International Journal of Obesity (1999) 23, 304±311
ß 1999 Stockton Press All rights reserved 0307±0565/99 $12.00
http://www.stockton-press.co.uk/ijo
Energy intake and appetite are suppressed by
glucagon-like peptide-1 (GLP-1) in obese men.
E NaÈslund*1, B Barkeling2, N King3, M Gutniak4, JE Blundell3, JJ Holst5, S RoÈssner2 and PM HellstroÈm6,
1
Division of Surgery, Danderyd Hospital, Karolinska Institute, Stockholm, Sweden; 2Obesity Unit, Huddinge University Hospital,
Stockholm, Sweden; 3Biopsychology group, Leeds University, Leeds, UK; 4Multidisciplinary Pain Center, Kronan, Stockholm, Sweden;
5
Department of Medical Physiology, University of Copenhagen, Copenhagen, Denmark; and 6Department of Hepatology and
Gastroenterology, Karolinska Hospital, Stockholm, Sweden
BACKGROUND: Peripheral administration of glucagon-like peptide-1 (GLP-1) for four hours, to normal weight and
obese humans, decreases food intake and suppresses appetite.
OBJECTIVE: The aim of this study was to assess the effect of an eight hour infusion of GLP-1 on appetite and energy
intake at lunch and dinner in obese subjects.
DESIGN: Randomised, blinded cross-over design with intravenous infusion of GLP-1 (0.75 pmol kg71 min71) or saline.
SUBJECTS: Eight obese (body mass index, BMI, 45.5 2.3 kg=m2) male subjects.
MEASUREMENTS: Ad libitum energy intake at lunch (12.00 h) and dinner (16.00 h) after an energy ®xed breakfast
(2.4 MJ) at 08.00 h. Appetite sensations using visual analogue scales, (VAS) immediately before and after meals and
hourly in-between. Blood samples for the analysis of glucose, insulin, C-peptide, GLP-1 and peptide YY. Gastric
emptying after breakfast and lunch using a paracetamol absorption technique.
RESULTS: Hunger ratings were signi®cantly lower with GLP-1 infusion. The summed ad libitum energy intake at lunch
and dinner was reduced by 1.7 0.5 MJ (21 6%) by GLP-1 infusion (P ˆ 0.01). Gastric emptying was delayed by GLP-1
infusion, and plasma glucose concentrations decreased (baseline: 6.6 0.35 mmol=L; nadir: 5.3 0.15 mmol=L). No
nausea was recorded during GLP-1 infusion.
CONCLUSIONS: Our results demonstrate that GLP-1 decreases feelings of hunger and reduces energy intake in obese
humans. One possible mechanism for this ®nding might be an increased satiety primarily mediated by gastric vagal
afferent signals.
Keywords: appetite; food intake; gastric emptying; insulin; universal eating monitor
Introduction
Glucagon-like peptide-1 (7-36)amide (GLP-1) is a
peptide of 30 amino acids produced in, and secreted
from, L-cells of the intestinal mucosa into the circulation, after intake of a mixed meal.1,2 By exerting a
receptor-mediated insulinotropic action on the pancreatic b-cell in the postprandial state, GLP-1 is
considered an incretin. GLP-1 also inhibits gastric
emptying and acid secretion and is, as such, a candidate mediator of the `ileal brake'.1
GLP-1 has been shown to affect food intake. In the
rat brain, GLP-1 immunoreactive cell bodies and
nerve ®bres are found in areas of the brain that receive
afferents from the gastrointestinal tract via the vagal
nerves and control food intake.3,4,5,6 Intracerebroventricular (ICV) injection of GLP-1 in rodents, inhibits
food intake.7 ± 9 In addition, central administration of
GLP-1 has been shown to inhibit the intake of water.8
The ICV administration of exendin (9-39)amide, a
GLP-1 receptor antagonist, to satiated, but not fasted
rats, increases food intake.9 Indeed, administration of
*Correspondence: Dr Erik NaÈslund, Department of Surgery,
Danderyd Hospital, SE-182 88 Danderyd, Sweden.
Received 20 July 1998; revised 14 October 1998; accepted
27 October 1998
exendin (9-39)amide twice daily over 10 d, has been
shown to increase food intake and results in a signi®cant weight gain in treated rats.10
Studies in normal weight humans, have shown that
peripheral administration of GLP-1 decreases food
intake and suppresses appetite.11,12 Obese subjects
may have an attenuated release of GLP-1 in response
to meals.13,14 In a previous study, we demonstrated
that intravenous infusion of GLP-1, for four hours, to
six obese subjects, resulted in lower hunger ratings
over four hours after the start of infusion. Concomitantly with the start of GLP-1 infusion, a meal was
eaten on a universal eating monitor. No effect of GLP1 was seen on food intake parameters.15 The aim of
the present study was therefore to extend these studies
and to examine if GLP-1 administered intravenously
for 8 h to obese humans, affects appetite and energy
intake at an anticipated meal. In order to study one
possible mechanism of GLP-1 action on food intake,
the rate of gastric emptying was studied using a
paracetamol absorption technique.
Experimental subjects
Eight male obese subjects (mean s.e.m., body mass
index (BMI) 45.5 2.3 kg=m2 and aged 35.0 3.8 y)
GLP-1, appetite and energy intake
E NaÈslund et al
were recruited from patients attending the outpatient
Obesity Unit at the Huddinge University Hospital.
Two patients were taking medication for hypertension, but the remaining subjects were not taking any
medication. Informed consent was obtained from all
subjects and the study was approved by the Ethics
Committee of the Karolinska Hospital.
Material and methods
Study protocol
The subjects were instructed to fast from midnight
before the study and refrain from alcohol consumption
and exercise for 24 h prior to the study day. They
reported to the Obesity Unit at 07.15 h in the morning
and an indwelling catheter was placed in each antecubital vein. The study was performed in a randomised double-blind, cross-over fashion on two
occasions, one week apart. Simultaneously with the
onset of intake of a ®xed energy breakfast at 08.00 h,
either GLP-1 (0.75 pmol=kg=min)(Bachem AG,
Bubendorf, Switzerland) dissolved in 0.9% saline
containing 1% albumin (Albumin Kabi 200 mg=ml),
subjected to sterile ®ltration and stored at 7 20 C
until use or saline was started in one of the intravenous catheters and continued for eight hours. For
infusion, an Infusor SV (Baxter, Swinford, Eire)
5 ml=h was used. Free access to drinking water was
allowed throughout the day.
Meals and the universal eating monitor
Breakfast was served at 08.00 h and consisted of bread
(55 g), butter (15 g), cheese (60 g) and orange juice
(200 g). The energy content was 2.4 MJ (protein 22.6 g
(16 energy (E)%), carbohydrate 49.6 g (36 E%), fat
(30.9 g (48 E%)). At 12.00 h and 16.00 h an ad libitum
test meal was served on a universal eating monitor
(VIKTOR). The VIKTOR has been described in detail
elsewhere.16 In short, the VIKTOR-equipment consists of a hidden scale built into a table and connected
to a computer. A homogenous meal in excess, is
placed on a plate and the subjects are instructed to
eat until satis®ed. The total food intake, meal duration, eating rate (initial and total) and rate of deceleration were measured by VIKTOR. The test meal at
12.00 h was an industrially produced pasta dish (Pasta
Bolognese, NestleÂ, Bjuv, Sweden) with the addition of
6.5 g fat=100 g pasta. This resulted in a meal of
0.7 MJ=100 g (protein 5 g (13 E%), carbohydrate
15 g (38 E%) and fat 9 g (49 E%)). For dinner an
industrially produced Swedish hash with a standard
energy content of 0.7 MJ=100 g (protein 5 g (13 E%),
carbohydrate 14 g (37 E%) and fat 9 g (50 E%)),
consisting of diced meat, onions and potatoes (Oxpytt, NestleÂ, Bjuv, Sweden) was served.
Assessment of appetite and nausea
The desire to eat, hunger, fullness, prospective consumption and feeling of nausea, were studied with
visual analogue scales (VAS)17 15 min before, 15 min
after, and hourly in between, meals. After completion
of lunch and dinner, the subjects were asked to assess
the pleasantness of each meal using a VAS.
Assays for glucose, insulin, C-peptide, GLP-1 and
peptide YY
Blood samples were obtained for the analysis of
glucose and plasma concentrations of insulin, C-peptide, GLP-1 and peptide YY (PYY) at 07.40 h,
07.50 h, 08.00 h, 08.15 h, 08.30 h, 09.00 h, 10.00 h,
11.00 h, 11.45 h, 12.15 h, 12.30 h, 13.00 h, 14.00 h,
15.00 h, 15.45 h and 16.15 h. Blood samples were
collected in pre-chilled heparinised tubes. The samples were centrifuged at 4 C for 10 min at 2000 g.
Plasma was collected and stored at 7 20 C for
analysis in one series.
Glucose was analysed using an enzyme assay
(mutarotase and glucose dehydrogenase) (BoeringerMannheim GmbH, Mannheim, Germany) using a
Hitachi 917 automatic analyser.
Insulin was analysed using an enzyme immunoassay (DAKO Insulin Kit K6219, Copenhagen, Denmark). The assay cross-reacts to 0.3% with proinsulin
but not with C-peptide. The detection limit of the
assay was 3 mU=L and the intra-assay coef®cient of
variation was 8%.
C-peptide concentrations in plasma were determined by a radioimmunoassay (RIA) (Euro-Diagnostica, MalmoÈ, Sweden). The assay cross-reacts to
proinsulin by 41%, but not with insulin. The detection
limit was 50 nmol=L and the intra-assay coef®cient of
variation was 5%.
GLP-1-like immunoreactivity in plasma was studied using RIA speci®c for each terminus of the
molecule. N-terminal immunoreactivity was measured
using a newly described antiserum18 raised in rabbits
against synthetic proglucagon 78 ± 87 with a C-terminal cysteine coupled to keyhole limpet hemocyanin
using the cystein thiol method. The selected antiserum
(code 93242) was used in a ®nal dilution of 1:15000,
and has a detection limit of 5 pmol=L. The antiserum
has a cross-reactivity of approximately 10% with
GLP-1 (1 ± 36)amide and 0.1% with GLP-1 (8 ±
36)amide and GLP-1 (9 ± 36)amide. High performance liquid chromatography (HPLC) supports the
use of RIA with this speci®city to measure concentrations of intact GLP-1.19 C-terminal immunoreactivity
was measured using antiserum 89390,20 which has an
absolute requirement for the intact amidated C-terminus of GLP-1 (7 ± 36)amide, and cross-reacts
< 0.01% with truncated fragments, and 83% with
GLP-1 (9 ± 36)amide. For all assays, the intra-assay
coef®cient of variation was < 6%. GLP-1 (7 ±
36)amide was used as standard, and 125I-labelled
GLP-1 (7 ± 36)amide as tracer. Before analysis
305
GLP-1, appetite and energy intake
E NaÈslund et al
306
plasma was extracted with 70% ethanol (vol=vol, ®nal
concentration) before assay, giving recoveries of 75%.
Separation was achieved using plasmacoated charcoal.
PYY-like immunoreactivity was analysed using
antiserum code no. 8412-2II (a gift from R. HaÊkanson,
Dept of Pharmacology, University of Lund, Lund,
Sweden) raised in rabbits against synthetic porcine
PYY 1 ± 36 (Peninsula Europe, Merseyside, UK) as
previously described,21 but without conjugation to
carrier protein.22 The antiserum cross-reacts 100%
with human PYY. The detection limit of the assay
was 1 pmol=L and the intra-assay coef®cient of
variation was 5%.
Gastric emptying
The rate of gastric emptying, was assessed by utilising
paracetamol absorption from the intestine after oral
ingestion as a marker.23 Plasma concentrations of
paracetamol were assessed at the time points
described above, after ingestion of paracetamol
(1.5 g) dissolved in 200 ml of water with breakfast
and lunch. Analysis of paracetamol was assessed by
¯uorescence-immunoassay (IMX, Abbott Laboratories, Chicago, IL). The intra-assay coef®cient of
variation was 5%.
Statistics and calculations
All values are mean s.e.m. or median (range) as
appropriate and P < 0.05 was considered statistically
signi®cant. Appetite and nausea ratings (VAS) were
analysed by calculating inter-meal delta values
(change in rating between the value after breakfast
(08.15 h) and before lunch (11.45 h); and the value
after lunch (12.15 h) and before dinner (15.45 h)).
These delta values and the results of the VIKTOR,
were analysed using Wilcoxon signed rank test for
matched pairs. Plasma concentrations of glucose,
insulin, C-peptide, N-terminal GLP-1, C-terminal
GLP-1 and PYY were analyzed employing an
ANOVA with repeated, paired measures with time
and treatment as factors.
The gastric emptying pro®le was estimated after
conversion of plasma paracetamol concentration
values to cumulated values, re¯ecting total absorption
of the drug. The inverted absorption values signi®es
the absorption of the drug in relation to maximal
absorption during saline infusion. Each individual
gastric emptying curve was adapted to a third-degree
polynomial. For repeated paracetamol administration,
the remaining plasma concentrations from the ®rst
administration, were subtracted according to the elimination rate of the drug, according to its terminal
elimination phase. For statistical analysis the actual
plasma paracetamol concentrations values were used
and analysed employing an ANOVA with repeated,
paired measures with time and treatment as factors.
Results
Food intake and eating rate
After the energy ®xed breakfast, GLP-1 infusion
resulted in reduced food intake and slower overall
eating rate at the ad libitum lunch and dinner (Table
1). The summed reduction in energy intake at lunch
and dinner during GLP-1 infusion, was 1.7 0.5 MJ
(21 6% reduction) (P ˆ 0.01). There was no difference in eating parameters, such as initial eating rate or
eating curve (accelerating vs decelerating) measured
on the universal eating monitor during GLP-1 or
saline infusion.
Appetite ratings, test meal palatability and nausea
In the control treatment, hunger ratings decreased
after breakfast (08.00 h). Thereafter hunger ratings
rose gradually until before lunch (11.45 h), and
decreased again after lunch (12.00 h) to rise again
until before dinner (15.45 h). A similar pattern was
seen after GLP-1 infusion, but the ratings were lower.
GLP-1 infusion resulted in a smaller increase in
ratings of hunger (P < 0.05) and prospective consumption (P < 0.05) and a corresponding decrease in
ratings of fullness (P < 0.05) during the period
between breakfast and lunch. The difference in ratings
of desire to eat, was not quite signi®cant (P ˆ 0.07).
The change in appetite ratings between lunch and
dinner showed a similar pattern (hunger (P ˆ 0.01),
desire to eat (P < 0.05), prospective consumption
(P ˆ 0.07) and fullness (P ˆ 0.09) (Figure 1). Meal
palatability at lunch was similar during saline (79.4
(55 ± 100)) and GLP-1 (77.9 (56 ± 89)) (P ˆ 0.57)
infusion and at dinner during saline (69.0 (2 ± 91))
and GLP-1 (49 (2 ± 93)) (P ˆ 0.16) infusion. There
were no differences in feelings of nausea during GLP1 or saline infusion (Figure 2).
Gastric emptying
Gastric emptying was slower during GLP-1 infusion,
as re¯ected by a reduced absorption rate of paracetamol (P < 0.001). Both the rate of gastric emptying
after breakfast and after lunch was slower during
GLP-1 infusion, with approximately 50% reduced
Table 1
Results of the universal eating monitor (VIKTOR)
characteristics during infusion of saline (NaCl) and glucagonlike peptide-1 (GLP-1) in obese human subjects
Energy intake (MJ) Meal duration (min)
Lunch
NaCl
GLP-1
Dinner
NaCl
GLP-1
Eating rate (g/min)
4.3 (3.5 ± 4.8)
3.8 (2.1 ± 4.6)*
6.7 (5.1 ± 8.4)
7.0 (4.0 ± 10.1)
88.0 (74.3 ± 110.0)
78.0 (67.2 ± 106.1)*
3.1 (1.4 ± 5.0)
2.3 (0.8 ± 3.0)*
6.3 (3.7 ± 8.6)
4.3 (2.4 ± 5.9)
79.7 (40.6 ± 104.1)
68.7 (36.4 ± 83.6)*
Data shown as median (range). Wilcoxon signed rank test for
matched pairs. *P < 0.05 GLP-1 vs saline for lunch and dinner,
respectively.
GLP-1, appetite and energy intake
E NaÈslund et al
307
Figure 1 Mean s.e.m. appetite ratings (hunger, fullness, desire to eat, and prospective consumption) during glucagon-like peptide-1
(GLP-1, ®lled circles) and saline (®lled squares) infusion in eight obese male subjects. An energy ®xed breakfast (2.4 MJ) was served at
08.00 h and an ad libitum lunch and dinner at 12.00 h and 16.00 h, respectively. Data were analysed using Wilcoxon signed rank test for
matched pairs to determine the change in rating between breakfast and lunch, and lunch and dinner (between breakfast and lunch:
hunger (P < 0.05), fullness (P < 0.05), desire to eat P ˆ 0.07 and prospective consumption (P < 0.05); between lunch and dinner: hunger
(P ˆ 0.01), fullness (P ˆ 0.09), desire to eat (P < 0.05) and prospective consumption (P ˆ 0.07).
absorption of paracetamol compared to controls
(Figure 3).
Glucose, insulin, C-peptide, GLP-1 and PYY
Plasma glucose concentrations were elevated after
breakfast and lunch in the control setting, while
GLP-1 infusion resulted in lower glucose concentrations (Figure 4, P < 0.001). Plasma C-peptide concentrations were higher during saline infusion (P ˆ 0.01),
whereas no difference in plasma concentrations of
insulin were seen between GLP-1 and saline infusions
(Figure 4). PYY concentrations were lower during
infusion with GLP-1 (Figure 5, P < 0.001). Infusion of
GLP-1 at 0.75 pmol=kg=min resulted in a six-fold
elevation of C-terminal plasma GLP-1 concentrations
(approx. 20 vs 120 pmol=L), and a 2.5-fold elevation
of the biologically active N-terminal GLP-1 (approx.
6 vs 16 pmol=L) (P < 0.001, respectively) (Figure 5).
Figure 2 Mean s.e.m. nausea ratings during glucagon-like
peptide-1 (GLP-1, ®lled circles) and saline (®lled squares) infusion in eight obese male subjects. An energy ®xed breakfast
(2.4 MJ) was served at 08.00 h and an ad libitum lunch and dinner
at 12.00 h and 16.00 h, respectively. Data were analysed using
Wilcoxon signed rank test for matched pairs to determine the
change in rating between breakfast and lunch, and lunch and
dinner. (No signi®cant difference between breakfast and lunch,
and lunch and dinner, respectively).
Discussion
This study demonstrates that peripherally administered GLP-1 in slightly supraphysiological concentrations,24 decreased hunger and reduced energy intake in
GLP-1, appetite and energy intake
E NaÈslund et al
308
Figure 3 Gastric emptying pro®le during glucagon-like peptide1 (GLP-1) (after breakfast: ®lled circles; after lunch: open circles)
and saline (after breakfast: ®lled squares; after lunch: open
squares) infusion in eight obese male subjects. Analysis on
actual plasma paracetamol concentrations, by ANOVA with
repeated measures time, treatment and treatment time interaction effect, all P < 0.001.
obese men. Concomitantly, gastric emptying was
delayed by GLP-1.
These results extend and con®rm those seen in our
previous study, where GLP-1 was infused for four
hours in obese subjects.15 In that study, GLP-1 infusion was initiated at the start of a test meal and
appetite ratings were monitored for four hours. No
difference was seen in energy intake parameters, but
postprandial hunger ratings were lower during GLP-1
compared to saline infusions. In the present study, the
appetite rating pro®les were similar to those in our
short-term study,15 but in addition we found a signi®cant reduction in energy intake at lunch and
dinner. Reduced appetite ratings and decreased
energy intake have also been demonstrated in
normal weight male subjects during GLP-1 infusion
for four hours.11 It has previously been shown that the
eating rate increases as hunger increases25 and the
eating rate is low when the hunger drive is low.26 The
eating rate was slower both at lunch and dinner during
GLP-1 infusion, compared to saline. Thus it is possible that GLP-1 reduces the eating-drive.
It has been suggested that the effect of GLP-1 on
food intake in rodents may be due to a centrally
evoked taste aversion.27 We were, however, unable
to detect any change in palatability (pleasantness of
the meal) of the test meals in this study or our
previous shorter study15 or change in nausea ratings
during GLP-1 compared to saline infusion our obese
men. Similar results on palatability ratings were also
demonstrated in the recent study on the effect of GLP1 on food intake in normal weight men.11 The test
meal used in studies with the VIKTOR, is a standard
Figure 4 Mean s.e.m. plasma glucose, insulin and C-peptide
concentrations during glucagon-like peptide-1 (GLP-1, ®lled circles) and saline (®lled squares) infusion in eight obese male
subjects. An energy ®xed breakfast (2.4 MJ) was served at
08.00 h and an ad libitum lunch and dinner at 12.00 h and
16.00 h, respectively. By ANOVA with repeated measures, time,
treatment and treatment time interaction effect, all P < 0.001
(glucose), P < 0.001, P ˆ 0.47, P ˆ 0.01, respectively (C-peptide)
and P < 0.001, P ˆ 0.65, P ˆ 0.13, respectively (insulin).
Swedish lunch hash, which has a high fat content.16 In
order to have the same energy distribution at lunch
and dinner, fat was added to the pasta dish. As the test
meals used in this study are high in fat, it raises the
question of whether the GLP-1 effect on food intake is
dependent on a high fat diet. However, in the study of
Flint et al, 11 the fat content was lower and in our
short-term study, we could not detect any speci®c
selection of food items in a food choice list during
GLP-1 infusion compared to saline,15 suggesting that
the effect of GLP-1 is not limited to a high fat diet. In
this study, access to water was unlimited and no
GLP-1, appetite and energy intake
E NaÈslund et al
Figure 5 Mean s.e.m. plasma peptide YY (PYY) and glucagonlike peptide-1 (C-terminal GLP-1 and N-terminal GLP-1) concentration during glucagon-like peptide-1 (GLP-1, ®lled circles) and
saline (®lled squares) infusion in eight obese male subjects. An
energy ®xed breakfast (2.4 MJ) was served at 08.00 h and an ad
libitum lunch and dinner at 12.00 h and 16.00 h, respectively. By
ANOVA with repeated measures, time, treatment and treatment time interaction effect: all P < 0.001 for PYY, C- and Nterminal GLP-1.
measurements of drinking were made. In rodents, it
has been demonstrated that ICV administration of
GLP-1 inhibits intake of water.8 It is not clear if IV
GLP-1 has a similar effect in man.
Several mechanisms by which GLP-1 inhibits
energy intake and decreases feelings of hunger, can
be envisioned. Previous studies show that GLP-1
inhibits gastric emptying in normal weight volunteers,28,29 as well as obese subjects,15 and these results
were con®rmed in this study. The decreased PYY
concentrations in plasma, are also most likely to be a
result of this retarded rate of gastric emptying seen
during GLP-1 infusion. The slower gastric emptying,
results in a prolonged period of gastric distension and
a more sustained delivery of nutrients to the small
intestine. In humans, satiety is induced if an emulsion
of lipid is infused into the duodenum at the same time
as the stomach is distended,30 and hunger is suppressed if lipid is infused into the jejunum without
gastric distension.31 The gastrointestinal tract is
equipped with mechano- and chemosensitive receptors, which relay information to the central nervous
system via the vagus nerve.32 It is possible that the
decreased gastric emptying rate, seen after infusion of
GLP-1, results in a prolonged stimulation of mechanoand chemosensitive receptors in the stomach and
small intestine, which may lead to prolonged satiety
and decreased appetite.
GLP-1 has been shown to cross the blood-brain
barrier via the subfornical organ and area postrema of
the circumventricular organs.33 Activation of GLP-1
receptors in the areas of the brain which control
feeding by means of ICV injection of GLP-1 in
rodents results in inhibition of food intake.7,8,9
Thus, it is also possible that peripherally administered
GLP-1 exerts its effect via a central mechanism. Both
the GLP-1 receptor and GLP-1-like immunoreactivity
are found in the hypothalamus.34 There is, however,
no evidence of an increased hypothalamic postprandial GLP-1 secretion. Contrary to the assumption of a
central satiating effect of peripherally administered
GLP-1 on food intake, there is no effect on food
intake after intraperitoneal injection of GLP-1 in
rats.8,9 This could be interpreted to indicate that
GLP-1 receptors in the hypothalamus are not accessible for peripherally administered GLP-1. Yet, GLP1 is metabolized very rapidly in rats,35 which may
explain the ineffectiveness of peripherally administered GLP-1 on food intake in rodents. The study of
Gutzweiler et al 12 supports the concept of a central
mechanism of GLP-1 in man. Here, GLP-1 was
infused one hour prior to food intake in human
subjects and energy intake was reduced in a dosedependent manner. In addition, feelings of hunger
were reduced in the period before the test meal.
This suggests that peripheral in¯uences, such as
gastric signals of fullness and the activation of intestinal mechano- or chemoreceptors, are less likely to
play a signi®cant role for the satiating effect of GLP1. Taken together, these ®ndings indicate that GLP-1,
perhaps like cholecystokinin,36 may exert its effect on
food intake and appetite via interaction with sensory
®bers in the periphery, possibly in the vagus nerve.
It has previously been shown that some obese
subjects have an attenuated release of GLP-1 and
enteroglucagon, secreted in parallel with GLP-1
after a meal.13,15,37 Thus, it can be speculated that
the lower GLP-1 response in obese subjects after a
meal, may result in a shorter period of inter-meal
satiety and more frequent food intake to maintain an
appropriate individual satiety level. Therefore, GLP-1
administered to obese subjects, may have an anorectic
309
GLP-1, appetite and energy intake
E NaÈslund et al
310
and even weight-reducing effect if given over a
prolonged period of time. The blood glucose-lowering
effect of GLP-1 has not been shown to exhibit
tachyphylaxis, when administered intravenously for
seven days in man,38 and so it may be hypothesised
that an appetite-reducing effect of GLP-1 should not
exhibit tachyphylaxis. However, recent data on longterm ICV administration to rodents is con¯icting.
There is evidence of tachyphylaxis,39 whereas it has
been shown that the central pathways involved in the
control of food consumption, do not desensitise after
chronic exposure to GLP-1.40 To evaluate the physiological relevance of GLP-1 in weight control, a mouse
model with a null mutation of the GLP-1 receptor has
been developed.41 In these knock-out mice, all binding
sites and effects of GLP-1 were eliminated. However,
no obvious phenotype of these mice was found
compared to wild type mice. In control animals,
administration of GLP-1 ICV resulted in a similar
inhibition of food intake as in rats,8,9 whereas no
effect was seen in the knock-out mice.41 Thus, the
inhibitory effect of GLP-1 ICV on feeding seems to be
mediated via a speci®c receptor interaction in the
brain. The lack of expression of the GLP-1 receptor
in knock-out mice did not lead to any changes in
feeding behaviour, suggesting the presence of redundant signalling systems and potentially adaptive
responses in these knock-out mice, which may ®nd
an explanation in the postreceptor mechanisms by
which exogenous GLP-1 reduces food intake. Thus,
it is far too early to speculate if GLP-1 therapy will be
effective as a long-term treatment in morbid obesity.
None of the subjects in this study were on any
treatment for diabetes, but demonstrated elevated
fasting plasma glucose concentrations and hyperinsulinaemia. GLP-1 has previously been shown to reduce
plasma glucose concentrations29 and GLP-1 has been
suggested as a treatment for non-insulin dependent
diabetes mellitus (NIDDM).38 Thus, in addition to the
reduction in food intake seen by GLP-1, an additional
bene®cial effect of GLP-1 is the improvement in
glycaemic control seen during GLP-1 infusion in
these obese subjects.
Conclusion
In summary, we propose that GLP-1 has a physiological role as a regulator of inter-meal satiety in man,
and that GLP-1 administered to obese subjects over
eight hours reduces hunger and energy intake. This
effect of GLP-1 may be mediated through vagal
afferent pathways in concert with a prolonged gastric
emptying signalling satiety to the brain.
Acknowledgements
This study was supported by grants from the Swedish
Medical Research Council (No 7916), the Danish
Medical Research Council, the funds of the Karolinska Institute, the Professor Nanna Svartz Fund, the
Magnus Bergvall Fund, the Ruth and Richard Juhlin
Fund and the Swedish Medical Society.
References
1 Holst JJ. Glucagon-like peptide-1 (GLP-1) ± a newly discovered gastrointestinal hormone. Gastroenterology 1994;
107: 1848 ± 1855.
2 Bell GI, Santerre RF, Mullenbach GT. Hamster preproglucagon contains the sequence of glucagon and two related peptides. Nature 1983; 302: 716 ± 718.
3 Jin SLC, Han VKM, Simmons JG, Towle AC, Lauder JM,
Lund PK. Distribution of glucagon-like peptide-1 (GLP-1),
glucagon and glicentin in the rat brain: an immunocytochemical study. J Comp Neurol 1988; 271: 519 ± 532.
4 Kreymann B, Ghatei MA, Burnett P, Williams S, Kanse S,
Diani AR, Bloom SR. Characterization of glucagon-like peptide 1(7-36) amide in the hypothalamus. Brain Res 1989; 502:
325 ± 331.
5 Shimizu I, Hirota M, Ohboshi C, Shima K. Identi®cation and
localization of glucagon-like peptide 1 and its receptor in rat
brain. Endocrinology 1987; 121: 1076 ± 1082.
6 Kanse SM, Kreymann B, Ghatei MA, Bloom SR. Identi®cation and characterisation of glucagon-like peptide 1(736amide) binding sites in the rat brain and lung. FEBS Lett
1988; 241: 209 ± 214.
7 Lambert PD, Wilding PH, Ghatei MA, Bloom SR. A role for
GLP-1-(7-36)NH2 in the central control of feeding behaviour.
Digestion 1994; 54: 360 ± 361.
8 Tang-Christensen M, Larsen PJ, GoÈke R, Fink-Jensen A,
Jessop DS, Mùller M, Sheikh SP. Central administration of
GLP-1-(7-36) amide inhibits food and water intake in rats. Am
J Physiol 1996; 271: R848 ± R856.
9 Turton MD, O'Shea D, Gunn I, Beak SA, Edwards CMB,
Meeran K, Choi SJ, Taylor GM, Heath MM, Lambert PD,
Wilding JPH, Smith DM, Ghatei MA, Herbert J, Bloom SR. A
role for glucagon-like peptide-1 in the central regulation of
feeding. Nature 1996; 379: 69 ± 72.
10 Bloom SR. Glucagon-like peptide-1 and satiety. Reply to Van
Dijk G, Thiele TE, Seeley RJ, Woods SC, Bernstein IL.
Nature 1997; 385: 214.
11 Flint A, Raben A, Astrup A, Holst JJ. Glucagon-like peptide 1
promotes satiety and suppresses energy intake in humans.
J Clin Invest 1998; 101: 515 ± 520.
12 Gutzweiler JP, GoÈke B, Drewe J, Ketterer S, Handschin D,
Hildebrand P, Winterhalder R, Conen D, Beglinger C. Glucagon-like peptide-1 is a physiologic regulator of food intake in
humans. Gastroenterology 1997; 112: A1153 (Abstr).
13 Ranganath LR, Beety JM, Morgan LM, Wright JW, Howland
R, Marks V. Attenuated GLP-1 secretion in obesity: cause or
consequence. Gut 1996; 38: 916 ± 919.
14 NaÈslund E, GrybaÈck P, Backman L, Jacobsson H, Holst JJ,
Theodorsson E, HellstroÈm PM. Distal small bowel hormones:
correlation to fasting antroduodenal motility and gastric emptying. Dig Dis Sci 1998; 43: 945 ± 952.
15 NaÈslund E, Gutniak M, Skogar S, RoÈssner S, HellstroÈm PM.
Glucagon-like peptide-1 (GLP-1) increases the period of
postprandial satiety and slows gastric emptying in obese
humans. Am J Clin Nutr 1998; 68: 525 ± 530.
16 Barkeling B, RoÈssner S, BjoÈrvell H. Effects of a high-protein
meal (meat) and a high-carbohydrate meal (vegetarian) on
satiety measured by automated computerized monitoring on
subsequent food intake. Int J Obes 1990; 14: 743 ± 751.
17 Blundell J, Rogers P, Hill A. Evaluating the satiating power of
foods: Implications for acceptance and consumption. In:
Solms J (ed). Chemical composition and sensory properties
of food and their in¯uence on nutrition. Academic Press:
London, 1988; 205 ± 219.
GLP-1, appetite and energy intake
E NaÈslund et al
18 Gutniak MK, Larsson H, Heibner SJ, Juneskans TO, Holst JJ,
Ahren B. Therapeutic levels of glucagon-like peptide-1 can be
obtained in humans by a buccal tablet. Diabetes Care 1996;
19: 843 ± 848.
19 Deacon CF, Johnsen AH, Holst JJ. Degradation of glucagonlike peptide-1 invitro yields an N-terminally truncated peptide
which is a major endogenous metabolite in vivo. J Clin
Endocrinol Metab 1995; 80: 952 ± 957.
20 érskov C, Rabenhùj L, Kofod H, Wettergren A, Holst JJ.
Production and secretion of amidated and glycine-extended
glucagon-like peptide-1 (GLP-1) in man. Diabetes 1994; 43:
535 ± 539.
21 Holst JJ, Bersani M. Assays for peptide products of somatostatin gene expression. In: Michael Conn P (ed). Methods in
neurosciences, vol 5: neuropeptide technology. Academic
Press: San Diego, 1991, 3 ± 22.
22 Ekman R, Wahlestedt C, Bottcher G, Sundler F, HaÊkanson R,
Panula P. Peptide YY-like immunoreactivity in the central
nervous system of the rat. Regul Peptides 1986; 16: 157 ± 168.
23 Maddern G, Miners J, Collins PJ, Jamieson GG. Liquid gastric
emptying assessd by direct and indirect techniques: radionuclide labelled liquid emtpying compared with a simple
paracetamol marker method. Aust NZ J Surg 1985; 55:
203 ± 206.
24 érskov C, Wettergren A, Holst JJ. Secreation of the incretin
hormones glucagon-like peptide-1 (GLP-1) and gastric inhibitory polypeptide (GIP) correlates with insulin secretion in
normal man throughout the day. Scand J Gastroenterol
1996; 31: 665 ± 670.
25 Hill SW, McCutcheon NB. Contribution of obesity, gender,
hunger, food preference, and body size to bite size, bite speed
and rate of eating. Appetite 1984; 5: 73 ± 83.
26 Hylander B, Barkeling B, RoÈssner S. Eating behaviour in
continuous ambulatory peritoneal dialysis and hemodialysis
patients. Am J Kidney Dis 1992; 6: 592 ± 597.
27 Thiele TE, Van Dijk G, Camp®eld LA, Smith FJ, Burn P,
Woods SC, Bernstein IL, Seeley R. Central infusion of GLP-1,
but not leptin, produces conditioned taste aversion in rats. Am
J Physiol 1997; 272: R726 ± R730.
28 Wettergren A, Schjoldager B, Mortensen PE, Myhre J, Christiansen J, Holst JJ. Truncated GLP-1 (proglucagon 78 ± 107amide) inhibits gastric and pancreatic function in man. Dig Dis
Sci 1993; 38: 665 ± 673.
29 Nauck MA, Niedereichholz U, Ettler R, Holst JJ, érskov C,
Ritzel R, Schmiegel WH. Glucagon-like peptide-1 inhibition
of gastric emptying outweighs its insulinotropic effects in
healthy humans. Am J Physiol 1997; 273: E981 ± E988.
30 Khan MI, Read NW. The effect of duodenal lipid infusions
upon gastric pressure and sensory responses to balloon distension. Gastroenterology 1992; 102: 467 ± 473.
31 Sepple CP, Read NW. Gastrointestinal correlates of the
development of hunger in man. Appetite 1989; 13: 183 ± 191.
32 Schwartz GJ, Moran TH. Sub-diaphragmatic vagal afferent
integration of meal-related gastrointestinal signals. Neurosci
Biobehav Rev 1996; 20: 47 ± 56.
33 érskov C, Poulsen SS, Mùller M, Holst JJ. GLP-1 receptors in
the subfornical organ and the area postrema are accessible
to circulating glucagon-like peptide-1. Diabetes 1996; 45:
832 ± 835.
34 Navarro M, de Fonseca FR, Alvarez E, Chowen SA,
Zueco JA, Gomez R, Eng J, Blazquez E. Colocalization of
glucagon-like peptide-1 (GLP-1) receptors, glucose transporter
GLUT-2, and glucokinase mRNA in rat hyopthalamic cells:
evidence for a role of GLP-1 receptor agonists as an inhibitory signal for food and water intake. J Neurochem 1996;
67: 1982 ± 1991.
35 Kieffer TJ, McIntosh CHS, Pederson RA. Degradation of
glucose-dependent insulinotropic polypeptide and truncated
glucagon-like peptide 1 in vitro and in vivo by dipeptidyl
peptidase IV. Endocrinology 1995; 136: 3585 ± 3596.
36 Smith GP, Gibbs J. Satiating effects of cholecystokinin. Ann
NY Acad Sci 1994; 713: 236 ± 241.
37 Holst JJ, Schwartz TW, Lùvgreen NA. Dinurnal pro®le of
pancreatic polypeptide, pancreatic glucagon, gut glucagon
and insulin in human morbid obesity. Int J Obes 1983; 7:
529 ± 538.
38 Larsen J, Jallad N, Damsbo P. One-week continuous infusion
of GLP-1 (7 ± 37) improves glycemic control in NIDDM.
Diabetes 1996; 45 (Suppl 2): 233A (Abstr).
39 Donahey JCK, van Dijk G, Woods SC, Seeley RJ. Intraventricular GLP-1 reduces short- but not long-term food intake
or body weight in lean and obese rats. Brain Res 1998; 779:
75 ± 83.
40 Davis HR, Mullins DE, Pines JM, Hoos LM, France
CF, Compton DS, Graziano MP, Stbertz EJ, Strader CD,
Van Heek M. Effect of chronic central administration of
glucagon-like peptide-1 (7 ± 36) amide on food consumption
and body weight in normal and obese rats. Obes Res 1998;
6: 147 ± 156.
41 Scrocchi LA, Brown TS, MacLusky N, Brubaker PL,
Anerbach AB, Joyner AL, Drucker DJ. Glucose intolerance
but normal satiety in mice with a null mutation in the
glucagon-like peptide-1 receptor gene. Nat Med 1996; 2:
1254 ± 1258.
311