Leptin Treatment in Activity-Based Anorexia
Jacquelien J.G. Hillebrand, Maarten P. Koeners, Corine E. de Rijke, Martien J.H. Kas, and Roger A.H. Adan
Background: Activity-based anorexia (ABA) is considered an animal model of anorexia nervosa (AN). In ABA, scheduled feeding
together with voluntary access to a running wheel results in increased running wheel activity (RWA), hypophagia, and body weight
loss. Previously it was shown that leptin treatment reduced semi-starvation–induced hyperactivity in rats. The present study was
performed to confirm and extend this finding, to evaluate leptin’s effect on energy balance in ABA.
Methods: The effects of chronic leptin treatment (intracerebroventricular, 4 ?g/day) in ABA rats, ad libitum–fed running rats, and
sedentary rats exposed to ad libitum feeding or scheduled feeding were investigated.
Results: Leptin treatment decreased RWA in ABA rats. Additionally, leptin treatment reduced food intake and increased energy
expenditure by thermogenesis in ABA rats. Ad libitum–fed running/sedentary rats or food-restricted sedentary rats did not reduce
activity after leptin treatment, whereas all leptin-treated rats showed hypophagia. Body temperature was slightly increased in
leptin-treated food-restricted sedentary rats.
Conclusions: Although leptin treatment reduced RWA in ABA rats, it also prevented hypothermia and decreased food intake.
Altogether, this resulted in a stronger negative energy balance and body weight loss in leptin-treated ABA rats.
ature, gene expression
brand et al 2003; Kron et al 1978). Compared with other
psychiatric disorders, AN has the highest mortality rate (Sullivan
It is hypothesized that a biological drive or alterations in
certain physiologic parameters trigger food restriction and hy-
peractivity in AN. A candidate parameter is leptin, one of the
main peripheral signaling molecules regulating energy ho-
meostasis (Elmquist et al 1998; Schwartz et al 2000; Zhang et al
1994). Although genetic studies thus far have reported no
evidence for specific disturbances in the leptin gene in AN
patients (Hinney et al 1998), alterations in plasma leptin levels
might still interfere with the disease. Plasma leptin levels rapidly
decrease after body weight loss in humans and in rodents (Ahima
et al 1996; Maffei et al 1995) and can vary between extremely low
and within the low normal range in AN patients (Calandra et al
2003; Hebebrand et al 1997; Mantzoros et al 1997). On the other
hand, AN patients show relative hyperleptinemia during recov-
ery, as a result of rapid body weight gain (Hebebrand et al 1997;
Mantzoros et al 1997).
The activity-based anorexia (ABA) model is used to study
anorectic behavior in rodents and serves as an animal model of
AN (Hall and Hanford 1954; Routtenberg and Kuznesof 1967). In
ABA, voluntary wheel running in combination with scheduled
feeding leads to a paradoxical increase of running wheel activity
(RWA) and decrease of food intake, and as a result body weight
drops below 80%. Not only total RWA increases during ABA, but
the distribution of activity throughout the day changes as well.
norexia nervosa (AN) is a psychiatric disorder often
characterized by extreme hypophagia, body weight loss,
hyperactivity, and hypothermia (Casper et al 1991; Hebe-
Rats develop food-anticipatory activity, which in general takes
place 3 to 4 hours before scheduled feeding.
It has been reported by Exner et al (2000) that peripheral
leptin treatment (31 ?g/day) reduced hyperactivity in rats ex-
posed to the semi-starvation–induced hyperactivity model (SIH).
Leptin treatment also rescued rats from further hyperactivity
when SIH had already developed. Additionally, it was shown
that AN patients reported a higher subjective rating of motor
restlessness when their serum leptin levels were low. From these
data, it was proposed that leptin administration in acute AN
might reduce hyperactivity, thereby improving outcome of fur-
ther treatment. This proposal gives rise to some further ques-
tions, because the main physiologic role of leptin is to reduce
food intake. Leptin decreases food intake through stimulation of
transcript neurons and inhibition of neuropeptide Y (NPY)/agouti-
related protein (AgRP) neurons in the arcuate nucleus of the
hypothalamus (Cowley et al 2001; Friedman and Halaas 1998;
Schwartz et al 1997). Furthermore, it has been described that
leptin has metabolic effects (e.g., increased energy expenditure
and thermogenesis), thereby promoting weight loss independent
of hypophagia (van Dijk 2001). Whereas a reduction of hyper-
activity by leptin treatment would be beneficial to ABA (or SIH)
rats, a decline in food intake and increased thermogenesis would
be disadvantageous. The present study was performed to con-
firm and extend the findings of Exner et al by investigating the
effects of leptin treatment on RWA, food intake, and thermogen-
esis in rats exposed to the ABA model.
Methods and Materials
Female outbred Wistar WU rats (Harlan, Horst, The Nether-
lands) weighing 160 g upon arrival were individually housed in
a temperature- and humidity-controlled room (21°C ? 2°C)
under a 12-hour dark/light cycle (Zeitgeber time [ZT]12 ? lights
off). The ethics committee on the use and care of animals of
Utrecht University approved all described procedures. For ethical
reasons, it was decided that rats were to be removed from the
experiment when their body temperature was lower than 30°C
Leptin (rat, Sigma-Aldrich, Zwijndrecht, The Netherlands) (4
?g/day) was dissolved in sterile isotonic saline and was chroni-
From the Department of Pharmacology and Anatomy (JJGH, MPK, CEdR,
ical Center Utrecht, Utrecht; and Altrecht–Rintveld Eating Disorders
(JJGH), Zeist, The Netherlands.
Neuroscience, Department of Pharmacology and Anatomy, University
Medical Center Utrecht, Universiteitsweg 100, 3584 CG Utrecht, The
Netherlands; E-mail: firstname.lastname@example.org.
BIOL PSYCHIATRY 2005;58:165–171
© 2005 Society of Biological Psychiatry
cally infused (continuous for 5 days, 12 ?L/day) into the lateral
ventricle with osmotic minipumps (Alzet model 1007D, DURECT,
Cupertino California). This dose of leptin was previously shown
to result in physiologic effects without being aversive (Hulsey et
al 1998; Pal et al 2003; Thiele et al 1997).
One week after arrival, all rats received transmitters (TA10TA-
F40 Data Sciences International, St. Paul, Minnesota) in the
abdominal cavity under fentanyl/fluanisone (.1 mL/100 g IM;
Hypnorm, Janssen Pharmaceutica, Beerse, Belgium) and mid-
azolam (.05 mL/100 g IP; Dormicum, Hoffman-LaRoche, Mij-
drecht, The Netherlands) anesthesia. After surgery, rats were
treated with buprenorphin (.05 mL/100 g s.c.; Temgesic, Scher-
ing-Plough, Maarssen, The Netherlands) and saline (1 mL s.c.)
and were allowed to recover for 2 weeks.
For intracerebroventricular (ICV) surgery, rats were anesthe-
tized by fentanyl/fluanisone as indicated above, the head was
shaved, and the skull was exposed by a midline incision. A brain
infusion cannula (Alzet, Brain infusion kit 3–5 mm) was placed
into the lateral ventricle 1 mm lateral and 1 mm posterior from
bregma and fixed in place with two small screws and dental
cement. The cannula was connected by tubing (filled with
vehicle or leptin) to an osmotic minipump containing vehicle or
leptin. Minipumps were placed s.c. into the flank region of the rat
after overnight incubation at 37°C. After surgery, rats were
treated with buprenorphin and saline as indicated above.
The effects of leptin treatment were investigated in ABA rats,
ad libitum–fed running rats, ad libitum–fed sedentary rats, and
food-restricted sedentary rats. One week after arrival, all rats (n
? 64) received transmitters. After 2 weeks of recovery (day ?10),
32 rats were placed into running wheel cages for a training
period of 10 days with ad libitum food and water intake. Running
wheel activity was continuously registered with a cage registra-
tion program (Department of Biomedical Engineering, University
Medical Center Utrecht, The Netherlands). The remaining 32 rats
stayed in macrolon type 3 cages and had ad libitum access to
food and water. At the end of day ?2 (? light phase, ZT11),
transmitters of all rats were activated for baseline recordings of
body temperature and locomotor activity (LMA). At the end of
day ?1 (ZT11), running and sedentary rats were divided into
restricted (2? n ? 16) and ad libitum–fed (2? n ? 16) groups,
matched for body weight (average day-1: 224.0 ? 1.7 g) and
4-day RWA (average day-4: day-1: 4284.4 ? 155.8 revolutions).
After ICV surgery, food was removed from the cages of the
restricted groups (onset day 0, ZT12). The next days (days 1–4),
the restricted rats had 1-hour access to food (ZT12 to ZT13),
whereas water was continuously available. Body weight (ZT11)
and food intake (ZT13) were measured daily. At the end of day
4 (ZT11), rats were decapitated. Brains (from all rats) and
interscapular brown adipose tissue (BAT, from ABA rats) were
rapidly removed, quickly frozen in cold (?35°C) isopentane, and
stored at ?80°C.
In Situ Hybridization
Cryosections (coronal, 20 ?m) of the arcuate nucleus were
sliced with a cryostat (Leica, Rijswijk, The Netherlands) and
thaw-mounted onto ribonuclease-free Superfrost slides (Menzel,
Braunschweig, Germany). The slides were stored at ?80°C until
processing for in situ hybridization (ISH). All cryostat sections
were concurrently prepared for hybridization and used in the
same assay for each probe. Sections were fixed in 4% parafor-
maldehyde in phosphate buffered saline (PBS) for 10 min,
washed in PBS, pretreated with .25% acetic anhydride in .1 mol/L
triethanolamine, washed again in PBS, and dehydrated in graded
ethanol followed by 100% chloroform and 100% ethanol.33P-
labeled anti-sense ribonucleic acid (RNA) probes were made
with a 350-base pair (bp) rat POMC complementary deoxyribo-
nucleic acid (cDNA) fragment (Kas et al 2003), a 286-bp rat NPY
cDNA fragment (Ericsson et al 1987), and a 396-bp mouse AgRP
cDNA fragment (Kas et al 2003). The sections were hybridized
overnight at 72°C with 1 ? 106cpm probe in buffer containing
50% deionized formamide, 2? standard saline citrate (SSC), 10%
dextrane sulphate, 1? Denhardt’s solution, 5 mmol/L ethyl-
enediaminetetraacetic acid, and 10 mmol/L phosphate buffer,
after 5 min heating at 80°C. After hybridization, the sections were
washed in 5? SCC (short, 72°C) and .2? SSC (2 hours, 72°C) and
dehydrated in graded ethanol with 3 mol/L ammoniumacetate.
Sections were exposed to x-ray films (Kodak Bio-Max MR,
Kodak, Rochester, New York) for 5 days. The films were
developed, and film absorbance values (including a standard
curve) were semi-quantitatively analyzed with the Microcom-
puter Imaging Device (Imaging Research, St. Catharines, Ontario,
Quantitative Polymerase Chain Reaction
Total RNA was prepared from BAT of ABA rats with Trizol
Reagent (Invitrogen Breda, The Netherlands). Ribonucleic acid
was treated with deoxyribonuclease I and was reverse tran-
scribed to cDNA with oligodT and SuperScript II reverse tran-
scriptase (Invitrogen). The Lightcycler real time polymerase
chain reaction (PCR) system (Roche Diagnostics, Mannheim,
Germany) was used for amplification and quantification of
uncoupling protein 1 (UCP1) cDNA. Cyclophilin and ²-actin were
used as reference genes. An amount of cDNA corresponding to
40 ng of total RNA was amplified with the Mastermix from
Lightcycler-Faststart DNA Master SYBR Green I kit (Roche Diag-
nostics) and the appropriate primers. Optimal MgCl2concentra-
tions, annealing temperature, and optimal cDNA dilution (1:10)
were determined, resulting in PCR efficiencies of ?1.80 (Table
1). All samples were measured in duplicate. Expression of UCP1
was calculated as normalized ratio relative to a calibrator (Rn).
Thus, UCP1 expression was analyzed relative to reference genes,
and each sample was normalized to a calibrator sample (pooled
cDNA), which was run in the same experiment (Roche Diagnos-
All data are presented as mean ? SE. Data were analyzed with
SPSS 11.5 for Windows (SPSS, Chicago, Illinois) and were
controlled for normality and homogeneity.
One rat (leptin ABA) was removed from the experiment
before day 4 for ethical reasons and was excluded from analysis.
For all measurements, baseline levels were not significantly
different between vehicle-treated and leptin-treated groups.
Body temperature was measured with telemetry and analyzed as
average body temperature during at least 30 min of inactivity in
the light phase (ZT0–ZT3) (basal body temperature). Locomotor
activity was also measured by telemetry and includes influences
of RWA. Relative body weight, food intake, RWA, basal body
temperature, and LMA were analyzed by general linear model
repeated-measures analysis with Huynh-Feldt correction for
Mauchlys sphericity effects, followed by t tests. Final body
weight, cumulative (day 0–4) food intake, total LMA, ISH data,
166 BIOL PSYCHIATRY 2005;58:165–171
J.J.G. Hillebrand et al
and quantitative PCR data were analyzed by independent t tests.
Differences were considered significant at p ? .05.
Effects of Leptin Treatment on Activity
Leptin treatment (4 ?g/day) significantly decreased RWA over
time in ABA rats [day: F(4,52) ? 20.38, p ? .001; day ? treatment:
F(4,52) ? 15.37, p ? .001]. Running wheel activity was reduced
during the dark phase [day: F(4,52) ? 15.24, p ? .001; day ?
treatment: F(4,52) ? 10.88, p ? .001] as well as during the light
phase [day: F(4,52) ? 11.21, p ? .001; day ? treatment: F(4,52)
? 9.87 p ? .001]. Whereas vehicle-treated ABA rats developed
hyperactivity and food-anticipatory activity in the hours before
feeding, leptin-treated ABA rats showed low levels of RWA.
In contrast, leptin treatment (4 ?g/day) did not significantly
affect RWA over time in ad libitum–fed rats [day: F(4,56) ? 6.66,
p ? .01; day ? treatment: F(4,56) ? .03, ns]. Neither running
during the dark phase nor running during the light phase was
affected in ad libitum–fed rats [dark day: F(4,56) ? 6.89, p ? .01;
day ? treatment: F(4,56) ? .02, ns; light day: F(4,56) ? 3.82, p ?
.03; day ? treatment: F (4,56) ? .76, ns) (Figure 1).
Locomotor activity measurements by telemetry allowed loco-
motion assessment in running rats and sedentary rats. Total LMA
was significantly decreased by leptin treatment in ABA rats [t(9)
? 2.14, p ? .03] but not in food-restricted sedentary rats [t(12) ?
?.75, ns]. Similar to that seen in the RWA data, leptin treatment
did not affect total LMA in ad libitum–fed running rats [t(14) ?
.40, ns] nor in ad libitum–fed sedentary rats [t(14) ? ?1.08, ns]
Effects of Leptin Treatment on Food Intake and Body Weight
Leptin treatment decreased cumulative food intake in ABA
rats [t(13) ? 9.37, p ? .001] and tended to decrease body weight
[t(13) ? 1.87, ns]. Ad libitum–fed running and ad libitum–fed
sedentary rats decreased food intake [t(14) ? 4.27, p ? .001 and
t(14) ? 7.48, p ? .001, respectively] and body weight [t(14) ?
Table 1. Primer Sequences, Magnesium Chloride (MgCl2) Concentration, Annealing Temperatures, Anticipated Size of the Amplified Products, and
GenBank Accession Codes of the Different Genes Studied by Quantitative Polymerase Chain Reaction
Size (bp) GenBank
UCP1 Forward CCACATAGGCGACTTGGA
4 63 79NM_012682
Figure 1. Running wheel activity (RWA) in rats after
vehicle or leptin (4 ?g/day) treatment (A, B) Distri-
bution of RWA during the day in (A) activity-based
anorexia (ABA) rats and (B) ad libitum–fed rats with
n ? 7/n ? 8). Leptin infusion started at day 0. (C, D)
or leptin treatment (white, n ? 7/n ? 8). (E, F) Total
light phase RWA per day in (E) ABA rats and (F) ad
libitum–fed rats with vehicle (black, n ? 8/n ? 8) or
leptin treatment (white, n ? 7/n ? 8). *Different
tests, p ? .05.
J.J.G. Hillebrand et al
BIOL PSYCHIATRY 2005;58:165–171 167
4.32, p ? .001 and t(14) ? 5.52, p ? .001, respectively] after
leptin treatment, whereas leptin-treated food-restricted sedentary
rats showed a decreased cumulative food intake [t(14) ? 2.40, p
? .03] with unchanged body weight [t (14) ? 1.23, ns] (Table 2).
Effects of Leptin Treatment on Thermogenesis
Body temperature was analyzed during periods of physical
inactivity in the light phase. As a result of negative energy
balance, body temperature dropped in ABA rats and food-
restricted sedentary rats. Leptin treatment significantly affected
body temperature over time in ABA rats [day F(4,40) ? 31.5, p ?
.001; day ? treatment: F(4,40) ? 13.65, p ? .01]. Starvation-
induced hypothermia was prevented by leptin treatment during
days 0–3 (all p ? .03); however, on day 4 body temperature of
leptin-treated ABA rats suddenly (within 1 day) dropped as
compared with vehicle-treated ABA rats (p ? .03). In food-
restricted sedentary rats, body temperature was also influenced
by leptin treatment over time [day: F(4,48) ? 49.48, p ? .001; day
? treatment: F(44,48) ? 12.52, p ? .001]. Leptin treatment
prevented hypothermia on days 2–4 (all p ? .01) in food-
restricted sedentary rats. In contrast, leptin treatment did not
influence body temperature over time in ad libitum–fed running
[day F(4,56) ? 19.50, p ? .001; day ? treatment: F(4,56) ? 1.62,
ns] or in ad libitum–fed sedentary rats [day: F(4,56) ? 17.98, p ?
.001; day ? treatment: F (4,56) ? .83, ns] (Figure 2).
Because leptin-treated ABA rats showed a sudden drop of
body temperature on day 4, the experiment was terminated.
Severe exhaustion of the leptin-treated rats was visible by
absence of visceral fat pads and extreme reddish BAT. Therefore,
the expression of UCP1 in BAT was analyzed by quantitative
PCR. UCP1 expression was 2.4-fold increased in leptin-treated
ABA rats (Rn? 4.0 ? .6) as compared with vehicle-treated ABA
rats (Rn? 1.7 ? .5) [t(6) ? ?2.89, p ? .03].
Effects of Leptin Treatment on Arcuate Nucleus Gene
In situ hybridization on brain slices showed that leptin
treatment increased POMC messenger RNA (mRNA) levels [t(11)
? ?3.51, p ? .01] and reduced AgRP [t(11) ? 2.99, p ? .03] and
NPY mRNA levels [t(11) ? 2.68, p ? .04] in ABA rats. Food-
restricted sedentary rats treated with leptin also had increased
POMC mRNA levels [t(9) ? ?2.61, p ? .03] and decreased AgRP
mRNA levels [t(9) ? 2.25, p ? .05], but NPY mRNA levels were
not significantly affected [t(9) ? 1.22, ns]. Leptin treatment in ad
libitum–fed running and ad libitum–fed sedentary rats only
Table 2. Cumulative Food Intake, Relative Body Weight, and Total Locomotor Activity (LMA) After 5 Days of Leptin (4 ?g/Day) or Vehicle Treatment in
Activity-Based Anorexia (ABA) Rats, Ad Libitum–Fed Running Rats, Food-Restricted Sedentary Rats, and Ad Libitum–Fed Sedentary Rats
Cumulative Food Intake
(Day 0–4) (g)
Final Body Weight
(% of Day ?1)
ABA Vehicle (n ? 8)
Leptin (n ? 7)
Vehicle (n ? 8)
Leptin (n ? 8)
Vehicle (n ? 8/6)
Leptin (n ? 8)
Vehicle (n ? 8/6)
Leptin (n ? 8/5)
20.7 ? 1.1
9.2 ? .3a
64.0 ? 2.8
44.7 ? 3.9a
15.4 ? 1.0
10.6 ? .9a
53.9 ? 2.4
29.4 ? 1.9a
81.3 ? 1.6
76.2 ? 1.3
101.4 ? .6
93.8 ? 1.6a
87.8 ? .6
86.8 ? .6
101.2 ? 1.0
93.5 ? 1.0a
4549.2 ? 1142.2
1819.9 ? 182.5a
3840.2 ? 627.7
3566.6 ? 268.9
1306.3 ? 61.7
1397.0 ? 93.5
1464.5 ? 53.7
1647.7 ? 161.4
Ad Libitum Running
Ad Libitum Sedentary
In some groups, LMA was not obtained from each rat.
aDifferent from vehicle treatment. T test, p ? .05.
Figure2.Body temperature in rats after vehicle or
leptin (4 ?g/day) treatment. Body temperature
was measured during 30 min of inactivity in the
early light phase. (A) Body temperature per day in
activity-based anorexia (ABA) rats after vehicle
(black, n ? 7) or leptin (white, n ? 5) treatment.
(B) Body temperature per day in ad libitum–fed
running rats after vehicle (black, n ? 8) or leptin
(white, n ? 8) treatment. (C) Body temperature
per day in food-restricted rats after vehicle (black,
n ? 6) or leptin (white, n ? 8) treatment. (D) Body
temperature per day in ad libitum–fed sedentary
rats after vehicle (black, n ? 8) or leptin (white, n
? 8) treatment. *Different from vehicle. Repeated
measurements followed by t tests, p ? .05. Note a
different y-axis in A.
168 BIOL PSYCHIATRY 2005;58:165–171
J.J.G. Hillebrand et al
resulted in trends towards upregulation of POMC and downregu-
lation of AgRP and NPY gene expression (Figure 3).
Leptin treatment suppressed RWA in the ABA model. This
effect was specific for ABA rats, because RWA and LMA were not
decreased in ad libitum–fed (running and sedentary) rats after
leptin treatment. Additionally, leptin treatment did not affect LMA
in food-restricted sedentary rats.
The effect of leptin treatment (4 ?g/day ICV) on RWA in ABA
rats seems to be solid and strong and confirms the suppression of
RWA in SIH as reported before (Exner et al 2000). The un-
changed activity levels in leptin-treated ad libitum–fed rats also
corresponds with earlier reports (Fox and Olster 2000; Surwit et
al 2000). Food-restricted sedentary rats did not reduce activity
when treated with leptin. Thus, reduced activity after leptin
treatment was specific for ABA. The mechanism underlying the
activity-reducing effects of leptin treatment in ABA rats is not
clear. Previously it was shown that leptin reverses food-restric-
tion–induced sensitization to brain stimulation reward (Carr
1996; Fulton et al 2000). Leptin also attenuates heroin seeking
after food restriction (Shalev et al 2001). This suggests that leptin
plays a role in behavioral allocation, which in the ABA model
would mean that leptin might decrease the rewarding value of
wheel running in ABA. The presence of leptin receptors in the
midbrain dopamine system (a system involved in motivational
behavior) and reduced activity of midbrain dopamine neurons
after leptin treatment support this view (Figlewicz et al 2003;
Krugel et al 2003).
The effect of leptin treatment on RWA in ABA rats was more
pronounced than described before in SIH (Exner et al 2000). This
might be explained by differences in the feeding schedules that
were used (60% of baseline food intake vs. 1-hour feeding) and
dose and route of leptin administration (peripherally 31 ?g/day
vs. ICV 4 ?g/day). Furthermore, we used female rats, which are
relatively more sensitive to leptin treatment than male rats (Clegg
et al 2003).
Leptin treatment decreased food intake in ad libitum–fed
(running and sedentary) rats but also in food-restricted sedentary
rats and even more in ABA rats, indicating that leptin has large
effects on feeding behavior even during a strong homeostatic
drive to eat (Ahima et al 1996; Velkoska et al 2003). Despite the
negative energy balance due to ABA, leptin still increased POMC
and decreased AgRP and NPY mRNA expression in the arcuate
nucleus. Similar results for POMC and AgRP mRNA expression
were obtained in food-restricted sedentary rats; however, NPY
expression was not significantly changed.
Leptin treatment prevented the starvation-induced decrease of
body temperature during the first days of ABA. Thus, on one hand
leptin-treated ABA rats showed reduced energy output due to
relative hypoactivity, whereas on the other hand they showed
increased energy output by generating heat. The experiment
showed that the effects of the reduced energy expenditure through
and the relative hyperthermia. On day 4, severe hypothermia was
observed in leptin-treated ABA rats, and therefore the experiment
was terminated. Whereas leptin treatment did not influence body
temperature in ad libitum“fed rats (running and sedentary), it did
prevent the starvation-induced decrease of body temperature in
food-restricted rats without running wheels. Thermogenic effects of
leptin have been described before (van Dijk 2001). Leptin deficient
temperature in a cold environment. Leptin treatment of ob/ob mice
increases body temperature (Harris et al 1998). The thermogenic
response of leptin is attributed to increased sympathetic activation
of BAT thermogenesis (Collins et al 1996; Satoh et al 1998). Indeed,
increased expression of UCP1 in BAT of leptin-treated ABA rats as
compared with vehicle-treated ABA rats was observed.
Recently it was demonstrated that chronic ICV AgRP(83-132)
treatment increased survival in ABA rats (Kas et al 2003).
Regarding inhibitory actions of leptin on AgRP expression and
opposite actions of leptin and AgRP on energy intake, the
suppressing effects of leptin on SIH (Exner et al 2000) and the
stimulatory effects of AgRP(83-132)on survival in the ABA model
(Kas et al 2003) seemed contradictory before. Here it was shown,
however, that leptin treatment in ABA positively influenced
energy balance by decreasing RWA, but at the same time it
negatively influenced energy balance in ABA by decreasing food
intake and increasing thermogenesis through increased UCP1
expression. Previously, Kas et al (2003) reported that AgRP(83-132)
Figure 3. Pro-opiomelanocortin (POMC), agouti-re-
lated protein (AgRP), and neuropeptide Y (NPY)
(4 ?g/day). (A) Arcuate nucleus gene expression of
POMC, AgRP, and NPY in ABA rats after vehicle
(black, n ? 6) or leptin (white, n ? 7) treatment. (B)
Arcuate nucleus gene expression of POMC, AgRP,
and NPY in ad libitum-fed running rats after vehicle
(black, n ? 7) or leptin (white, n ? 7) treatment. (C)
Arcuate nucleus gene expression of POMC, AgRP,
? 5) or leptin (white, n ? 6) treatment. (D) Arcuate
ad libitum-fed sedentary rats after vehicle (black, n
? 7) or leptin (white, n ? 8) treatment. *Different
from vehicle. T test, p ? .05.
J.J.G. Hillebrand et al
BIOL PSYCHIATRY 2005;58:165–171 169
positively influenced survival by increasing food intake without
affecting RWA. Treatment with AgRP(83-132)also inhibited starva-
tion-induced hypothermia (Kas et al 2003). It has also been
reported, however, that AgRP(83-132)treatment decreases BAT
UCP1 levels (Small et al 2001), opposite to the effect of leptin
(Satoh et al 1998). Thus, the beneficial effects of AgRP(83-132)on
ABA are mediated differently than by leptin, which only by
decreasing RWA positively affects energy balance (and thus
The strong effects of leptin treatment on food intake, RWA, and
treatment. Reduced RWA after leptin treatment supports the notion
that a decrease in serum leptin levels triggers hyperactivity in ABA
rats. Also in AN, a relationship between (endogenous) serum leptin
levels and activity levels has been demonstrated. Anorexia nervosa
patients display the highest activity levels when their serum leptin
levels are lowest, whereas their activity levels are decreased when
body weight and leptin levels are increased (Exner et al 2000;
Holtkamp et al 2003). Variances in hyperactivity levels among AN
patients correlate with serum leptin levels but not with body mass
index (BMI), suggesting that hypoleptinemia, rather than a reduced
BMI, might underlie physical hyperactivity (Holtkamp et al 2003).
After weight gain, AN patients show relative hyperleptinemia,
which seems to be related to renewed weight loss and relapse
(Hebebrand et al 1997; Holtkamp et al 2004; Mantzoros et al 1997).
In summary, we demonstrated that leptin treatment strongly
decreased RWA in ABA; however, decreased RWA was outweighed
by reduced food intake and increased energy output by thermo-
genesis. Food intake suppression and increased thermogenesis
resulted in a rapid worsening of the physical state of the rats. Thus,
in the present study, with prolonged scheduled feeding, chronic
leptin treatment was not effective in rescuing rats from ABA.
Animal data from the present study and from Exner et al are
convincing and show that hyperactivity in animals receiving sched-
uled feeding can be reduced by leptin treatment. One should,
however, be careful in extrapolating these findings to the human
situation of AN. Although endogenous leptin levels in AN patients
has yet been provided that leptin therapy also reduces hyperactivity
in humans. Considering the animal data, leptin might have great
potential as a medication for treatment of severe hyperactivity (and
potentially for severe hypothermia) in AN patients; however, if
earlier death. Initial treatment in AN patients should ensure a high
food intake and body weight gain. Considering the fact that this is
already difficult to achieve in medication-free AN patients, treating
these patients with a strong anorexic agent, like leptin, provides a
serious additional concern. We believe, however, that leptin treat-
ment could be beneficial in extremely active AN patients by
decreasing hyperactivity and making them more susceptible to
further treatment. As indicated above, sufficient food intake (e.g.,
nasogastric) and possibly a heated environment during leptin
treatment should be ensured.
We thank J. H. Brakkee for assistance with the animal work.
JJGH was supported by Netherlands Organization of Scientific
Research grant 9033175, The Netherlands.
Ahima RS, Prabakaran D, Mantzoros C, Qu D, Lowell B, Maratos-Flier E, et al
(1996): Role of leptin in the neuroendocrine response to fasting. Nature
esis of anorexia nervosa and bulimia. Eat Weight Disord 8:130–137.
Carr KD (1996): Feeding, drug abuse, and the sensitization of reward by
metabolic need. Neurochem Res 21:1455–1467.
Casper RC, Schoeller DA, Kushner R, Hnilicka J, Gold ST (1991): Total daily
Clegg DJ, Riedy CA, Smith KA, Benoit SC, Woods SC (2003): Differential
sensitivity to central leptin and insulin in male and female rats. Diabetes
leptin in fat regulation. Nature 380:677.
Cowley MA, Smart JL, Rubinstein M, Cerdan MG, Diano S, Horvath TL, et al
(2001): Leptin activates anorexigenic POMC neurons through a neural
network in the arcuate nucleus. Nature 411:480–484.
nervous system pathways underlying responses to leptin. Nat Neurosci
et al (1987): Detection of neuropeptide Y and its mRNA in megakaryo-
cytes: Enhanced levels in certain autoimmune mice. Proc Natl Acad Sci
(2000): Leptin suppresses semi-starvation induced hyperactivity in rats:
Implications for anorexia nervosa. Mol Psychiatry 5:476–481.
Figlewicz DP, Evans SB, Murphy J, Hoen M, Baskin DG (2003): Expression of
nigra (VTA/SN) of the rat. Brain Res 964:107–115.
Fox AS, Olster DH (2000): Effects of intracerebroventricular leptin adminis-
rats. Horm Behav 37:377–387.
Friedman JM, Halaas JL (1998): Leptin and the regulation of body weight in
mammals. Nature 395:763–770.
by leptin. Science 287:125–128.
Hall JF, Hanford PV (1954): Activity as a function of a restricted feeding
Harris RB, Zhou J, Redmann SM Jr, Smagin GN, Smith SR, Rodgers E, et al
(1998): A leptin dose-response study in obese (ob/ob) and lean (?/?)
Hebebrand J, Blum WF, Barth N, Coners H, Englaro P, Juul A, et al (1997):
Leptin levels in patients with anorexia nervosa are reduced in the acute
stage and elevated upon short-term weight restoration. Mol Psychiatry
(1998): No evidence for involvement of the leptin gene in anorexia
nervosa, bulimia nervosa, underweight or early onset extreme obesity:
Identification of two novel mutations in the coding sequence and a
novel polymorphism in the leptin gene linked upstream region. Mol
(2004): High serum leptin levels subsequent to weight gain predict
renewed weight loss in patients with anorexia nervosa. Psychoneuroen-
al (2003): Elevated physical activity and low leptin levels co-occur in
(i.c.v.) administration of mouse leptin in rats: Behavioral specificity and
effects on meal patterns. Physiol Behav 65:445–455.
Kas MJ, van Dijk G, Scheurink AJ, Adan RA (2003): Agouti-related protein
prevents self-starvation. Mol Psychiatry 8:235–240.
Kron L, Katz JL, Gorzynski G, Weiner H (1978): Hyperactivity in anorexia
nervosa: A fundamental clinical feature. Compr Psychiatry 19:433–440.
Krugel U, Schraft T, Kittner H, Kiess W, Illes P (2003): Basal and feeding-
in obese and weight-reduced subjects. Nat Med 1:1155–1161.
170 BIOL PSYCHIATRY 2005;58:165–171
J.J.G. Hillebrand et al
Mantzoros C, Flier JS, Lesem MD, Brewerton TD, Jimerson DC (1997): Cere- Download full-text
brospinal fluid leptin in anorexia nervosa: correlation with nutritional
status and potential role in resistance to weight gain. J Clin Endocrinol
Pal R, Sahu A (2003): Leptin signaling in the hypothalamus during chronic
central leptin infusion. Endocrinology 144:3789–3798.
Routtenberg A, Kuznesof AW (1967): Self-starvation of rats living in activity
Satoh N, Ogawa Y, Katsuura G, Numata Y, Masuzaki H, Yoshimasa Y, et al
(1998): Satiety effect and sympathetic activation of leptin are mediated
by hypothalamic melanocortin system. Neurosci Lett 249:107–110.
Schwartz MW, Seeley RJ, Woods SC, Weigle DS, Campfield LA, Burn P, et al
(1997): Leptin increases hypothalamic pro-opiomelanocortin mRNA ex-
pression in the rostral arcuate nucleus. Diabetes 46:2119–2123.
Schwartz MW, Woods SC, Porte D Jr, Seeley RJ, Baskin DG (2000): Central
nervous system control of food intake. Nature 404:661–671.
induced relapse to heroin seeking. J Neurosci 21:RC129.
Effects of chronic central nervous system administration of agouti-re-
lated protein in pair-fed animals. Diabetes 50:248–254.
Sullivan PF (1995): Mortality in anorexia nervosa. Am J Psychiatry 152:1073–
Surwit RS, Edwards CL, Murthy S, Petro AE (2000): Transient effects of long-
term leptin supplementation in the prevention of diet-induced obesity
in mice. Diabetes 49:1203–1208.
Central infusion of GLP-1, but not leptin, produces conditioned taste
aversions in rats. Am J Physiol 272:R726–R730.
van Dijk G (2001): The role of leptin in the regulation of energy balance and
adiposity. J Neuroendocrinol 13:913–921.
Velkoska E, Morris MJ, Burns P, Weisinger RS (2003): Leptin reduces food
Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM (1994):
J.J.G. Hillebrand et al
BIOL PSYCHIATRY 2005;58:165–171 171