0021-972X/01/$03.00/0
The Journal of Clinical Endocrinology & Metabolism
Copyright © 2001 by The Endocrine Society
Vol. 86, No. 3
Printed in U.S.A.
Effect of Protein Ingestion on the Glucose Appearance
Rate in People with Type 2 Diabetes*
M. C. GANNON, J. A. NUTTALL†, G. DAMBERG‡, V. GUPTA‡,
AND
F. Q. NUTTALL
Section of Endocrinology, Metabolism, and Nutrition (M.C.G., F.Q.N.), Metabolic Research Laboratory,
Veterans Affairs Medical Center, Minneapolis, and Departments of Medicine (M.C.G., F.Q.N.) and
Food Science and Nutrition (M.C.G.), University of Minnesota, Minneapolis, Minnesota 55417
ABSTRACT
Amino acids derived from ingested protein are potential substrates
for gluconeogenesis. However, several laboratories have reported that
protein ingestion does not result in an increase in the circulating
glucose concentration in people with or without type 2 diabetes. The
reason for this has remained unclear. In people without diabetes it
seems to be due to less glucose being produced and entering the
circulation than the calculated theoretical amount. Therefore, we
were interested in determining whether this also was the case in people
with type 2 diabetes. Ten male subjects with untreated type 2 diabetes
were given, in random sequence, 50 g protein in the form of very lean beef
or only water at 0800 h and studied over the subsequent 8 h.
Protein ingestion resulted in an increase in circulating insulin,
A
LL OF THE amino acids commonly found in proteins
can, at least in part, be converted into glucose, with
the exception of leucine. Indeed, conversion of amino acids
derived from either endogenous or exogenous proteins is the
major source of new glucose formation. Lactate and pyruvate
are converted to glucose but they also are derived from
glucose; thus, they do not result in net glucose formation.
They are a means of shuttling carbohydrate-derived energy
sources to and from the various organs in the body. Glycerol
derived from triacylglycerol metabolism also may be used
for net glucose production, but quantitatively it is of only
minor importance (1).
It has been known for many years that 50 – 80 g glucose can
be derived from 100 g ingested protein (2). The amount of
potential glucose produced depends on the amino acid composition. For beef skeletal muscle protein, this has been calculated to be 56 g glucose/100 g protein (2). However, it also
has been demonstrated that ingestion of proteins results in
little or no increase in circulating glucose concentration in
nondiabetic people or in people with type 2 diabetes mellitus
(3–10). The reason for this remains unclear. In normal young
men, we previously have reported that the lack of a rise in
glucose is due to the production of less glucose than pre-
Received September 28, 2000. Revision received November 6, 2000.
Accepted November 9, 2000.
Address correspondence and requests for reprints to: Mary C. Gannon, Ph.D., Director, Metabolic Research Laboratory (111G), Veterans
Affairs Medical Center, One Veterans Drive, Minneapolis, Minnesota
55417. E-mail: [email protected].
* Supported by a grant from the Minnesota Beef Council and Merit
Review Research Funds from the Department of Veterans Affairs.
† Medical student.
‡ Fellow in Endocrinology.
C-peptide, glucagon, ␣ amino and urea nitrogen, and triglycerides; a
decrease in nonesterified fatty acids; and a modest increase in respiratory quotient.
The total amount of protein deaminated and the amino groups
incorporated into urea was calculated to be ⬃20 –23 g. The net amount
of glucose estimated to be produced, based on the quantity of amino
acids deaminated, was ⬃11–13 g. However, the amount of glucose
appearing in the circulation was only ⬃2 g. The peripheral plasma
glucose concentration decreased by ⬃1 mM after ingestion of either
protein or water, confirming that ingested protein does not result in
a net increase in glucose concentration, and results in only a modest
increase in the rate of glucose disappearance. (J Clin Endocrinol
Metab 86: 1040 –1047, 2001)
dicted (9). Subsequently, we wanted to determine whether
this also was the case in people with type 2 diabetes. The
present data indicate that even less glucose is produced in
these subjects. Part of these data have been presented previously in abstract form (11).
Materials and Methods
Ten male volunteers with untreated type 2 (noninsulin dependent)
diabetes were studied in the Special Diagnostic and Treatment Unit
(SDTU). Volunteers had a mean age of 65 yr, with a range of 41– 81 yr.
Mean body mass index was 30, with a range of 24 – 44 (Table 1). The
volunteers gave written informed consent, and the study was approved
by the Veterans Affairs Medical Center, Minneapolis, and the University
of Minnesota Committees on Human Subjects. All volunteers had ingested a diet containing at least 200 g carbohydrate/day, as well as their
usual food energy intake for the 3 days before the study.
Volunteers were admitted to the SDTU on the evening before the
study and were given a standardized meal at 1700 h. Subsequently, only
water ad libitum was allowed. The following morning at 0300 h, two
indwelling venous catheters were inserted, one in the antecubital vein
and the other in the dorsum of the contralateral hand. The catheters were
kept patent with a slow infusion of 0.9% saline. An infusion of tritiumlabeled glucose (d-[3⫺3H]-glucose) and 14C bicarbonate (NEN Life Science Products, Boston, MA) was started at 0300 h at a constant rate of
⬃0.11 ␮Ci glucose/min and ⬃0.52 ␮Ci 14C HCO3/min, respectively.
This was continued until 1600 h.
At 0800 h, either 50 g beef protein (tenderloin, ⬍5% fat, 236 g raw
weight), or water alone was given in random order. This was consumed
within 15 min. During the subsequent 8-h period, the volunteers were
allowed to drink water ad libitum. Arterialized blood samples were
drawn from the hand hourly from 0300 – 0700 h and then every 15 min
until 0800 h. Subsequently, blood samples were drawn at 15-min intervals for 90 min, then at 30-min intervals for 150 min, and every hour for
the final 4 h of the study. Subjects were studied over an 8-h period to
assure complete absorption of the 50 g protein. The time interval between studies was at least 2 weeks to allow adequate time for the
isotopes to be cleared from the body, but less than 3 months.
Plasma glucose was determined by a glucose oxidase method using
1040
DIETARY PROTEIN AND TYPE 2 DIABETES
1041
TABLE 1. Patient characteristics
Patient
Age
(yr)
Height
(m)
Weight
(kg)
%IBW
BMI
(kg/m2)
tGHb
Duration Db
(months)
1
2
3
41
57
67
1.75
1.87
1.70
136
95
80
193
117
120
44.3
27.1
27.6
9.0
8.6
6.9
2
2
29
4
70
1.83
88
115
26.3
8.1
21
5
6
54
81
1.70
1.73
92
80
140
117
31.9
26.8
6.8
6.9
79
9
7
54
1.75
121
172
39.4
6.8
9
8
70
1.93
106
128
28.5
9.1
1
9
10
77
76
1.85
1.66
81
66
103
106
23.8
24.0
8.3
6.9
24
16
65
4.0
1.78
0.03
94.5
6.7
131
9.3
30.0
2.1
7.7
0.3
24
7.7
Mean
SEM
Glucose
Concomitant diseases
Bipolar psychosis
None
Hypertension; hypercholesterolemia; coronary heart disease
Hypertension; coronary heart disease; atrial fibrillation
Coronary heart disease
Hypertension; hypothyroidism
(currently euthyroid on replacement)
Hypertension; hypothyroidism
(currently euthyroid on replacement)
Hypertension; hypothyroidism
(currently euthyroid on replacement)
Peptic ulcer disease
Hypercholesterolemia; hypertension; osteoarthritis
Insulin
Visit 1
Visit 2
Visit 1
Visit 2
94
157
136
102
140
143
59
35
14
79
24
24
141
116
10
24
96
104
99
116
13
17
24
20
109
109
52
66
129
149
43
47
130
99
137
87
12
5
13
4
The tGHb reference range is 4.5– 6.5. IBW, Ideal body weight; BMI, body mass index.
a Beckman glucose analyzer with an O2 electrode (Beckman Coulter,
Inc., Fullerton, CA). Serum immunoreactive insulin was measured by a
standard double-antibody RIA method using kits produced by Endotech
(Louisville, KY). Glucagon was determined by RIA using 30-K antiserum purchased from Health Sciences Center (Dallas, TX). Serum nonesterified fatty acids (NEFAs) were determined by the colorimetric assay
of Duncombe (12). Triglycerides were determined using an EktaChem
analyzer (Eastman Kodak Co., Rochester, NY). Plasma lactate was determined by the method of Hohorst using lactate dehydrogenase (13).
Plasma and urine urea nitrogen were determined using the Ortho Vitros
950 instrument. The total ␣-amino nitrogen was determined by the
method of Goodwin (14). ␣ Amino nitrogen concentration is a measure
of the total concentration of amino acids in serum.
The amount of protein oxidized was determined by quantifying the
urine urea nitrogen excreted over the 8 h of the study in association with
the change in amount of urea nitrogen retained endogenously. The latter
was calculated by determining the change in plasma urea nitrogen
concentration at 8 h and correcting for plasma water by dividing by 0.94.
In this calculation it is assumed that there is a relatively rapid and
complete equilibration of urea in total body water. Total body water as
a percentage of body weight was calculated using the equation of
Watson et al. (15). The overall assumption is that a change in plasma urea
concentration is indicative of a corresponding change in total body water
urea concentration. As indicated later, this may or may not be entirely
correct. The sum of total urea nitrogen in urine and body water was
divided by 0.86 to account for 14% lost to metabolism in the gut (16).
A respiratory quotient (RQ) was determined for periods of 10 min or
more at 0730 and 0830 h, then at half-hour intervals until 1200 h and
hourly intervals until 1600 h, by measuring the O2 consumed and the
CO2 produced using a Deltatrac instrument. A protein RQ of 0.83 was
used based on the RQ of individual amino acids (17) and the amino acid
composition of beef muscle (18). The O2 consumed and CO2 produced
as a result of carbohydrate and fat oxidation were calculated using the
nomogram produced by Lusk (19).
Rates of peripheral glucose appearance (Ra) were calculated using the
nonsteady-state equations of Steele et al. (20) as modified by deBodo et
al. (21). To correct for noninstantaneous mixing of glucose, a correction
factor of Vp ⫽ 0.65 was used (22). The volume of distribution for glucose
was considered to be 26% of body weight. In humans the volume has
been variously reported to be from 24 –37% (23). Volume distributions
over this range have little effect on the final Ra (see Ref. 24). The fasting
baseline data used in presentation of the results represents the mean of
the data obtained from the four blood samples obtained from 0700 to
0800 h for each individual. The steady-state equation was used for
calculation of Ra over the 0700- to 0800-h baseline period. Quantitation
of the subsequent 8-h integrated glucose Ra was determined as the area
above or below the mean of the fasting Ra over the 0700- to 0800-h time
frame. The area was calculated by the trapezoid rule (25) using a program developed for this purpose in our laboratory (26). The glucose
disappearance rate (Rd) was calculated using the equation Rd ⫽ (Ra) ⫺
(rate of change of the glucose pool) (27).
The rate of gluconeogenesis was estimated by determining the incorporation of 14C from infused 14C-NaHCO3 into glucose as described
by McMahon et al. (28). There is debate regarding how accurately this
method, as well as other tracer methods, reflects the actual gluconeogenic rate. The method used generally is considered to underestimate
the gluconeogenic rate. Potential problems in interpretation of the data
using this and other methods have been reviewed previously by others
(28 –32). However, because each subject served as his or her own control,
the data can be used for comparative purposes regardless of the potential
limitations of the method.
Statistics were done by multivariate ANOVA using the Minitab computer program, followed by post hoc t tests when data were significantly
different. Because of the large number of comparisons at individual time
points, and the concern over type I errors, the criterion of significance
was set at P of 0.01 or less.
Results
As expected, when the subjects ingested only water (fasting controls) there was a gradual decrease in serum glucose
concentration over the 8 h of the study (33). When the subjects ingested 50 g beef protein there was a small initial and
transient increase in glucose, but by 2.5 h the glucose concentration had decreased and continued to decrease until the
end of the study. Over the last 5.5 h, the concentration was
slightly less than when only water was ingested (Fig. 1).
The insulin concentration decreased only slightly over
time when water was ingested. When beef was ingested
there was a prompt rise in mean insulin concentration (⬃3fold). In contrast to the prompt return of the glucose concentration to the fasting control value at 2.5 h, the insulin
concentration was still at a maximum at that time. The insulin
concentration did not return to a fasting value until 7 h after
1042
GANNON ET AL.
FIG. 1. Plasma glucose response. The mean glucose concentration ⫾
SEM after ingestion of 50 g protein (solid line) or water only (broken
line) in 10 males with untreated type 2 diabetes (top). Bottom, The
change from baseline. The baseline is indicated by a light horizontal
broken line in the bottom panel.
the meal (Fig. 2). The C-peptide curve (Table 2) was similar
to that of the insulin curve (Fig. 2).
The glucagon concentration increased after protein ingestion, as expected. It remained unchanged when only water
was ingested (Fig. 3).
The glucose appearance rate slowly decreased both following water ingestion and following protein ingestion (Fig.
4). Beginning after the first half hour, the glucose appearance
rate tended to be slightly higher after protein ingestion.
In 8 of the 10 subjects the incorporation of 14C from 14CHCO3 into glucose was calculated as an estimate of the gluconeogenic rate (Fig. 5). There was a gradual increase over
the 8-h period of the study, both when the subjects were
fasting (control) and when they received protein. The rate
was modestly higher after protein ingestion, but it was not
statistically significant.
The ␣-amino nitrogen concentration decreased slightly after only water was ingested (Fig. 6). After protein ingestion,
it increased rapidly indicating a rapid digestion of the protein
JCE & M • 2001
Vol. 86 • No. 3
FIG. 2. Serum insulin response. The mean insulin concentration ⫾
SEM after ingestion of 50 g protein (solid line) or water only (broken
line) in 10 males with untreated type 2 diabetes (top). Bottom, The
change from baseline. The baseline is indicated by a light horizontal
broken line in the bottom panel. *, Statistically different from water
only (P ⱕ 0.01); †, statistically different from the 0800-h time point
(P ⱕ 0.01).
and absorption of the resulting amino acids. It returned to
near the control value by the 7th h, indicating that digestion,
absorption, and metabolism were largely completed.
After a short delay there was an increase in plasma urea
nitrogen following ingestion of beef. The increase persisted
to the end of the study (Fig. 7). Interestingly, the uric acid
concentration also increased after beef ingestion (Table 2).
The urea nitrogen excreted in the urine over the 8-h period
was 3.4 ⫾ 0.3 g and 4.4 ⫾ 0.3 g for water and protein ingestion, respectively (data not shown).
The NEFA concentration decreased after protein ingestion
(Table 2), and this was largely the mirror image of the insulin
concentration (Fig. 2).
The RQ decreased modestly when only water was ingested. It increased modestly following the ingestion of protein. The curve was similar to the insulin curve and was the
mirror image of the NEFA concentration, indicating a sub-
DIETARY PROTEIN AND TYPE 2 DIABETES
1043
TABLE 2. Metabolite concentrations
Hour
0800
0815
0830
0845
0900
0915
0930
1000
1030
1100
1130
1200
1300
1400
1500
1600
a
b
C-peptide (nM)
Water
Beef
0.78 ⫾ 0.12
0.90 ⫾ 0.11
0.78 ⫾ 0.12
0.96 ⫾ 0.13a
0.82 ⫾ 0.12
1.02 ⫾ 0.12b
0.82 ⫾ 0.13
1.18 ⫾ 0.16a,b
0.81 ⫾ 0.14
1.36 ⫾ 0.19a,b
0.79 ⫾ 0.13
1.46 ⫾ 0.24a,b
0.73 ⫾ 0.11
1.49 ⫾ 0.12a,b
0.75 ⫾ 0.12
1.44 ⫾ 0.20a,b
0.80 ⫾ 0.12
1.43 ⫾ 0.22a,b
0.81 ⫾ 0.13
1.46 ⫾ 0.23a,b
0.80 ⫾ 0.13
1.41 ⫾ 0.22a,b
0.78 ⫾ 0.12
1.28 ⫾ 0.18a,b
0.79 ⫾ 0.14
1.14 ⫾ 0.19
0.76 ⫾ 0.13
1.00 ⫾ 0.16
0.74 ⫾ 0.14
0.90 ⫾ 0.13
0.76 ⫾ 0.17
0.86 ⫾ 0.13
F(1294) ⫽ 191.75, P ⬍ 0.000
NEFA (␮M)
Uric acid (mM)
Water
Beef
Water
0.4 ⫾ 0.02
0.4 ⫾ 0.03
0.4 ⫾ 0.02
0.4 ⫾ 0.03
0.4 ⫾ 0.02
0.4 ⫾ 0.03
0.4 ⫾ 0.02
0.5 ⫾ 0.03b
0.4 ⫾ 0.02
0.5 ⫾ 0.04b
0.4 ⫾ 0.02
0.5 ⫾ 0.04b
0.4 ⫾ 0.02
0.5 ⫾ 0.04b
0.4 ⫾ 0.02
0.5 ⫾ 0.04b
0.4 ⫾ 0.02
0.5 ⫾ 0.04b
0.4 ⫾ 0.02
0.5 ⫾ 0.04b
0.4 ⫾ 0.02
0.5 ⫾ 0.04b
0.4 ⫾ 0.02
0.4 ⫾ 0.03
0.4 ⫾ 0.02
0.4 ⫾ 0.03
0.4 ⫾ 0.02
0.4 ⫾ 0.03
0.4 ⫾ 0.02
0.4 ⫾ 0.03
0.4 ⫾ 0.02
0.4 ⫾ 0.03
F(1294) ⫽ 28.54, P ⬍ 0.000
Triacylglycerol (mM)
Beef
638 ⫾ 56
650 ⫾ 45
691 ⫾ 67
656 ⫾ 70
693 ⫾ 68
699 ⫾ 85
631 ⫾ 57
627 ⫾ 62b
595 ⫾ 52
555 ⫾ 57b
643 ⫾ 72
514 ⫾ 55b
665 ⫾ 90
471 ⫾ 53b
707 ⫾ 101
419 ⫾ 53b
749 ⫾ 87
415 ⫾ 34a,b
667 ⫾ 70
430 ⫾ 41a,b
683 ⫾ 48
437 ⫾ 41a,b
664 ⫾ 38
572 ⫾ 56
670 ⫾ 41
658 ⫾ 68
798 ⫾ 58
814 ⫾ 65b
748 ⫾ 64
803 ⫾ 55b
783 ⫾ 76
833 ⫾ 65b
F(1294) ⫽ 24.64, P ⬍ 0.000
Water
Beef
2.27 ⫾ 0.33
2.63 ⫾ 0.45
2.31 ⫾ 0.33
2.81 ⫾ 0.49
2.33 ⫾ 0.33
2.72 ⫾ 0.47
2.28 ⫾ 0.35
2.62 ⫾ 0.44
2.26 ⫾ 0.32
2.71 ⫾ 0.46
2.24 ⫾ 0.31
2.60 ⫾ 0.43
2.26 ⫾ 0.29
2.68 ⫾ 0.44
2.27 ⫾ 0.31
2.75 ⫾ 0.43
2.25 ⫾ 0.29
2.79 ⫾ 0.44
2.28 ⫾ 0.31
2.87 ⫾ 0.45b
2.25 ⫾ 0.31
3.02 ⫾ 0.49b
2.27 ⫾ 0.31
3.06 ⫾ 0.47b
2.26 ⫾ 0.29
3.04 ⫾ 0.49b
2.27 ⫾ 0.32
3.30 ⫾ 0.61b
2.32 ⫾ 0.33
3.08 ⫾ 0.52b
2.27 ⫾ 0.29
2.94 ⫾ 0.53b
F(1294) ⫽ 102.3, P ⬍ 0.000
Statistically significantly different from control (P ⬍ 0.01).
Statistically significantly different from 0800 baseline value (P ⬍ 0.01).
stitution of carbohydrate and/or protein for fatty acids in the
fuel mixture being oxidized (Fig. 8).
Using the Lusk Table (19), the integrated mean nonprotein
fuel mixture oxidized over the 8 h of the study when only
water was ingested was 27% carbohydrate and 73% fat. When
protein was ingested it was 31% carbohydrate and 69% fat.
There also was a late rise in triacylglycerol concentration
when protein was ingested (Table 2). In the absence of protein ingestion the mean concentration was stable.
Discussion
As indicated previously, it has been reported several times
that protein ingestion does not raise the circulating glucose
concentration or raises it only modestly (3–10). The reason for
this has been unclear.
In 1971, it was suggested that protein ingestion did not
raise the circulating glucose concentration because an increased production and release of glucose from the liver was
balanced by an increased uptake and utilization of glucose
by peripheral tissues (34). The mechanism proposed was that
an increased circulating glucagon concentration, resulting
from the ingestion of protein, would stimulate glucose production from amino acids in the liver. The increased insulin
concentration resulting from the ingestion of protein then
would stimulate peripheral tissues, primarily skeletal muscle, to remove the glucose produced and to store it as glycogen (34). The latter is a well known effect of high concentrations of insulin. However, using direct hepatic vein
catheterization techniques, a significant increase in glucose
production in the splanchnic bed after protein ingestion
could not be demonstrated either in dogs (35) or in humans
(36).
In a previous study in which normal young males ingested
50 g protein in the form of casein, we reported that the
ingested protein did increase the glucose appearance rate,
but this was considerably less than expected from the
amount of amino acids deaminated and the resulting amino
groups synthesized into urea (9). In the present study, the
increase in glucose appearance rate in subjects with untreated type 2 diabetes was even less when 50 g beef protein
was ingested (Fig. 4). Based on the glucose appearance rate
data integrated over 8 h, the calculated amounts of glucose
entering the circulation were a mean of 70.8 g and 68.2 g,
when protein or water were ingested, respectively. This difference was statistically significant (P ⬍ 0.001). Thus, the
amount of glucose appearing in the circulation due to protein
ingestion was 2.6 g. In normal young males ingesting cottage
cheese protein, 10 g glucose were calculated to have appeared over the 8 h of the study (9). Thus, isotope dilution
data as well as direct catheterization data indicate that protein ingestion does not result in a major glucagon-stimulated
increase in glucose production and release into the circulation with a consequent increased glucose removal rate due
to an increased insulin concentration. However, a small increase in glucose appearance does occur. The RQ data (Fig.
8) suggest that the additional small amount of glucose appearing in the circulation was largely used as fuel.
To calculate the mass of urea retained at the end of the
study, the concentration of urea nitrogen in serum water was
multiplied by total body water. This was added to the urea
nitrogen excreted to calculate the amount of protein catabolized. In these older, obese subjects the mean total body
water was estimated to be 47 ⫾ 2 L or 50 ⫾ 0.6% of body
weight using the formula of Watson et al. (15).
The calculated mean total amount of protein metabolized
was 20 g after water ingestion and 42 g after protein ingestion, or a net amount of 22 g metabolized (deaminated) as a
result of the ingestion of the protein. The change in urea
concentration in body water based on the change in plasma
concentration made up a significant proportion of the total
calculated change in urea flux.
Use of the final plasma urea nitrogen concentration to
estimate the amount of urea retained in the body water
assumes that the rate at which the additional urea produced
equilibrates in body water is rapid and does not significantly
affect the calculations (i.e. it represents the maximal amount
1044
GANNON ET AL.
FIG. 3. Plasma glucagon response. The mean glucagon concentration ⫾ SEM after ingestion of 50 g protein (solid line) or water only
(broken line) in 10 males with untreated type 2 diabetes (top). Bottom,
The change from baseline. The baseline is indicated by a light horizontal broken line in the bottom panel. *, Statistically different from
water only (P ⱕ 0.01). †, statistically different from the 0800-h time
point (P ⱕ 0.01).
of urea that could be retained). In the present study, the urea
equilibration rate was not determined. It has been estimated
to be ⬃40 min. However, two thirds of the final equilibration
occurred within 5 min (37). In dogs (38) and in cats with the
ureters ligated, it also was 30 – 45 min (39). The equilibration
rate is similar to that of D2O (40). Thus, use of changes in
serum urea concentration to calculate changes in protein
oxidation rate, as used here and by others (41), may represent
a modest overestimation of the rate at which proteins are
being deaminated. If we assume a urea pool size of 40% of
body weight to allow for incomplete mixing of urea in total
body water, the mean amount of protein-derived amino acids catabolized would be 20 g, instead of 23 g. Thus, these are
likely to represent minimum and maximum values for this
process (i.e. 20 g and 23 g).
Based on the above calculations, and assuming that for
each gram of protein deaminated, 0.56 g of glucose may be
produced (2), the expected glucose production would be
⬃11–13 g. This is considerably more than the amount of
glucose appearing in the blood as a consequence of the in-
JCE & M • 2001
Vol. 86 • No. 3
FIG. 4. Glucose rate of appearance. The mean rate of glucose appearance (Ra) ⫾ SEM after ingestion of 50 g protein (solid line) or water
only (broken line) in 10 males with untreated type 2 diabetes (top).
Bottom, The change from baseline.
gested protein. The gluconeogenic rate as estimated by the
incorporation of 14C HCO3 into glucose was nearly the same
whether water or protein was ingested (Fig. 5). Again, the net
amount of glucose produced because of the protein meal, as
calculated using the incorporation of 14C HCO3 into glucose,
was only ⬃2 g. Even assuming entrance into the gluconeogenic pathway of some amino acids not detected by 14C
HCO3 incorporation, etc. (1, 29, 30, 32, 42, 43), a maximal
increase in protein-stimulated gluconeogenesis is likely to
have yielded less than 4 g over the 8-h period. Because there
was only a very modest increase in glucose appearance in the
circulation, the ingested protein-derived amino acids most
likely replaced other gluconeogenic substrates.
The switching from endogenous gluconeogenic substrates
to absorbed gluconeogenic substrates has been observed after iv administration of gluconeogenic substrates (44, 45) and
after ingestion of fructose (24, 46) and galactose (47). In
addition to protein, the latter are the other major gluconeogenic substrates derived from dietary sources. Their ingestion also results in only a modest increase in appearance of
glucose in the circulation. Diversion of glucose into glycogen
cannot be ruled out but would be unlikely in the presence of
DIETARY PROTEIN AND TYPE 2 DIABETES
FIG. 5. Rate of Gluconeogenesis. The mean rate of gluconeogenesis ⫾
SEM after ingestion of 50 g protein (solid line) or water only (broken
line) in eight males with untreated type 2 diabetes (top). Bottom,
Gluconeogenesis as a percentage of Ra.
a high glucagon concentration. In rats, intragastric administration of a large amount of protein actually resulted in
glycogenolysis (48).
The modest increase in glucose concentration and the intracellular switching from one gluconeogenic substrate to
another occurs even though the stimulated rise in insulin and
in glucagon was considerably different following ingestion
of these fuels (24, 46, 47). The mechanism remains to be
determined.
In overnight fasted normal subjects, it has been reported
that ⬃35–50% of the glucose being produced comes from
gluconeogenesis (30, 32, 43). It also has been reported that the
gradual decrease in glucose production with fasting is due
to a progressive reduction in the rate of glycogenolysis. There
was little change in the rate of gluconeogenesis (43). In the
present study, the initial proportion of the glucose appearance rate due to gluconeogenesis as estimated by the incorporation of 14C-HCO3 into glucose, also was ⬃35% in these
overnight fasted subjects with mild type 2 diabetes, but gradually increased over the 8 h of the study. As indicated, the
method used underestimates the true gluconeogenic rate.
With continued starvation there was a progressive de-
1045
FIG. 6. ␣ Amino nitrogen response (total amino acids). The mean ␣
amino nitrogen concentration ⫾ SEM after ingestion of 50 g protein
(solid line) or water only (broken line) in 10 males with untreated type
2 diabetes (top). Bottom, The change from baseline (bottom). The
baseline is indicated by a light horizontal broken line in the bottom
panel. *, Statistically different from water only (P ⱕ 0.01); †, statistically different from the 0800-h time point (P ⱕ 0.01).
crease in glucose appearance rate and an increase in relative
gluconeogenesis rate. Thus, the results are different than
those reported in nondiabetic subjects. Also, in contrast to
nondiabetic subjects where starvation for an additional 8 h
beyond the overnight fast had only a minor effect on glucose
concentration (9) in the present study, the glucose concentration decreased continuously, indicating the importance of
glycogenolysis in maintaining the elevated, overnight fasting
glucose concentration and glucose appearance rate in people
with type 2 diabetes (Fig. 1). An indirect estimate of the
importance of glycogenolysis in the maintenance of an elevated glucose concentration in people with type 2 diabetes
with short-term fasting also has been reported from our
laboratory previously (33).
The decrease in glucose concentration in association with
a decrease in glucose production rate when only water was
ingested occurred in association with an increase in NEFA
concentration but in the absence of a significant change in
insulin concentration. These changes suggest development
1046
GANNON ET AL.
FIG. 7. Plasma urea nitrogen response. The mean urea nitrogen concentration ⫾ SEM after ingestion of 50 g protein (solid line) or water
only (broken line) in 10 males with untreated type 2 diabetes (top).
Bottom, The change from baseline. The baseline is indicated by a light
horizontal broken line in the bottom panel. *, Statistically different
from water only (P ⱕ 0.01); †, statistically different from the 0800-h
time point (P ⱕ 0.01).
of a modest degree of hepatic and adipose insulin resistance
due to fasting over the 8-h time period. This downward trend
was only temporarily interrupted by protein ingestion, even
though the insulin concentration remained elevated (Fig. 2).
An increase in insulin concentration would have been expected to decrease the glucose production rate (49). Presumably, an inhibiting effect on glucose production by insulin
was just balanced by the elevated glucagon concentration.
However, the precision of this balance is surprising and
suggests that other factors are playing a role.
The glucose disappearance rate also decreased progressively with fasting, and this was modestly greater than the
decline in glucose appearance rate. It was calculated that the
difference in glucose Ra and Rd resulted in 5 g glucose being
used in excess of the glucose appearance rate over the 8 h of
the study, both when the subjects were given protein or only
water. The decrease in Rd indicated a progressive conversion
from a carbohydrate-based fuel mixture to a more fatty acidbased mixture. This was confirmed by the RQ data.
JCE & M • 2001
Vol. 86 • No. 3
FIG. 8. Total RQ (CO2 produced/O2 consumed). The mean RQ ⫾ SEM
after ingestion of 50 g protein (solid line) or water only (broken line)
in 10 males with untreated type 2 diabetes (top). Bottom, The change
from baseline. The baseline is indicated by a light horizontal broken
line in the bottom panel.
The increase in uric acid concentration after beef ingestion
is intriguing. It may have been due to the rapid oxidation to
uric acid of purines or their derivatives present in the beef
muscle, although an accelerated rate of endogenous purine
metabolism cannot be ruled out. In beef muscle, purine nitrogen has been estimated to be ⬃0.06% by weight (50). An
increased glucagon concentration has been reported to increase the rate of uric acid excretion, whereas a high insulin
concentration has been reported to decrease it. However,
neither resulted in a change in plasma uric acid concentration
(51, 52). Ingestion of protein sources free of nucleic acids also
did not result in a change in plasma uric acid concentration,
but uric acid excretion was stimulated (53). Thus, the results
were similar to those observed with a raised glucagon concentration (52). Unfortunately, urine uric acid excretion was
not quantified in the present study.
The increase in triacylglycerol concentration induced by beef
ingestion also is of interest. It may have been due to the small
DIETARY PROTEIN AND TYPE 2 DIABETES
amount of fat present in the beef. However, based on literature
data, this is unlikely. An amount of fat more than twice that
present in the beef ingested did not raise the triacylglycerol
concentration (54). In addition, an increase in triacylglycerol
concentration was noted after ingestion of egg white (fat free)
and very low-fat cottage cheese (8). Thus, ingested protein per
se may result in a rise in triacylglycerol concentration. Whether
a change in the synthesis or removal rate is primarily responsible for the change remains to be determined.
Acknowledgments
We thank the patients for volunteering for these studies. We also
thank Mary Adams, M.T.; Kelly Jordan, B.A.; Heidi Hoover, R.D.; and
the staff of the SDTU for excellent technical expertise; Drs. Robert Rizza
and Peter Butler for advice and helpful discussions; Dr. Michael Kuskowski for advice on statistical analyses; and Claudia Durand for expert
clerical assistance.
References
1. Esenmo E, Chandramouli V, Schumann WC, Kumaran K, Wahren J, Landau
BR. 1992 Use of 14CO2 in estimating rate of hepatic gluconeogenesis. Am J
Physiol. 263:E36 –E41.
2. Janney NW. 1915 The metabolic relationship of the proteins to glucose. J Biol
Chem. 20:321–347.
3. MacLean H. 1924 Modern methods in the diagnosis and treatment of glycosuria and diabetes. London: Constable & Co. Ltd.
4. Conn JW, Newburgh LH. 1936 The glycemic response to isoglucogenic quantities of protein and carbohydrate. J Clin Invest. 15:665– 671.
5. Nuttall FQ, Mooradian AD, Gannon MC, Billington CJ, Krezowski PA. 1984
Effect of protein ingestion on the glucose and insulin response to a standardized oral glucose load. Diabetes Care. 7:465– 470.
6. Westphal SA, Gannon MC, Nuttall FQ. 1990 The metabolic response to
glucose ingested with various amounts of protein. Am J Clin Nutr. 52:267–272.
7. Nuttall FQ, Gannon MC. 1991 Plasma glucose and insulin response to macronutrients in non-diabetic and NIDDM subjects. Diabetes Care. 14:824 – 838.
8. Gannon MC, Nuttall FQ, Lane JT, Burmeister LA. 1992 Metabolic response
to cottage cheese or egg white protein, with or without glucose in type 2
diabetic subjects. Metabolism. 41:1137–1145.
9. Khan MA, Gannon MC, Nuttall FQ. 1992 Glucose appearance rate following
protein ingestion in normal subjects. J Am Coll Nutr. 11:701–706.
10. Gannon MC, Nuttall FQ, Grant CT, Ercan-Fang S, Ercan-Fang N. 1998 Stimulation of insulin secretion by fructose ingested with protein in people with
untreated NIDDM. Diabetes Care. 21:16 –22.
11. Gannon MC, Damberg G, Gupta V, Nuttall JA, Nuttall FQ. 1999 The glucose
appearance rate following ingestion of protein by people with type 2 diabetes.
Diabetes. 48:A295 (Abstract).
12. Duncombe WC. 1964 The colorimetric micro-determination of non-esterified
fatty acids in plasma. Clin Chim Acta. 9:122–125.
13. Hohorst HJ. 1965 L-(⫹)-lactate. Determination with lactic dehydrogenase and
DPN. In: Bergmeyer HU, ed. Methods of enzymatic analysis. New York:
Academic Press; 266 –270.
14. Goodwin JF. 1968 The colorimetric estimation of plasma amino nitrogen with
DFNB. Clin Chem. 14:1080 –1090.
15. Watson PE, Watson ID, Batt RD. 1980 Total body water volumes for adult
males and females estimated from simple anthropometric measurements. Am J
Clin Nutr. 33:27–39.
16. Hamberg O, Vilstrup H. 1994 Effects of insulin and glucose on urea synthesis
in normal man, independent of pancreatic hormone secretion. J Hepatol.
21:381–387.
17. Livesey G, Elia M. 1988 Estimation of energy expenditure, net carbohydrate
utilization, and net fat oxidation and synthesis by indirect calorimetry: evaluation of errors with special reference to the detailed composition of fuels. Am J
Clin Nutr. 47:608 – 628.
18. First Data Bank. 1998 Nutritionist IV (computer software). San Bruno, CA:
Hearst Corp.
19. Lusk G. 1924 Animal calorimetry: analysis of the oxidation of mixtures of
carbohydrate and fat. A correction. J Biol Chem. 59:41– 42.
20. Steele R, Wall J, deBodo JR, Altszuler H. 1956 Measurement of size and
turn-over rate of body glucose pool by the isotope dilution method. Am J
Physiol. 187:15–24.
21. de Bodo RC, Steele R, Altszuler N, Dunn A, Bishop JS. 1963 On the hormonal
regulation of carbohydrate metabolism: studies with 14C-glucose. Recent Prog
Horm Res. 19:445– 488.
22. Hetenyi G, Norwich KH. 1974 Validity of the rates of production and utili-
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
1047
zation of metabolites as determined by tracer methods in intact animals. Fed
Proc. 33:1841–1848.
Searle GL. 1976 The use of isotope turnover techniques in the study of carbohydrate metabolism in man. Clin Endocrinol Metab. 5:783– 804.
Nuttall FQ, Khan MA, Gannon MC. 2000 Peripheral glucose appearance rate
following fructose ingestion in normal subjects. Metabolism. 49:1565–1571.
Fuller G, Parker RM. 1964 Approximate integration. Applications 13–16. In:
Analytical geometry and calculus. Princeton, NJ: Van Nostrand Publishers;
367–368.
Gannon MC. 1991 A computer program to calculate the rate of appearance of
glucose in the peripheral circulation following infusion of labeled glucose.
©University of Minnesota, Patents and Licensing.
Bugianesi E, Kalhan S, Burkett E, Marchesini G, McCullough A. 1998 Quantification of gluconeogenesis in cirrhosis: response to glucagon. Gastroenterology. 115:1530 –1540.
McMahon M, Marsh HM, Rizza RA. 1989 Effects of basal insulin supplementation on disposition of a mixed meal in obese patients with NIDDM.
Diabetes. 38:291–303.
Radziuk J. 1982 Sources of carbon in hepatic glycogen synthesis during absorption of an oral glucose load in humans. Fed Proc. 41:110 –116.
Katz J. 1985 Determination of gluconeogenesis in vivo with 14C-labeled substrates. Am J Physiol. 248:R391–R399.
Turk D, Alzaid A, Dinneen S, Nair KS, Rizza R. 1995 The effects of noninsulin-dependent diabetes mellitus on the kinetics of onset of insulin action
in hepatic and extrahepatic tissues. J Clin Invest. 95:755–762.
Haymond MW, Sunehag AL. 2000 The reciprocal pool model for the measurement of gluconeogenesis by use of [U-13C] glucose. Am J Physiol.
278:E140 –E145.
Gannon MC, Nuttall FQ, Lane JT, Fang S, Gupta V, Sandhofer C. 1996 Effect
of twenty-four hours of starvation on plasma glucose and insulin concentrations in people with untreated non-insulin-dependent diabetes mellitus. Metabolism. 45:492– 497.
Unger RH. 1971 Glucagon physiology and pathophysiology. N Engl J Med.
285:443– 449.
Ishida T, Chow J, Lewis RM, Kartley CJ, Entman M, Field JB. 1983 The effect
of ingestion of meat on hepatic extraction of insulin and glucagon and hepatic
glucose output in conscious dogs. Metabolism. 32:558 –567.
Wahren J, Felig P, Hagenfeldt L. 1976 Effect of protein ingestion on splanchnic
and leg metabolism in normal man and in patients with diabetes mellitus. J Clin
Invest. 57:987–999.
San Pietro A, Rittenburg D. 1953 A study of the rate of protein synthesis in
humans. I. Measurement of the urea pool and urea space. J Biol Chem.
201:445– 451.
Painter EE. 1940 Total body water in the dog. Am J Physiol. 129:744 –755.
Kornberg HL, Davies RE, Wood DR. 1952 Measurement of total body-water
with urea. Nature. 169:502–503.
Moore FD. 1946 Determination of total body water and solids with isotopes.
Science. 104:157–160.
Tappy L, Owen OE, Boden G. 1988 Effect of hyperinsulinemia on urea pool
size and substrate oxidation rates. Diabetes. 37:1212–1216.
Kelleher JK. 1986 Gluconeogenesis from labeled carbon: estimating isotope
dilution. Am J Physiol. 250:E296 –E305.
Chandramouli V, Ekberg K, Schumann WC, Kalhan SC, Wahren J, Landau BR.
1997 Quantifying gluconeogenesis during fasting. Am J Physiol. 273:E1209 –E1215.
Steele R, Winkler B, Altszuler N. 1971 Inhibition by infused glycerol of
gluconeogenesis from other precursors. Am J Physiol. 221:883– 888.
Jahoor F, Peters EJ, Wolfe RR. 1990 The relationship between gluconeogenic
substrate supply and glucose production in humans. Am J Physiol.
258:E288 –E296.
Delarue J, Normand S, Pachiaudi C, Beylot M, Lamisse F, Riou JP. 1993 The
contribution of naturally labelled 13C fructose to glucose appearance in humans. Diabetologia. 36:338 –345.
Gannon MC, Khan MA, Nuttall FQ. 2001 Glucose appearance rate after the
ingestion of galactose. Metabolism. 50:93–98.
Gannon MC, Nuttall FQ. 1987 Oral protein hydrolysate causes liver glycogen
depletion in fasted rats pretreated with glucose. Diabetes. 36:52–58.
Alzaid AA, Dinneen SF, Turk DJ, Caumo A, Cobelli C, Rizza RA. 1994
Assessment of insulin action and glucose effectiveness in diabetic and nondiabetic humans. J Clin Invest. 94:2341–2348.
Burian R, Schur H. 1900 Ueber die Stellung der Purinkörper im menschlichen
Stoffwechsel. Arch f d ges Physiol. 80:241–350.
Quinones-Galvan A, Natali A, Baldi S, et al. 1995 Effect of insulin on uric acid
excretion in humans. Am J Physiol. 268:E1–E5.
Yamamoto T, Moriwaki Y, Takahashi S, et al. 1999 Effect of amino acids on
the plasma concentration and urinary excretion of uric acid and uridine.
Metabolism. 48:1023–1027.
Garrel DR, Verdy M, Petit-Clerc C, Martin C, Brulé D, Hamet P. 1991 Milk
and soy protein ingestion: acute effect on serum uric acid concentration. Am J
Clin Nutr. 53:665– 669.
Jeppesen J, Ida Chen YD, Zhou MY, Wang T, Reaven GM. 1995 Effect of
variations in oral fat and carbohydrate load on postprandial lipemia. Am J Clin
Nutr. 62:1201–1205.