~Eq/L by two hours of lipid infusion, then declined to 381 +_ 41 /xEq/L by one hour after cessation of the infusion (Fig. 1). FFA values were significantly increased over baseline by one hour of lipid infusion. Similarly, in group B, FFA values increased eightfold over baseline by the end Of the infusion period. All differences from baseline were statistically significant and showed a dose-response effect between groups. The baseline serum glucose variability was 6% for both groups. During the baseline period the three serum
Glycemic response to lipid infusion
109
FFA (~.EqlL) 2200
INTRALIPID
DOSE
91kglhrl I
o GroupA=0.25 (n=7)
t,~;?~B=~176
: a:00'0055 V~r B:pe;Si;:B~
1800.
1400
/
/I* / I
*
I
-
I000
600
200
I
LIPID INFUSION
I
0
.
60
I
I
I
120
180
I
240
Baseline "
9
TIME
(Minutes}
Fig. 1. Effect of lipid infusion on serum free fatty acid (FFA) concentrations. glucose values were averaged and designated as prelipid values. The mean of these prelipid values in group A was 98 + 8 mg/dl (M _+ SEM). Following lipid infusion, a steady and sustained rise in serum glucose concentration was observed (Fig. 2, A). After 120 minutes of study the serum glucose values were statistically higher than the baseline values. The peak value of 121 +_ 9 mg/dl occurred at 180 minutes; it persisted at that level for one hour after discontinuation of lipid infusion. Similar changes, but of a greater magnitude, were observed for the serum glucose values of the infants in group B. Mean baseline glucose concentration averaged 79 _+ 9 mg/dl, which increased to a peak of 124 _+ 13 after two hours of lipid administration and was unchanged one hour after discontinuation of the lipid infusion (122 _+ 8 mg/dl). As in group A, the glucose concentration w a s significantly elevated compared to baseline values after 120 minutes of study. The glucose data were also examined as a percent increment over baseline (Fig. 2, B). For group A, the increase in serum glucose values was 14% at 120 and at 180 minutes. At 240 minutes the increment was 24% above the baseline values. In group B, a greater increment from baseline values was observed: 29% at 120 minutes, 63% at 180 minutes, and 65% at 240 minutes. Between the two
110
Vileisis, Cowett, a n d Oh
The Journal o f Pediatrics January 1982
SERUM GLUCOSE (mg/dl)
Table. Clinical characteristics of study infants Patients groups
]50 t IO
.
sss
7O I 50
LIPID INFUSION
f
0
l
60
120
180
240
Baseline
TIME
(Minutes)
Lipid test dose (gm/ kg/hr) Birth weight (gin) Gestational age (wk) Chronologic age (days) Dextrose intake (gm/ kg/day) Amino acid intake (gm/kg/day)
A
B
(n = 7)
(n = 9)
0.25
0.50
1,350 + 50 31.6 + 0.7 7.0 _-2-0.4 12.4 _+ 0.7
1,200 _+ 80 30.1 _+ 0.5 7.0 _+ 0.8 15.3 _+ 1.2
1.4 + 0.1
1.7 + 0.2
M 4- SEM, no significant differences were observed between the two
groups. % INCREASE 80"
IN G L U C O S E
FROM
BASELINE
T
~[
INTRALIPlD DOSE q/kg/hr
o GroupAoo.2s~n='7~ 60" 40 "
9 Group B =0.SO(n =9) M tSEM p
20. O"
0
T
I ,
/
~
~-?. . . . . 2 , + -
[
LIPID INFUSION
I
610
I
I
Baseline
120
TIME
")1"
"1~
",.z"/" , s
k.o]
180
I
240
(Minutes)
Fig. 2. A, Effect of lipid infusion on serum glucose concentrations. B, Effect of lipid infusion on serum glucose concentrations. expressed as a percent increase from baseline. study groups, the increment was statistically greater in group B at 150, 180, and 240 minutes of the study period. Serum insulin levels increased from a preinfusion value of 13.4 _ 2.0 /~U/ml to 21.3 _+ 2.4 /xU/ml in group A after one hour of triglyceride emulsion (P < 0.05) (Fig. 3). However, the 180- and 240-minute levels were not statistically different from baseline in group A. In grouP B a statistically significant rise in serum insulin concentration was observed at 180 and 240 minutes following the intravenous infusion of lipid. DISCUSSION At least three physiologic events resulting from increased fatty acid availability (secondary to lipid oxidation) may account for the increment in serum glucose concentration that was induced by lipid administration in our studies: (1) decreased insulin activity in the peripheral tissues; (2) reduced rates of glycolysis; and (3) enhanced gluconeogenesis. In vitro studies in perfused rat heart and diaphragm muscle have documented decreased insulin effectiveness
in these tissues in the presence of increased free fatty acid or ketone body availability? Another possible mechanism to explain the increase in serum glucose concentration is a reduction of glycolysis caused by high F F A concentrations, as shown by Randle et al' and by S61ing et a12 A decrease in glyeolysis results from inhibition of phosphofructokinase, hexokinase, and pyruvate oxidation. The accelerated rate of fatty acid oxidation is responsible for increased acetyl C o A / C o A and N A D H / N A D ratios, which in turn account for the inhibition of key glycolytic enzymes. -~The third possible biochemical mechanism has been postulated by Ferr6 et al '~ in the fasted newborn rat. Free fatty acid availability and oxidation are essential to provide adequate ATP, NADH, and acetyl CoA Ibr maintenance of rapid rates of gluconeogenesis. Thus, enhanced gluconeogenesis following fat infusion may serve as an additional explanation for the observed increase in serum glucose concentrations. However, conflicting results have been reported in other species 6 and in the adult human. Schalch and Kipnis ~ found that after tripling plasma F F A concentrations, glucose disappearance rates decreased markedly during intravenous glucose tolerance tests. However, Pelkonen et aP found that improved glucose tolerance accompanied elevated F F A values. In these studies, elevated F F A concentrations were achieved by means of oral ingestion of fat plus administration of heparin, a condition which may not be strictly comparable to intravenous fat infusion. In our study, we did not document fatty acid oxidation. However, based on studies utilizing respiratory quotients and ketone body levels following lipid infusion in ~he premature infantJ ~- ~ it is very likely that increased lipid oxidation did occur and probably was responsible for the glycemic response. Three clinical studies evaluated the effect of intrave-
Volume I00 Number 1
Glycemic response to lipid infusion
SERUM INSULIN (,uU/ml)
nous lipid infusion on glucose metabolism. In studies of adults ~3 a n d of adolescents, '4 decreased glucose tolerance was observed as a direct consequence o f increased F F A utilization. Mesty/m et al 1~ reported a small increase in blood glucose concentrations c o n c o m i t a n t with fat infusion in infants o f a larger birth weight (x = 1,736 gm) who were receiving 5 or 10% glucose solutions. O u r findings in the very low-birth-weight i n f a n t are consistent with these observations. W e c a n n o t discount a c o n t r i b u t i o n from the glycerol c o m p o n e n t in lntralipid in causing an i n c r e m e n t in blood glucose concentration, as suggested by W o l f et al. 12 However, in glycerol tolerance tests in older children, a m u c h larger a n d more rapid infusion of glycerol t h a n administered here resulted only in a n average increase in serum glucose concentration o f 9 _+ 1 m g / d l . '~ M a n y factors (glycerol) 7 ketone bodies, 1. F F A t h e m selves, 19 or small increments in serum glucose ~~ m a y play a role in the increase in serum insulin concentration noted in our study infants. O u r study design does not p e r m i t specific designation of which factor plays the predomin a n t role. In summary, in the low-birth-weight neonate, the metabolism o f intravenous lipid emulsions significantly alters glucose homeostasis, a n d may result in a glycemic
response. The authors thank Ms. B. Kelley and Mrs. T. Bienieki for their skillful technical assistance and Ms. D. Perry for secretarial assistance. REFERENCES 1. Randle PJ, Garland PB, Hales CN, and Newsholme EA: The glucose fatty acid cycle, Lancet 1:364, 1963. 2. Ruderman NB, Toews CJ, and Shafrir E: Role of free fatty acids in glucose homeostasis, Arch Intern Med 123:299, 1969. 3. Pelkonen R, Miettinen TA, Taskinen MR, and Nikkila EA: Effect of acute elevation of plasma glycerol, triglyceride and FFA levels on glucose utilization and plasma insulin, Diabetes 17:76, 1968. 4. Schalch DS, and Kipnis DM: Abnormalities in carbohydrate tolerance associated with elevated plasma nonesterifled fatty acids, J Clin Invest 44:2010, 1965. 5. Friedman B, Goodman EH, and Weinhouse S: Effect of insulin and fatty acids on gluconeogenesis in the rat, J Biol Chem 242:3260, 1967. 6. Seyffert WA, and Madison LL: Physiologic effects of metabolic fuels on carbohydrate metabolism I. Acute effect of elevation of plasma free fatty acids on hepatic glucose output, peripheral glucose utilization, serum insulin and plasma glucagon levels, Diabetes 16:765, 1967. 7. Hales CN, and Randle PH: Immunoassay of insulin with insulin antibody precipitates, Biochem J 88:137, 1963. 8. Tokumitsu Y, Kondoh T, and Ui M: Radiochemical assay of free fatty acid in blood using Cobalt 57 as a tracer, Anal Biochem 81:488, 1977. 9. SOling HD, Willms B, Friericks D, and Kelineke J: Regula-
11 1
INTRALIPID DOSE g/kcl/hr I o GroupA=0,25 (n=7) I 60
9w Group v , ~ v B = 0 . 5 0 (n =9) M •tSEM
I
* pp
I
+ p,~O.O p <0.05
I "~ +
Group
5O
40
3O
20
I0
[
LIPID INF.
I
I
I
I
60
120
180
240
]
Baseline
TIME (Minutes) Fig. 3. Effect of lipid infusion on serum insulin concentrations.
10.
11.
12.
13.
14.
15.
16.
tion of gluconeogenesis by fatty acid oxidation in isolated perfused livers of nonstarved rats, Eur J Biochem 4:364, 1968. Ferr6 P, Pegorier JP, Marliss EB, and Girard JR: Influence of exogenous fat and gluconeogenic substrates on glucose homeostasis in the newborn rat, Am J Physiol 234:E129, 1978. Rubecz I, Mesty~in J, Varga P, and Klujber L: Energy metabolism, substrate utilization and nitrogen balance in parenterally fed postoperative neonates and infants, J PEDIATR 98:42, 1981. Wolf H, Berg WV, Kerstan J, Lausmann S, Ley HG, Lohr H, Melichar V, and Otten A: Metabolic responses to I.V. fat in the newborn, in Hahn R, Segal S, and Israel S, editors: Role of fat in intravenous feeding of the newborn, Quebec, 1979, Pharmacia Ltd., p 49. Felber JP, and Vannoti A: Effects of fat infusion in glucose tolerance and insulin plasma levels, Med Exp 10:153, 1964. Persson B, Sterky G, and Thorell J: Effect of fat infusion on plasma glucose FFA, glycerol and insulin levels during intravenous and oral glucose tolerance tests, Prog Biochem Pharmacol 3:403, 1967. Mestyfin J, Rubecz 1, and Solt~sz G: Changes in blood glucose, free fatty acids and amino acids in low birth weight infants receiving intravenous fat emulsion, Bio Neonate 30:74, 1976. Senior B, and Loridan L: Studies in liver glycogenesis, with particular reference to the metabolism of intravenously administered glycerol, N Engl J Med 279:958, 1968.
112
Vileisis, Cowett, and Oh
17. Miettinen TA, Pelkonen R, Vesenne F, and Nikkila EA: Effects of plasma triglyceride, FFA, glycerol levels on plasma insulin and its response, Diabetologia 2:232, 1966. 18. Balanasse EO, Ooms HA, and Lambilliotte JP: Evidence for a stimulating effect of ketone bodies on insulin secretion in man, Horm Metab Res 2:371, 1970. 19. Crespin SR, Greenough WB, and Steinberg D: Stimulation
The Journal of Pediatrics January 1982
20.
of insulin secretion by infusion of free fatty acids, J Clin Invest 48:1934, 1969. Pelkonen R, Taskinen MR, and Nikkila EA: Early responses of plasma insulin to small doses of intravenous glucose: Effect of obesity. J Clin Endocrinol Metab 39i418, 1974.
January 1982 The J o u r n a l o f P E D I A T R I C S
Glycemic response to lipid infusion in the premature neonate The effect of lipid infusion on glucose homeostasis in the preterm newborn infant was evaluated. Seven infants were given a test dose of 0.25 gm/kg/hour o f lipid emulsion. Their response was characterized by: (1) a twofold increase in serum free fatty acid concentrations, (2) a small, transient rise in insulin values, and (3) a sustained increase in serum glucose eoneentration (mean increment in serum glucose was 24% over baseline). Nine infants received a test dose o f 0.5 g m / k g / h r o f lipid. Their response was similar to that in the lower infusion group, but o f a greater magnitude: an eightfoM increase in free fatty acids', sustained increase in serum insulin concentration, and a mean increment in serum glucose values of 65% over baseline. Increased lipid availabifity in the low-birth-weight newborn infant plays a significant role in promoting an increase in serum glucose concentrations.
R i t a A. Vileisis, M . D . , R i c h a r d M. Cowett, M.D., and W i l l i a m Oh, M.D.,* P r o v i d e n c e , R . L
ADEQUATE CALORIC INTAKE in the premature infant
is often difficult to achieve without intravenous nutritional supplementation. Consequently, many infants receive lipid emulsion in addition to glucose-amino acid mixtures in an attempt to maintain positive caloric and nitrogen balance. In experimental models, increased lipid availability, both in vivo and in vitro, has been reported to produce contradictory effects on carbohydrate metabolism. 1. ~ For example, both improved :' and worsened 4 glucose tolerance tests have been described, as have both increases s and decreases 6 in blood glucose concentration after lipid administration. Insulin responses to lipid ingestion and infusion have also been variable, ranging from little change to significant increases caused by high concentrations of plasma free fatty acids? In the human newborn infant, the role of lipid infusion in glucose regulation remains undefined. Thus, the specific aim of this study
From the Department o f Pediatrics, Women and Infants Hospital o f Rhode Island, Brown University Program in Medicine. Supported in part by Major Research Program, HD11343 National Institute o f Child and Human Development. *Reprint address: Women and lnfants Hospital of RL 50 Maude St., Providence, RI 02908.
VoL lO~No. 1, pp. 108-112
was to define the effect of lipid infusion on glucose homeostasis in the premature human infant. MATERIALS
AND METHODS
The study population included 16 clinically stable premature infants who were receiving glucose-amino acid Abbreviation used FFA: free fatty acid
]
mixtures parenterally. To provide additional caloric supplementation, the primary physician had prescribed intravenous lipid infusion as part of nutritional management. To evaluate the difference in response to two different doses of lipid emulsion, study infants were divided into two groups, based on the test dose of lipid administered. The clinical characteristics of each study group are listed in the Table. One infant in each group was small for gestational age. In group A (infants receiving 0.25 gm/ kg/hour) four infants were receiving theophylline for apnea of prematurity. All patients were giveh nothing by mouth except one, who had received two small feedings and was fasted for four hours prior to study. One patient was receiving supplemental oxygen. Three infants of group B (0.5 gm/kg/hour) were receiving theophylline therapy for apnea. The three patients who were begun on small oral feedings were also fasted for four to six hours
0022-3476/82/010108+05500.50/0@
1982 The C. V. Mosby Co.
Volume 1O0 Number 1
prior to initiation of the study protocol. Three patients in this group were receiving supplemental oxygen. The mean serum indirect bilirubin value in all study patients was 5.3 _+ 0.4 mg/dl (M _+ SEM); none of the infants was receiving phototherapy. The protocol was begun after informed parental consent was obtained. Patients received the parenteral nutrition solution without lipid via constant infusion pump for 12 to 18 hours prior to the initiation of the study. During the first hour of study, baseline serum samples were obtained via heelstick for the following determinations: glucose ( x 3), free fatty acid, and insulin. To qualify as being in a steady state, the maximal allowable variability in the three serum glucose values was < 10%. After the baseline period, group A patients received 0.25 gm/kg/hr for two hours of lipid emulsion (t0% Intraiipid, Cutter Laboratories, Irvine, Calif.) via a second constant infusion syringe pump. Blood samples were obtained every hour for FFA and insulin values in addition to serum glucose determinations taken every half hour. One hour after termination of the lipid infusion, while the patients were still receiving glucose and amino acids, final glucose, insulin, and FFA values were determinedl Group B patients followed the identical protocol except that they received 0.5 gm/kg/hour of lipid emulsion. Maximum blood withdrawal from each infant was 2 ml during the entire study protocol. Throughout the study period, no Patient received any feedings or theophylline medication. No episodes of apnea or hypothermia were observed. Glucose was determined by the glucose oxidase method on a YSI glucose analyzer (Yellow Springs Instrument Co., Inc., Yellow Springs, Ohio): .insulin, by radioimmunoassayT; and FFA, by a radioactive tracer technique. ~ Statistical analysis between groups A and B was performed using the unpaired t test; data between baseline and study period were analyzed by paired t test. Significance was assigned to the P < 0.05 level. RESULTS Serum FFA concentration increased twofold in group A from a baseline of 263 _ 40 /~Eq/L (M _+ SEM) to 554 _+ 43 /~Eq/L by two hours of lipid infusion, then declined to 381 +_ 41 /xEq/L by one hour after cessation of the infusion (Fig. 1). FFA values were significantly increased over baseline by one hour of lipid infusion. Similarly, in group B, FFA values increased eightfold over baseline by the end Of the infusion period. All differences from baseline were statistically significant and showed a dose-response effect between groups. The baseline serum glucose variability was 6% for both groups. During the baseline period the three serum
Glycemic response to lipid infusion
109
FFA (~.EqlL) 2200
INTRALIPID
DOSE
91kglhrl I
o GroupA=0.25 (n=7)
t,~;?~B=~176
: a:00'0055 V~r B:pe;Si;:B~
1800.
1400
/
/I* / I
*
I
-
I000
600
200
I
LIPID INFUSION
I
0
.
60
I
I
I
120
180
I
240
Baseline "
9
TIME
(Minutes}
Fig. 1. Effect of lipid infusion on serum free fatty acid (FFA) concentrations. glucose values were averaged and designated as prelipid values. The mean of these prelipid values in group A was 98 + 8 mg/dl (M _+ SEM). Following lipid infusion, a steady and sustained rise in serum glucose concentration was observed (Fig. 2, A). After 120 minutes of study the serum glucose values were statistically higher than the baseline values. The peak value of 121 +_ 9 mg/dl occurred at 180 minutes; it persisted at that level for one hour after discontinuation of lipid infusion. Similar changes, but of a greater magnitude, were observed for the serum glucose values of the infants in group B. Mean baseline glucose concentration averaged 79 _+ 9 mg/dl, which increased to a peak of 124 _+ 13 after two hours of lipid administration and was unchanged one hour after discontinuation of the lipid infusion (122 _+ 8 mg/dl). As in group A, the glucose concentration w a s significantly elevated compared to baseline values after 120 minutes of study. The glucose data were also examined as a percent increment over baseline (Fig. 2, B). For group A, the increase in serum glucose values was 14% at 120 and at 180 minutes. At 240 minutes the increment was 24% above the baseline values. In group B, a greater increment from baseline values was observed: 29% at 120 minutes, 63% at 180 minutes, and 65% at 240 minutes. Between the two
110
Vileisis, Cowett, a n d Oh
The Journal o f Pediatrics January 1982
SERUM GLUCOSE (mg/dl)
Table. Clinical characteristics of study infants Patients groups
]50 t IO
.
sss
7O I 50
LIPID INFUSION
f
0
l
60
120
180
240
Baseline
TIME
(Minutes)
Lipid test dose (gm/ kg/hr) Birth weight (gin) Gestational age (wk) Chronologic age (days) Dextrose intake (gm/ kg/day) Amino acid intake (gm/kg/day)
A
B
(n = 7)
(n = 9)
0.25
0.50
1,350 + 50 31.6 + 0.7 7.0 _-2-0.4 12.4 _+ 0.7
1,200 _+ 80 30.1 _+ 0.5 7.0 _+ 0.8 15.3 _+ 1.2
1.4 + 0.1
1.7 + 0.2
M 4- SEM, no significant differences were observed between the two
groups. % INCREASE 80"
IN G L U C O S E
FROM
BASELINE
T
~[
INTRALIPlD DOSE q/kg/hr
o GroupAoo.2s~n='7~ 60" 40 "
9 Group B =0.SO(n =9) M tSEM p
20. O"
0
T
I ,
/
~
~-?. . . . . 2 , + -
[
LIPID INFUSION
I
610
I
I
Baseline
120
TIME
")1"
"1~
",.z"/" , s
k.o]
180
I
240
(Minutes)
Fig. 2. A, Effect of lipid infusion on serum glucose concentrations. B, Effect of lipid infusion on serum glucose concentrations. expressed as a percent increase from baseline. study groups, the increment was statistically greater in group B at 150, 180, and 240 minutes of the study period. Serum insulin levels increased from a preinfusion value of 13.4 _ 2.0 /~U/ml to 21.3 _+ 2.4 /xU/ml in group A after one hour of triglyceride emulsion (P < 0.05) (Fig. 3). However, the 180- and 240-minute levels were not statistically different from baseline in group A. In grouP B a statistically significant rise in serum insulin concentration was observed at 180 and 240 minutes following the intravenous infusion of lipid. DISCUSSION At least three physiologic events resulting from increased fatty acid availability (secondary to lipid oxidation) may account for the increment in serum glucose concentration that was induced by lipid administration in our studies: (1) decreased insulin activity in the peripheral tissues; (2) reduced rates of glycolysis; and (3) enhanced gluconeogenesis. In vitro studies in perfused rat heart and diaphragm muscle have documented decreased insulin effectiveness
in these tissues in the presence of increased free fatty acid or ketone body availability? Another possible mechanism to explain the increase in serum glucose concentration is a reduction of glycolysis caused by high F F A concentrations, as shown by Randle et al' and by S61ing et a12 A decrease in glyeolysis results from inhibition of phosphofructokinase, hexokinase, and pyruvate oxidation. The accelerated rate of fatty acid oxidation is responsible for increased acetyl C o A / C o A and N A D H / N A D ratios, which in turn account for the inhibition of key glycolytic enzymes. -~The third possible biochemical mechanism has been postulated by Ferr6 et al '~ in the fasted newborn rat. Free fatty acid availability and oxidation are essential to provide adequate ATP, NADH, and acetyl CoA Ibr maintenance of rapid rates of gluconeogenesis. Thus, enhanced gluconeogenesis following fat infusion may serve as an additional explanation for the observed increase in serum glucose concentrations. However, conflicting results have been reported in other species 6 and in the adult human. Schalch and Kipnis ~ found that after tripling plasma F F A concentrations, glucose disappearance rates decreased markedly during intravenous glucose tolerance tests. However, Pelkonen et aP found that improved glucose tolerance accompanied elevated F F A values. In these studies, elevated F F A concentrations were achieved by means of oral ingestion of fat plus administration of heparin, a condition which may not be strictly comparable to intravenous fat infusion. In our study, we did not document fatty acid oxidation. However, based on studies utilizing respiratory quotients and ketone body levels following lipid infusion in ~he premature infantJ ~- ~ it is very likely that increased lipid oxidation did occur and probably was responsible for the glycemic response. Three clinical studies evaluated the effect of intrave-
Volume I00 Number 1
Glycemic response to lipid infusion
SERUM INSULIN (,uU/ml)
nous lipid infusion on glucose metabolism. In studies of adults ~3 a n d of adolescents, '4 decreased glucose tolerance was observed as a direct consequence o f increased F F A utilization. Mesty/m et al 1~ reported a small increase in blood glucose concentrations c o n c o m i t a n t with fat infusion in infants o f a larger birth weight (x = 1,736 gm) who were receiving 5 or 10% glucose solutions. O u r findings in the very low-birth-weight i n f a n t are consistent with these observations. W e c a n n o t discount a c o n t r i b u t i o n from the glycerol c o m p o n e n t in lntralipid in causing an i n c r e m e n t in blood glucose concentration, as suggested by W o l f et al. 12 However, in glycerol tolerance tests in older children, a m u c h larger a n d more rapid infusion of glycerol t h a n administered here resulted only in a n average increase in serum glucose concentration o f 9 _+ 1 m g / d l . '~ M a n y factors (glycerol) 7 ketone bodies, 1. F F A t h e m selves, 19 or small increments in serum glucose ~~ m a y play a role in the increase in serum insulin concentration noted in our study infants. O u r study design does not p e r m i t specific designation of which factor plays the predomin a n t role. In summary, in the low-birth-weight neonate, the metabolism o f intravenous lipid emulsions significantly alters glucose homeostasis, a n d may result in a glycemic
response. The authors thank Ms. B. Kelley and Mrs. T. Bienieki for their skillful technical assistance and Ms. D. Perry for secretarial assistance. REFERENCES 1. Randle PJ, Garland PB, Hales CN, and Newsholme EA: The glucose fatty acid cycle, Lancet 1:364, 1963. 2. Ruderman NB, Toews CJ, and Shafrir E: Role of free fatty acids in glucose homeostasis, Arch Intern Med 123:299, 1969. 3. Pelkonen R, Miettinen TA, Taskinen MR, and Nikkila EA: Effect of acute elevation of plasma glycerol, triglyceride and FFA levels on glucose utilization and plasma insulin, Diabetes 17:76, 1968. 4. Schalch DS, and Kipnis DM: Abnormalities in carbohydrate tolerance associated with elevated plasma nonesterifled fatty acids, J Clin Invest 44:2010, 1965. 5. Friedman B, Goodman EH, and Weinhouse S: Effect of insulin and fatty acids on gluconeogenesis in the rat, J Biol Chem 242:3260, 1967. 6. Seyffert WA, and Madison LL: Physiologic effects of metabolic fuels on carbohydrate metabolism I. Acute effect of elevation of plasma free fatty acids on hepatic glucose output, peripheral glucose utilization, serum insulin and plasma glucagon levels, Diabetes 16:765, 1967. 7. Hales CN, and Randle PH: Immunoassay of insulin with insulin antibody precipitates, Biochem J 88:137, 1963. 8. Tokumitsu Y, Kondoh T, and Ui M: Radiochemical assay of free fatty acid in blood using Cobalt 57 as a tracer, Anal Biochem 81:488, 1977. 9. SOling HD, Willms B, Friericks D, and Kelineke J: Regula-
11 1
INTRALIPID DOSE g/kcl/hr I o GroupA=0,25 (n=7) I 60
9w Group v , ~ v B = 0 . 5 0 (n =9) M •tSEM
I
* pp
I
+ p,~O.O p <0.05
I "~ +
Group
5O
40
3O
20
I0
[
LIPID INF.
I
I
I
I
60
120
180
240
]
Baseline
TIME (Minutes) Fig. 3. Effect of lipid infusion on serum insulin concentrations.
10.
11.
12.
13.
14.
15.
16.
tion of gluconeogenesis by fatty acid oxidation in isolated perfused livers of nonstarved rats, Eur J Biochem 4:364, 1968. Ferr6 P, Pegorier JP, Marliss EB, and Girard JR: Influence of exogenous fat and gluconeogenic substrates on glucose homeostasis in the newborn rat, Am J Physiol 234:E129, 1978. Rubecz I, Mesty~in J, Varga P, and Klujber L: Energy metabolism, substrate utilization and nitrogen balance in parenterally fed postoperative neonates and infants, J PEDIATR 98:42, 1981. Wolf H, Berg WV, Kerstan J, Lausmann S, Ley HG, Lohr H, Melichar V, and Otten A: Metabolic responses to I.V. fat in the newborn, in Hahn R, Segal S, and Israel S, editors: Role of fat in intravenous feeding of the newborn, Quebec, 1979, Pharmacia Ltd., p 49. Felber JP, and Vannoti A: Effects of fat infusion in glucose tolerance and insulin plasma levels, Med Exp 10:153, 1964. Persson B, Sterky G, and Thorell J: Effect of fat infusion on plasma glucose FFA, glycerol and insulin levels during intravenous and oral glucose tolerance tests, Prog Biochem Pharmacol 3:403, 1967. Mestyfin J, Rubecz 1, and Solt~sz G: Changes in blood glucose, free fatty acids and amino acids in low birth weight infants receiving intravenous fat emulsion, Bio Neonate 30:74, 1976. Senior B, and Loridan L: Studies in liver glycogenesis, with particular reference to the metabolism of intravenously administered glycerol, N Engl J Med 279:958, 1968.
112
Vileisis, Cowett, and Oh
17. Miettinen TA, Pelkonen R, Vesenne F, and Nikkila EA: Effects of plasma triglyceride, FFA, glycerol levels on plasma insulin and its response, Diabetologia 2:232, 1966. 18. Balanasse EO, Ooms HA, and Lambilliotte JP: Evidence for a stimulating effect of ketone bodies on insulin secretion in man, Horm Metab Res 2:371, 1970. 19. Crespin SR, Greenough WB, and Steinberg D: Stimulation
The Journal of Pediatrics January 1982
20.
of insulin secretion by infusion of free fatty acids, J Clin Invest 48:1934, 1969. Pelkonen R, Taskinen MR, and Nikkila EA: Early responses of plasma insulin to small doses of intravenous glucose: Effect of obesity. J Clin Endocrinol Metab 39i418, 1974.