Alteration of Glucose and Insulin Metabolism
in Congenital Heart Disease
By GERSHON HAIT, M.D., MARNA CORPUS, M.D., FRANcOIs R. LAMARRE, M.D.,
SHANG-HSIEN YUAN, M.D., JINDRICH KYPSON, M.D.,
AND GRACE CHENG, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
SUMMARY
Children with left-to-right shunt, with and without congestive heart failure, were
found to have impaired glucose tolerance tests (GTT). In cyanotic children normal
levels of glucose were found in association with abnormally high levels of insulin
following oral GTT. Several possible mechanisms are proposed to explain the different
glucose tolerance alterations: (1) Suppression of insulin release appeared to partially
explain the low levels of insulin in congestive heart failure. This suppression may be
related to the high levels of circulating norepinephrine found in these patients. (2) Excessive clearance of insulin by the lung may also be responsible for the reduced arterial
insulin levels in patients with left-to-right shunt, and underclearance of insulin for the
abnormally higher arterial insulin levels in patients with right-to-left shunts in whom a
significant amount of venous blood has bypassed the lung. (3) Hypoxia of the pancreas
and the liver in cyanotic patients and those with congestive heart failure may explain
the reduction of insulin levels in the hepatic vein following i.v. glucose tolerance tests.
An excess production of a glucagonlike gastrointestinal factor in cyanotic children may
partially explain the abnormally high levels of insulin following oral GTT.
Additional Indexing Words:
Gastrointestinal factor
Cyanotic heart disease
Congestive heart failure
P.ulmonary clearance of insulin
Hepatic vein
Glucagonlike substance
and a normal or slightly elevated glucose
levels were found following oral glucose
tolerance tests.3
The aim of the present study was to
examine the possibility that gastrointestinal,
hepatic or pulmonary factors may play a role
in the metabolic alterations in children with
congenital heart disease with and without
heart failure. In an attempt to clarify the role
of the gastrointestinal tract, glucose and
insulin levels were compared following oral
and intravenous glucose tolerance tests
(GTT). To assess the role of the liver,
simultaneous hepatic venous and systemic
arterial glucose and insulin levels were compared following i.v. GTT. The role of the lung
as a possible insulin clearing organ was
investigated by comparing insulin levels in
RECENTLY we demonstrated that children in congestive heart failure have
impaired oral glucose tolerance tests. Suppression of insulin release appeared to be an
important contributing mechanism for this
abnormality.1' 2 However, in children with
cyanotic heart disease and diminished pulmonary blood flow, an abnormally high insulin
From the Department of Pediatrics, Albert Einstein
College of Medicine, New York, New York.
Supported by Training Grant HE 05532 of the
National Heart and Lung Institute, the New York
Heart Association, and New York State Heart
Assembly grants-in-aid.
Address for reprints: Dr. Gershon Hait, Pediatric
Cardiology, Albert Einstein College of Medicine, 1300
Morris Park Avenue, Bronx, New York 10461.
Received January 13, 1971; revision accepted for
publication March 27, 1972.
Circulation, Volume XLVI, August 1972
333
HAIT ET AL.
334
pulmiioniary arterial anid aortic blood samples
during ix. GTT.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Methods
Glucose tolerance tests (GTT) were performed
on 54 patients raniging in age from 12 days to 16
years. In 36 of the patients oral GTT was
performed. In 20, both oral and i.v. GTT were
performed. In 17, only i.v. GTT was done. In four
of the patients, oral GTT was repeated three
times to coincide with an observed increase in the
intensity of the cyanosis or an improvement of the
congestive heart failure.
The oral tests were carried out both on the
wards and in the outpatient department of the
Bronx Municipal Hospital Center. Intravenous
GTT were performed during cardiac catheterization. Children with a family history of diabetes
were excluded. For the purpose of the study, a
positive family history was taken to mean
diabetes mellitus in a parent or grandparent. The
diagnosis of congestive heart failure (CHF) was
based on the presence of several of the following
clinical signs and symptoms: cardiac enlargetachycardia, decreased peripheral pulsation, growth failure, sweating, tachypnea,
dyspnea with effort, cough, rales, and hepatomegaly. All children with CHF required medical
management for their heart failure and were
treated with digoxin. Whenever possible a glucose
tolerance test was performed before institution of
the drug. The patients were classified clinically
into four groups (table 1):
Group I consisted of 16 children who
represented the control group. Ten were healthy
normal children without significant cardiac involvement. The six children in this group who
were subjected to cardiac catheterization either
had normal hemodynamics or small left-to-right
shunts. All were between the 3rd and 90th
percentile of the Stuart growth chart for weight
and height.
Group II consisted of 9 noncyanotic children
with significant left-to-right shunts but uncomplicated by congestive heart failure. Five were
catheterized. All were between the 10th and 97th
percentile for weight.
ment,
able 1
Catheterization Data
Group
Control:
Normal (2)
VSD (2)
ASD (1)
Coarct (1)
II. Uncomplicated CHD:
VSD (3)
VSD + PI (1)
A-V com (1)
III. CHF:
VSD (6)
A-V com (4)
Coarct + PDA (3)
OP (1)
MI (1)
MI + TI + PDA (1)
SV + TGV (1)
IV. Cyanotic:
PS + VSD (6)
PS + DORV (1)
PS + TGV (1)
PS + TA (1)
Age
(year)
Weight
Capacity
(vol %)
A-V diff
(vol %)
6
3.04
-1.15
37.5
16.2
14.0
-1.l
a
4.02
1.63
51.4
i15.9
17
1.98
-0.78
7.4
3.67
1.19
a5.9
N
Oxygen
Art 02 sat
(%)
Qp/QS
3.1
i .3
-
93.2
-1.4
1.23:1
0. 15
15.9
-0.4
3.2
-0.1
94.9
0.6
1.86:1
i0.31
15.1
0.9
4.1
0.4
88.3
*14.10
-2.2
2.27:1
0.34
9.4
20.3
-2.2
2.7
77.2
0.6:1
-0.3
-3.7
-0.14
(%)
I.
9
Abbreviations: CHD = congenital heart disease; CHF = congestive heart failure; VSD = ventricular septal
defect; ASD = atrial septal defect; coarct = coarctation of the aorta; PI = pulmonary insufficiency; A-V comr =
persistent common atrioventricular canal; PS = pulmonary stenosis; DORV = double-outlet right ventricle; TGV
= transposition of the great vessels; TA = tricuspid atresia; PDA = patent ductus arteriousus; OP = ostium primum;
MI = mitral insufficiency; TI = tricuspid insufficiency; SV = single ventricle; N = number of cases.
Circulation, Volume XLVI, August 1972
GLUCOSE AND INSULIN IN CHD
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Group III consisted of 17 patients with
noncyanotic heart disease and varying degrees of
congestive heart failure. All were below the 10th
percentile for weight.
Group IV consisted of 12 children with various
types of cyanotic heart disease and diminished
pulmonary flow, but without evidence of heart
failure. Nine were catheterized. All were between
the 3rd and 50th percentile for weight.
The children had been eating as usual with no
missed meals during the previous 3 days. Oral
glucose tolerance tests were performed after an
overnight fast. During the first months of life,
fasting was limited to 6-8 hours. In subjects older
than 3 months, fasting was for 10-14 hours.
Glucose was given as a 25% solution, chilled and
flavored. The doses used were 2.5 g/kg for infants
0-18 months; 2.0 g/kg for 18 months to 3 years;
1.75 g/kg for 3-12 years of age; and 1.25 g/kg for
children over the age of 12 years.4
Blood samples were obtained before and Xi, 1,
2, 3, and 4 hours after the glucose administration.
In infants, capillary blood was obtained from the
heel. In the older children a scalp vein needle was
inserted into an antecubital vein for blood
sampling. A very slow infusion of normal saline
was given to prevent clotting in the catheter
throughout the 4-hour period of the test. While no
vigorous physical activity was permitted in these
children during the procedure, a quiet state was
not always achieved. Most infants cried while
capillary blood sampling specimens were obtained.
Blood for glucose and insulin determination
was placed immediately in heparinized, sterile
plastic tubes standing in ice water, then
centrifuged in the cold. The serum collected was
kept frozen at -19°C until analyzed. Concentrations of glucose were measured by the enzyme for
true glucose determination using 0.1 ml of
serum.5 Radioimmunoassay determination of plasma insulin was performed by a modified method
of Yalow and Berson.6 The plasma insulin levels
in samples obtained simultaneously from the
hepatic vein and aorta, or from the pulmonary
artery and the aorta, were measured at the same
time.
Intravenous glucose tolerance tests were done
in 17 children during cardiac catheterization. Six
children were from group I, five children from
group III, and six from group IV. None of these
children received general anesthesia. No premedication was given to infants less than 3 months of
age. Older children received 1 mg/kg of meperidine hydrochloride (Demerol) and 1 mg/kg of
hydroxyzine (Vistaril) by the im route. The inguinal region was infiltrated with 0.5-2 ml of 1.5%
of lidocaine (Xylocaine) with 1:200,000 epinephrine, and the saphenous vein, superficial femoral
Circulation, Volume XLVI, August 1972
335
vein, and artery were isolated. Under fluoroscopic
control an NIH or Elecatheter was introduced via
the femoral vein and advanced to the hepatic
vein. Special care was taken to avoid contamination of hepatic blood by the blood from the
inferior vena cava and to avoid wedging of the
catheter in the hepatic vein, since under these
circumstances mixed arterial and portal venous
blood may bypass the surrounding sinusoidal bed
and enter the catheter.7 A purse-string suture was
placed on the superficial femoral artery, and an
NIH catheter was introduced and advanced
retrograde into the thoracic aorta. Intravenous
glucose tolerance tests were then performed
by injecting a 25% solution of dextrose
over a period of 3 min into a small vinyl catheter
placed in the iliac vein. Children less than 2 years
received 0.75 g/kg of glucose, and older children
received 0.5 g[kg up to a total of 25 g.
Simultaneous hepatic venous and aortic samples
were obtained at control, then at 5, 10, 15, 20,
30, 45, and 60 min following onset of dextrose
infusion. Preparation and determinations of
plasma glucose and insulin were performed as
described above.
A correlation between the changes in plasma
glucose as well as for insulin in relation to time
was derived by employing a coefficient (K) of
glucose assimilation for arterial (Ka) and for
hepatic venous (Kh) blood, respectively. After
glucose mixed through the vascular space during
the first 0-15 min, the glucose concentration falls
in a curve approximating a negative exponential
function.8 Several methods have been proposed
for analyzing this curve. All are based on the
assumption that it is a semilogarithmic curve. We
have used the formula:
(log C - log C) log 10
K
10(1)
10(2)
n
t2 - tl
Where C = concentration of glucose in mg per
100 ml of plasma, t2 = 60 min, t1 = 15 min,
n = number of samples obtained. The value K
obtained was expressed as mg%/min.
The relative impermeability of cell membranes
to glucose throughout the body makes it possible
to calculate the diffusion space of glucose which
prevails during the first few minutes following
rapid intravenous injection of glucose. This is
feasible since the rapid rise in blood glucose is
insufficient in the first few minutes to induce a
hypersecretion of insulin.
By this method the extracellular glucose
compartment in the control patients has been
found to be very similar to that obtained with
sodium thiocyanate or other technics.9 The
extracellular glucose compartment (diffusion
space) (Vg) was determined by using the
HAIT ET AL.
336
,(,
e'
L.
4
C:
-H 1-1
nt
-Ht
N
'
-11
C'
-H
T1.
11
ri
5
e2
0
_/1
6
.
c.
O
O
Q
c c
+V
.
S
"-
i
V6r
~
~
P
._
V
Q
0
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
$.
-T
9000
11+ t
CD ^1
t` 1=
a 1-_ N i
&6OAi
oj
N
""
Tf
0n
.
.
0
h
eq
Cl
-0
0o
,~
C~ 4- 0Od
H.C
s
-qO-adOC
Q0 nt o
t-
-
N
0
~1-
Q
t./
b
-S
0
c]
C~0~-0t=
C~ 0
0
X
M
0
*
0
0
2
*2
t'-t.
:i
00
m
4m
ot-°~
0
b£
~
e
0
.b_
=, o
0
O
Vd < 1'
n
_Cx
0
.5
S
;002 -o
0t-
-
i
A-'
~Rn td
M.~ °
"-
.~-
O
4-
C2
0
o t- o Q
t, CS c) c,
0o~
z b O-H
22
-
-
*
E"
-H
*
*
Q+OC
*
.
C~ O
22
2e
0
E"
C4-
-Q
z
C)
a
0
0
C.
0
0
V)o
o
-
oU
22
o
=0-2
;:
*_
0
W
V
Circulation, Volume XLVI, August 1972
GLUCOSE AND INSULIN IN CHD
337
0
0
MC
o- ~m
.0CO~
c
oo cc
~~~~C
._ 0-
CI~~~
-H
of
cQC.o
m. tcmX<+
~~~~~-'~~~
Cn
cm
N
t- l!
m
s
OO oc CQ i N i en i noOn
i.
Z
~~~~~~C
4-
a
o
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
00
CC
n nm
nc i_0
+
c
1
C ONiN~
-
01C
CO~~~~~~~~~~~~~~~~~~~~~~~~~c
0
02
.
cq
cq
n
+O
P.4
i~
o -X
>
L0
i~t-C~
Cq
1:
0it:
>
°O}
OOmeD
:
L^v
:
D
C e
0
0 0t
0.0..L? n
L
Cirulaion
Voum XLI Auus
19723
CD
-
>
CO
CO
t
m c
c
CO~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C
CO~
CO~~~~~~~~~~~~~~~~~~~~~~~~~~~~C
ut-
Vu
XLVl,
Augu
> ex)
CCcc
s
>
>+t
197
2
CC
HAIT ET AL.
338
and 2 hours. As the blood glucose rose from
the fasting level, plasma insulin increased
steadily to a peak value of 45.6 + 7.2 in the
control at M hour, 31.9 ± 6 in the uncomplicated group at 1 hour, 12± 2.8 in those with
CHF at 1 hour, and 67.9 ± 11.6 ,uU/ml in the
CY group at 1 hour. The net gain of plasma
insulin in the uncomplicated group was lower
than control at 3 hour (P<0.05) (fig. 2).
However, in the CHF group the net gain
insulin level was much lower than control at X
hour (P<0.001), at 1 hour (P<0.001), and
at 2 hours (P<0.005), respectively (fig. 2).
On the other hand, in the CY group the net
gain insulin level was higher than control at X,
1, and 2 hours but was statistically significant
only at 1 hour (P < 0.05). The lowest
recorded peak plasma insulin level was 2.5
,uU/ml in children in CHF, and the highest
level of 134 ,uU/ml was seen in the CY group
in which Q2 was the
formula Vg-(A Q-C
quantity of glucose introduced, A was the
extrapolation of the assimilation line to time 0,
and Co was the initial glucose concentration9 (see
fig. 4).
Results
Oral Glucose Tolerance Tests
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
The mean plasma glucose level was higher
than the control in children with uncomplicated CHD (P < 0.025) at 1 and 2 -hours and
in children with congestive heart failure
(P < 0.005, P < 0.02, and P < 0.005) at X, 1,
and 2 hours, respectively (tables 2, 3; figs. 1,
2). The mean plasma glucose levels in children
with cyanosis (CY) were higher than in the
control at M through 3 hours but were not
statistically significant. The mean plasma
glucose levels in CY were lower than the mean
glucose levels in children with CHF at 2, 1,
I+ S.E.
*-
CONTROL
O----O CYANOTIC
*----O UNCOMPLICATED C.H.D.
C.H.D. WITH C.H.F.
.-....
ORAL G.T.T.
I.V. G.T.T.
300
HEPATIC
r VEIN
, 250
E
LU
O
200
U
D
0 150
LI)
< 100
CL
J
I
1
2
3
TIME (hours)
4
I
I 1
J1l
I
I
I
I .
AL I
0 5 10 15 20 30 45 60 0 5 10 15 20 3045 60
TIME (minutes)
Figure 1
Mean plasma concentrations (+ SE) of glucose in response to oral and i.v. glucose tolerance
control, cyanotic, uncomplicated congenital heart disease, and in children with congestive heart failure.
tests in
Circulation, Volume XLVl, August 1972
GLUCOSE AND INSULIN IN CHD
339
NET GAIN PLASMA INSULIN
I.V G.TT.
ISE
*
CONTROL
0---< CYANOTIC
F
C.H.D
-_3-® CUNCOMPLICATED
.
H.D WITH C.H.F.
AORTA
z
HEPATIC
VEIN
z
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
5
T I ME (hours)
10
15
20
TI ME (minutes)
TIME (minutes)
Figure 2
Mean net gain of plasma concentration (+ sE) of insulin in response to oral and i.v. glucose
tolerance tests in control, cyanotic, uncomplicated congenital heart disease, and in children
with congestive heart failure.
in one case at X and in a second case at 1 hour.
The total net gain of plasma insulin released
throughout the GTT in the various groups was
calculated by planimetry from the total area
under the insulin curve obtained following
oral GTT. Taking the total net gain of plasma
insulin in the control group as 100%, the net
gain in the CHF group was 20%, in the
uncomplicated group 65%, and in the CY
group 142%, respectively (table 4).
Intravenous Glucose Tolerance Tests
The glucose assimilation coefficients of the
arterial (Ka) and the hepatic venous blood
(Kh) were 1.84 and 1.34% in the control
group; 0.94 and 0.86% in the CHF group; and
1.50 and 1.36% in the CY group, respectively
Table 4
Total Net Gain of Plasma Insulin following GTT
Group
Oral
(% normal)
Systemic vein
Normal
Uncomplicated CHD
CHF
Cyanotic
100
65
20
142
Circulation, Volume XLVI, August 1972
Intravenous
(% normal)
Aorta
Hepatic vein
100
100
28
62
47
44
(tables 2-4; figs. 1-3). The mean (+ SE) fasting
plasma insulin concentrations in the aorta and
hepatic vein were 11.5 3 and 12.5 + 4 in the
control group, 16.2 7.7 and 9.2 2.4 ,U/ml
in the CY group, and 24.6 + 8.3 and 20 + 10
,uU/ml in the CHF group, respectively. After
injection of glucose in the control group, a
mean peak plasma insulin concentration of
42.7 + 14.9 ,uU/ml, was found in the aorta
and 53.8 + 11.5 ,uU/ml in the hepatic vein. In
congestive heart failure, mean peak insulin of
34.6 + 9.2 and 39.2 + 3.9 ,U/ml, and in the
cyanotic group 35.7 + 15.4 and 24.6 5.7
,U/ml, were found in the aorta and the
hepatic vein, respectively. After injection of
glucose, the total net gain of arterial insulin
release was only 28% in CHF and 62% in CY as
compared with the control (table 4). The
total net gain of insulin released from the
hepatic vein in the group with CHF was 47%
and in the group with CY 44% of the control
group (table 4). The assimilation coefficients
for arterial (Kia) and hepatic (Kih) insulin in
the control group were 1.75 and 1.68%; in the
group with CHF 0.06 and 0.86%; and in the
CY group 1.63 and 1.59%/min, respectively. In
the control group, with the exception of
HAIT ET AL.
340
MEAN (±S.E.) INSULIN LEVEL IN
CYANOTIC,C.H.D. WITH C.H.F.,
AND CONTROL GROUPS (I.VGTT.)
55
Cl
50
Hepatic Vein
A---4
Cyanotic
Aorta
@--
Hepatic
El Control
Aorta
c C.H.D.withb---
45
C.H.F.
Hepatic Vein
Aorta
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
I S.E
±
40 _
rTI
351=
30
C.L:
z
D
254
(I)
z
'<20i
LI)
0
L- 15.
j
1
I
~1
10
11
5
L-
5
-1
10
15
20
MINUTES
30
45
60
Figure 3
Simultaneous mean (-+- SE) insulin levels from the aorta and hepatic vein following i.v. GTT
in control, cyanotic, and children with congestive heart failure.
samples taken at 20 min, the mean plasma
insulin levels in the hepatic vein were always
higher than those in the systemic artery. A
reverse relationship, however, was found in
the CY group. In congestive heart failure a
normal relationship between plasma insulin
levels in the aorta and hepatic vein was
observed during the 5-, 10-, and 15-min
samples, and a reverse relationship in the
fasting, 30-, 45-, and at the 60-min samples.
The extracellular glucose compartnent represented 20% of the body weight in the
control, 27% in the CY, and 40% in the group
with CHF (fig. 4). The results found in the
control group were similar to those obtained
by other methods.9
Circulation, Volume XLVI, August 1972
GLUCOSE AND INSULIN IN CHD
341
RELATIONSHIP BETWEEN BODY WEIGHT (Kg) AND
GLUCOSE DIFFUSION SPACE
1616 ---0 CONTROL
CYANOTIC
W
14
W
o- o-- C .H.F.
'00w40%
12
Q
-C
VgL=
0
10
1
_J
~~--27%
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
8
20%
6
4
2
5
10
15
20
25
Kg BODY WEIGHT
30
35
40
Figure 4
Relationship between body weight (kg) and glucose diffusion space (extracellular compartment)
in control, cyanotic, and congestive heart failure. Note the twofold increase in this space in
children with congestive heart failure.
Discussion
Several abnormalities of glucose and insulin
metabolism were observed in this study.
Infants in congestive heart failure had abnormally high glucose levels following oral GTT
(fig. 1). Plasma glucose levels up to 310
mg/100 ml were observed. The result of our
studies indicated that suppression of insulin
release served as a contributing mechanism
for this abnormality (fig. 2). In the group of
uncomplicated CHD with moderate left-toright shunts, slightly higher than normal levels
of glucose and significantly lower than normal
levels of insulin were observed (figs. 1, 2, 6).
These metabolic alterations could therefore be
regarded either as a more sensitive index of
Ciculation, Volume XLVI, August 1972
heart failure than the usual clinical criteria or
as the direct result of left-to-right shunt. Our
results from four control patients as well as
other preliminary findings10 suggest that in
addition to the known insulin-deactivating
systems in the liver, kidney, muscle, and other
tissues, a significant amount of insulin is
cleared by the lung. The reduction of insulin
across the lung in the control group was
estimated at about 25% (fig. 5). Furthermore,
we observed a significant correlation between
plasma insulin levels and pulmonary blood
flow (fig. 6), and this may provide a partial
explanation for the abnormally low arterial
insulin levels in patients with moderate and
large left-to-right shunts as a result of
HAIT ET AL.
342
SIMULTANEOUS INSULIN LEVELS
FOLLOWING I.V. GTT
A.M. 9 YEARS
* Aorta
O Pulmonary Artery
:D
zE
:D
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
z
LI)
C-J
L
0
1 1
L
1JYI,1
5 10 15 20 30 45 60
MINUTES
Figure 5
Plasma insulin obtained and arrayed simultaneously
from the pulmonary artery and aorta in a 9-year-old
patient with insignificant cardiac lesion.
excessive clearance by the lung. Similarly, in
the CY group, the elevation of arterial insulin
level is due to the underclearance of insulin by
the lung because of reduced pulmonary blood
flow (fig. 6). However, the plasma insulin
levels in the aortic and hepatic venous blood
following i.v. GTT were too small to allow a
similar conclusion.
In patients with CHF the marked reduction
in insulin level observed in the hepatic vein
suggests therefore that mechanisms other than
lung clearance are involved. Possible mechanisms may include reduction in the rate of
incorporation of amino acid into insulin,
excessive destruction of insulin by the liver,
suppression of pancreatic release of insulin, or
some combination of these factors. The rate of
incorporation of labeled amino acid into
insulin in the islet tissue was shown to be
dependent on the amino acid concentration
and the presence of oxygen.11 In the present
study, the group with CHF had a mean
extracellular compartment twice as large as
that in the control group (fig. 4). Therefore, it
is conceivable that in cardiac failure the
congestion of the viscera and the resulting
tissue hypoxia may indeed be responsible for
the decreased rate of insulin production.
Normally, insulin is secreted into the portal
system and carried to the liver where much of
it is destroyed by insulinase before coming
into contact with the peripheral tissues. The
possibility exists that in CHF, because of the
reduced rate of blood flow12 through the liver,
a greater degree of insulin destruction by
insulinase may occur.
Norepinephrine, which is found in excess in
the plasma and urine of patients with
congestive heart failure,13 may inhibit the
release of insulin through its specific alphaadrenergic receptor effect.2' 14, 1; The important inhibitory control of insulin by the
sympathetic nervous system had been demonstrated in cardiogenic shock,"' hypothermia,"7
pheochromocytoma,18 and most recently in
congestive heart failure. 19
The possibility that protein-calorie deficiency alone may be an important factor in
suppression of insulin release20-25 must also be
considered. Such observations have been made
in children suffering from kwashiorkor,22 in
severely malnourished African adults,23, 24 and
in experimental animals.20 25 Postmortem evidence for deficiency of insulin secretion was
demonstrated by the reduced cell size and degeneration of the beta cells. In addition, a
low beta/alpha cell ratio in the islets of Langerhans have been found in man with malignant malnutrition26 and in protein-calorie deficient pigs.27
The reason for the significantly higher levels
of insulin and glucose in the peripheral
circulation following oral glucose load in
children with cyanotic heart disease is unclear.
It suggests an antagonistic action to the
utilization of blood glucose by the peripheral
tissues. The elevated plasma insulin levels in
Circulation, Volume XLVI, August 1972
GLUCOSE AND INSULIN IN CHD
343
PLASMA INSULIN LEVELS AS FUNCTION OF Qp/Qs
140
0
* CONTROL
O CYANOTIC
* UNCOMPLICATED C.H.D.
o C.H.D. WITH C.H.F.
120_
± S.E.
100
0
:D
80 I
0O
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Z 60
-
a
0
<
40 I _
-J
a-
20k
0
0
0
-
l
0
l
0 .5 1
o
0
0
2
0
l
l
l
3
4
5
6
Qp/Qs
Figure 6
Mean plasma insulin levels as a function of the pulmonary/systemic flow, 1 hour following
oral glucose tolerance tests. The mean point representing the cyanotic group is significantly
higher (P < 0.05), and that which represents the CHF group is significantly lower (P < 0.001)
than the mean control.
these children also indicate excessive production of insulin. Humoral antagonists to insulin
such as growth hormone and adrenocortical
hormones may also result in elevated levels of
circulating insulin.28 The levels of such
antagonists may conceivably be increased in
the blood of cyanotic children. Another
possibility is the production of an altered,
ineffective insulin molecule which reacts with
antibodies as measured immunochemically,
but the function of which is impaired to the
extent that larger concentrations are required
for effective activity. In the light of our
findings of significant insulin clearance by the
lung, the increased arterial insulin levels
Circulation, Volume XLVI, August 1972
observed in the cyanotic group following oral
GTT could reflect the amount of insulin which
had bypassed the lung (fig. 6).
Contrary to the findings of increased insulin
release following oral GTT, the intravenous
GTT studies of the cyanotic group showed
abnormally low levels of insulin in the hepatic
vein (fig. 2). These low levels were similar to
those seen in the group with CHF (fig. 2). A
probable common cause for the decrease in
insulin release following i.v. GTT in both
groups is hypoxia. Arterial oxygen tension is
tolerably reduced in severe cyanotic heart
disease and in infants with CHF.29 However,
when adaptive mechanisms are exhausted,30-32
HAIT ET AL.
3'44
tissue oxygen tension is lowered to critical
levels. The existence of a total-body oxygen
deficit in CHF is particularly acute in the
splanchnic region. Oxygen extraction in that
region at rest often is already near maximal,
and a decrease in splanchnic blood flow with
even mild exercise may result in frank
ischemia of this area.33 Pancreatic hypoxia in
CHF and in severe cyanotic CHD must
therefore exert a profound effect on the
function of this organ, of which the secretory
activity may be flow dependent34' 35 and insulin biosynthesis by the pancreas is oxygen deDownloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
pendent.1'
Intriguing, however, is the fact that under
condition of tissue hypoxia, in the CY group
insulin release could be abnormally increased
following oral GTT and decreased during i.v.
GTT (fig. 2), and in CHF insulin release is
abnormally decreased in both oral and i.v.
GTT. An important difference between oral
and intravenous glucose tolerance tests in
normal individuals is due at least in part to the
greater release of immunologically active
insulin from the pancreas in the oral test. A
gastrointestinal factor which was shown to be
glucagon is also released from the gastrointestinal wall during glucose absorption and
stimulates the release of insulin from the
pancreatic islet cells.56 It is therefore possible
that a glucagonlike substance is produced in
excess in the GI tract of children with CY,
resulting in a greater stimulation of insulin
release. An analogous mechanism was postulalated for nondiabetic obese children who have
abnormally high levels of insulin, normal
glucose levels, and recently shown to have
high levels of glucagon.37
Fasting blood sugar38' 39 and glucose levels
following oral GTT40 in individuals with heart
disease have been reported previously. Very
low fasting blood sugars were observed in
three elderly patients with congestive heart
failure.38 These patients, however, were in the
terminal phase of their disease, and blood was
obtained just before or during coma. All
patients showed evidence of liver disease and
unequivocal diffuse hepatic damage on histologic examination. Similar values were report-
ed just before death in a group of newborn
infants in congestive heart failure.39 Unlike
the usual child in congestive heart failure, this
group consisted of newborn infants with the
hypoplastic left ventricle syndrome, known to
die within the first few days of life, probably
because of inadequate perfusion of blood to
the myocardium and to other vital organs.
Histologic studies of such infants4' also
revealed hepatic necrosis which may have
contributed to the development of hypoglycemia.
In CHF, fasting arterial insulin levels were
higher, reached lower peaks than controls, and
had a markedly reduced assimilation coefficient (Ki = 0.06) (fig. 3). At 30 min when the
hepatic venous levels of insulin dropped to
fasting level, arterial insulin levels remained
unchanged, suggesting poor insulin clearance
by the liver, kidney, muscle, and other tissues
(fig. 3). In view of the poor insulin degradation, it is possible that the use of exogenous
insulin as a therapeutic measure may not
confer a beneficial effect. The findings observed in congestive heart failure indicated
that insulin release as well as its degradation is
altered and explains their markedly suppressed arterial glucose assimilation coefficient
(Ka = 0.96). Such a state of catabolism
superimposed upon a deficient diet could
contribute to the growth retardation of
children with CHF.
References
1. HAIT G, GRUSKIN AB, PAULSEN EP: Impaired
2.
3.
4.
5.
glucose tolerance and insulin suppression in
children with congestive heart failure. Circulation 36 (suppl II): II-129, 1967
HAIT G, GRUSKIN AB, PAULSENT EP: Insulin
suppression in children with congestive heart
failure. Pediatrics. In press
HArr G, CoRPus M, LAMARRE FR: Abnonnal
insulin release in cyanotic heart disease. Pediat
Res 2: 287, 1968
CORNBLATH M, SCHWARTZ R: Disorders of
carbohydrate metabolism in infancy. In Major
Problems of Clinical Pediatrics, edited by
Schaffer-Consult AJ. Philadelphia, W. B.
Saunders Co., 1966, vol 3
HUGGET AG, NIXON DA: Enzymatic determinations of blood glucose. Biochem J 66: 12,
1957
Circulation, Volume XLVI, August 1972
GLUCOSE AND INSULIN IN CHD
6. HERBERT V, LAU K, GOTTLIEB CW, BLEICHER
SJ: Coated charcoal immunoassay of insulin. J
Clin Endocrinol 25: 1375, 1965
Clinical aspect of the hepatic
circulation. Gastroenterology 48: 790, 1965
7. LEvY CM:
8. FRANCKSON JRM, OoMs HA, CONRAD BV,
BASTENIC PA: Physiologic significance of the
intravenous glucose tolerance test. Metabolism
11: 482, 1962
9. FRANCKSON JRM, CONRAD BV, BASTENIC PA:
Measurement of the free glucose diffusion
space in man by the rapid intravenous glucose
tolerance test. Acta Endocr (Kobenhavn) 32:
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
463, 1959
10. RUBENSTEIN AH, ZWI S, MILLER K: Insulin and
the lung. Diabetologia 4: 235, 1968
11. BAUER CE, LAZAROW A: Studies on the isolated
islet tissue of fish: IV. In vitro incorporation of
C14 and H3 labeled amino acids into Goosefish
islet tissue proteins. Biol Bull 121: 425, 1961
12. DONALD KW: Hemodynamics in chronic congestive heart failure. J Chronic Dis 9: 476,
1959
13. CHIDSEY CA, BRAUNWALD E, MoRRow AG:
Catecholamine excretion and cardiac stores of
norepinephrine in congestive heart failure.
Amer J Med 39: 442, 1965
14. PORTE D JR, GRABER A, KUZUYA T, WILLIAMS
RH: The effect of epinephrine on immunoreactive insulin levels in man. J Clin Invest 45:
228, 1966
15. PORTE D JR, WILLIAMS RH: Inhibition of
insulin release by norepinephrine in man.
Science 152: 1248, 1966
16. ALLISON SP, CHAMBERLAIN MJ, HINTON P:
Intravenous glucose tolerance, insulin, glucose
and free fatty acid levels after myocardial
infarction. Brit Med J 4: 776, 1969
17. BAUM D, PORTE D JR: Adrenergic regulation of
the inhibition of insulin release in hypothermia.
Diabetes 17: 298, 1968
18. WILBUR JF, TURTLE JR, CRANE NA: Inhibition
of insulin secretion by a phaeochromocytoma.
Lancet 2: 733, 1966
19. MAJID PA, SAXTON C, DYKES JRW, GALVIN MC,
TAYLOR SH: Autonomic control of insulin and
the treatment of heart failure. Brit Med J 4:
328, 1970
20. HEARD CRC, PLATT BS: The nutritional
aetiology of fatty liver. In Metabolism and
Physiological Significance of Lipids, edited by
Dawson RMC, Rhodes ND. London, John
Wiley and Sons, Ltd., 1964, p 435
21. HEARD CRC, TURNER MR, PLATT BS: Diabetic
like changes in carbohydrate metabolism in
dogs fed diets of suboptimal protein value.
Proc Nutr Soc 23: 6, 1964
Circulation, Volume XLVI, August 1972
345
22. SLONE D, TAITZ LS, GILCHRIST GS: Aspect of
carbohydrate metabolism in kwashiorkor. Brit
Med J 1: 32, 1961
23. HOLMES EG, TROWELL HC: Formation of
hepatic glycogen in normal Africans and in
those suffering from malignant malnutrition.
Lancet 1: 395, 1948
24. GILLMAN J, CILLMAN T: Perspectives in Human
Malnutrition. New York, Grune & Stratton,
Inc., 1951
25. HEARD CRC: Effects of severe protein-caloric
deficiency on the endocrine control of carbohydrate metabolism. Diabetes 15: 78, 1966
26. STEWARD RJC, HEARD CRC: Pancreatic islet cells
and blood-sugar regulation in pigs maintained
on low-protein diets. Proc Nutr Soc 18: 10,
1959
27. STEWARD RJC, HEARD CRC: Protein malnutrition
and disorders of the endocrine glands morphological changes. Acta Endocr (Kobenhavn)
(suppl) 51: 1313, 1960
28. PERLEY M, KIPNIs DM: Effect of glucocorticoid
on plasma insulin. New Eng J Med 274: 1237,
1966
29. TALNER NS, SANYAL SK, HALLORAN KH,
GARDNER TH, ORDWAY NK: Congestive heart
failure in infancy: I. Abnormalities in blood
gases and acid-base equilibrium. Pediatrics
35: 20, 1965
30. KORNER PI: Circulatory adaptations in hypoxia.
Physiol Rev 39: 687, 1959
31. BURCHELL HB, TAYLOR BE, KNUTSON JRB,
WOOD EH: Circulatory adjustments to the
hypoxemia of congenital heart disease of the
cyanotic type. Circulation 1: 404, 1950
32. HUssON G, OTIS AB: Adaptive value of
respiratory adjustments to shunt hypoxia and to
altitude hypoxia. J Clin Invest 39: 270,
1957
33. BIsHoP JM, DONALD KW, WADE OL: Changes
in oxygen content of hepatic venous blood
during exercise in patients with rheumatic
heart disease. J Clin Invest 34: 1114, 1955
34. BABKIN BP: The influence of the blood supply
on pancreatic secretion. J Physiol (London)
59: 152, 1924
35. BARLOW OW: A comparison of the blood
pressure, kidney volume and the pancreatic
secretory response following the vein administration of various secretin preparations. Amer
J Physiol 81: 182, 1927
36. UNGER RH, OHNEDA A, VALVERDE I, EISENTRAUr
AM, EXTON J, HARRIS V, JONES AM, THOMPSON G: Characterization of the responses of
circulating glucagon-like immunoreactivity to
intraduodenal and intravenous administration
of glucose. J Clin Invest 47: 48, 1968
346
37. PAULSEN EP, LAWRENCE AM: Glucagon hypersecretion in obese children. Lancet 2: 110,
1968
38. MELLINKOFF SM, TUMULTY PA: Hepatic hypoglycemia. New Eng J Med 247: 745, 1952
39. BENZING C III, SCHUBERT W, HUG G, KAPLAN S:
Simultaneous hypoglycemia and acute congestive heart failure. Circulation 40: 209, 1969
HAIT ET AL.
40. ETTINGER PO, OLDEWURTEL HA, WEISSE AB,
REGAN TJ: Diminished glucose tolerance and
immunoreactive insulin response in patients
with nonischemic cardiac disease. Circulation
38: 559, 1968
41. WEINBERG AG, BOLANDE RP: The liver in
congenital heart disease. Amer J Dis Child
119: 390, 1970
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Circulaion, Volume XLVI, August 1972
Alteration of Glucose and Insulin Metabolism in Congenital Heart Disease
GERSHON HAIT, MARINA CORPUS, FRANCOIS R. LAMARRE,
SHANG-HSIEN YUAN, JINDRICH KYPSON and GRACE CHENG
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Circulation. 1972;46:333-346
doi: 10.1161/01.CIR.46.2.333
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX
75231
Copyright © 1972 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/46/2/333
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles
originally published in Circulation can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column
of the Web page under Services. Further information about this process is available in the
Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/