Central Regulation of Energy Homeostasis
Intelligent Design: How to Build the Perfect
Barry E. Levin
homeostasis intelligent design: how to build the perfect
survivor. Obesity. 2006;14(Suppl 5):192S–196S.
The perfect survivor must be able to eat and store as many
calories as possible when food is readily available as a
buffer against periods of scarcity. He must also reduce
energy expenditure when food is scarce and efficiently and
accurately restore lost adipose stores when food is again
available. These processes are dependent on information
relayed to a distributed central network of metabolic sensing
neurons through hard-wired neural, metabolic, and hor-
monal signals from the periphery. These sensing neurons
engage neuroendocrine, autonomic, and motor processes
involved in arousal, motor activity, and the ingestion, ab-
sorption, assimilation, storage, and expenditure of calories.
A raised threshold in these metabolic sensors for detecting
inhibitory signals from increasing adipose stores allows
continued intake of excess calories when they are readily
available. Unfortunately, this mechanism for surviving pe-
riods of feast and famine predisposes the perfect survivor to
become obese when highly palatable, energy dense foods
are readily available at low energetic cost. It further assures
that raised adipose stores are metabolically defended against
attempts to lower them. Thus, effective treatment of obesity
will only come with a better understanding of the physio-
logical, metabolic, and neurochemical processes that ensure
this defense of an elevated body weight.
BARRYE. Central regulationof energy
Key words: diet-induced obesity, set-point, metabolic
sensing, leptin, insulin
Survival of the species ultimately depends on the ability
of individual to eat and to reproduce. Of these, only inges-
tion is required for the survival of the individual. Most
mammals, including humans, evolved in an environment
where food was only intermittently available (1). Thus, the
perfect survivor has evolved central and peripheral systems
that are biased toward ingesting and storing as much food as
possible when it is available, conserving those stores when
it is not, and efficiently replenishing lost stores when food
is available again. Together, this set of traits describes the
“thrifty genotype” (2). However, the same traits that make
some individuals perfect survivors are those that predispose
them to become obese when highly palatable, energy dense
foods are readily available and can be obtained with mini-
mum expenditure of energy. The hypothesis of this review
is that the thrifty genotype depends on a raised threshold for
sensing and responding to catabolic signals from the periph-
ery such as leptin and insulin that normally limit food intake
and increase energy expenditure when adipose stores in-
crease. To gain the insights necessary to treat obesity effec-
tively, we must first have a better understanding of the
factors that promote this raised threshold. To do this we
must be able to answer five questions.
Question 1: What Are the Signals from the
Periphery and How Are They Sensed and
Integrated Within Systems that Regulate
The brain maintains a constant dialog with the external
environment through inputs from somatic sensation, taste,
smell, sight, and sound and the body by inputs from the
viscera. These signals are relayed to a variety of brain areas
through hard-wired neural connections and are comple-
mented by metabolic and hormonal inputs that reflect the
metabolic status of the body. In humans, there is also a
strong psycho-social component that affects ingestive be-
havior. All of these inputs are integrated within areas of the
Neurology Service, VA Medical Center, East Orange, New Jersey and Department of
Neurology and Neurosciences, New Jersey Medical School, Newark, New Jersey.
Address correspondence to Barry E. Levin, Neurology Service (127C), VA Medical Center,
385 Tremont Avenue, East Orange, NJ 07018-1095.
Copyright © 2006 NAASO
192S OBESITY Vol. 14 Supplement August 2006
brain involved in the control of energy homeostasis, motor
activity, memory and learning, motivation, reward, and
effect. The brain has evolved a specialized set of neurons
that integrate many of these signals from the periphery. First
described as “glucose sensing” neurons, it is now clear that
they are really metabolic sensors that have transporters,
enzymes, ion channels, and receptors that allow them to
sense and respond to signals from the periphery by altering
their membrane potential and firing rate (3). These meta-
bolic sensing neurons are clustered in sites scattered
throughout the brainstem and forebrain and are integrated
into a distributed network that links them to afferent and
efferent pathways involved in the control of energy ho-
(POMC) neurons in the hypothalamic arcuate nucleus are
prototypic metabolic sensors. Both use glucose as signaling
molecules, and both have receptors for peripheral hormones
such as insulin and leptin (4). The POMC neurons produce
?-melanocyte stimulating hormone whose release onto
melanocortin-3 and -4 receptors in the paraventricular nu-
cleus and lateral hypothalamus reduces food intake and
increases energy expenditure through their projections to
autonomic and neuroendocrine effector systems (5). Firing
of NPY neurons releases both NPY and agouti-related pep-
tide (AgRP); NPY is an anabolic peptide that strongly
stimulates ingestive behaviors and minimizes energy expen-
diture, whereas AgRP acts as a functional antagonist of
catabolic melanocortin receptors (5). Under homeostatic
conditions, leptin and insulin levels reflect the amount of
adiposity in the body (6). When their levels increase, they
produce a net catabolic state by inhibiting NPY and AgRP
and stimulating POMC gene transcription (5,6). During
states of negative energy balance, plasma levels of both
hormones fall rapidly so they no longer correlate with the
amount of adipose tissue. Reduction of leptin and insulin
disinhibits NPY and AgRP and inhibits POMC expression,
leading to a net anabolic state (5,6). Both neurons are also
dependent on the presence of insulin and leptin for out-
growth of their pathways during development and for rapid
changes in their synaptic connectivity in response to altered
energy homeostasis (7).
Question 2: Why Do We Eat Our Next Meal
at Times When We Are Not in a State of
Negative Energy Balance?
Short-term regulation of energy intake, expenditure and
storage engages a complex set of metabolic, physiological,
and behavioral functions that requires input from many
other metabolic-sensing neurons besides those which ex-
press NPY/AgRP and POMC. The individual must first
awaken and become hungry enough to forage for food.
Arousal is mediated by orexin (hypocretin), serotonin, and
norepinephrine neurons (8–10), whereas hunger and forag-
ing are regulated by a large variety of neuropeptides and
monoamines (6,11,12). Many animals also hoard food be-
fore ingestion, and this behavior is mediated by AgRP,
NPY, and opioids (13,14). Ingestion and absorption trigger
the release of hormones such as insulin that help assimilate
and store ingested foods and a host of gut peptides that act
as satiety signals (15). In rodents, ingestion is followed by
a “satiety sequence” in which animals groom and then
sleep; this sequence is also mediated by an additional set of
peptides and transmitters (15,16). Assimilation of food
leads to heat production by activation of the sympathetic
nervous system and metabolic processes within organs.
Excess ingested calories are stored primarily as glycogen
and fat. All of these carefully modulated processes are under
the control of the distributed network of metabolic sensing
neurons in the brain as modulated by feedback from the
Although we know a great deal about what happens when
we eat, it is less certain why we actually eat a given meal.
For example, we may eat breakfast because we are in a mild
state of negative energy balance, but it is unlikely that we
eat lunch for the same reason. There are several possible
reasons for short-term initiation of feeding: 1) a small
deficit of a metabolic substrate such as glucose; 2) with-
drawal of satiety factors; 3) circadian and ultradian rhythms
involved in light- and meal-entrained behaviors; 4) habit;
and 5) psychosocial factors. Any and all of these probably
play some role in initiation of individual meals in humans.
However, biorhythms are likely to play the most important
role. Both light and the temporal scheduling of meals can
entrain biological rhythms that regulate the function of
central and peripheral systems involved in energy ho-
Question 3: Why Do We Eat When We Are
in Negative Energy Balance?
When a state of negative energy balance develops be-
cause of a limited supply or increased expenditure, the body
must reduce its expenditure of calories and mobilize stored
substrate to fuel its metabolic needs. Reduced energy ex-
penditure is mediated by both a decline in sympathetically
mediated thermogenesis in tissues such as muscle and
brown adipose tissue (19) and activation of cellular energy
sensors such as 5?AMP-activated protein kinase, which
switch off ATP-consuming pathways and activate ATP-
regenerating pathways in response to cellular fuel depletion
(20). The end result of these changes is a reduction in
resting and diet-induced energy expenditure (21,22). On the
other hand, increased sympathetic activity is required to
1Nonstandard abbreviations: NPY, neuropeptide Y; POMC, proopiomelanocortin; AgRP,
Regulation of Energy Homeostasis, Levin
OBESITY Vol. 14 Supplement August 2006 193S
mobilize glycogen stores from liver and muscle and fatty
acids from adipose depots (23,24). Increased adipose sym-
pathetic activity causes leptin levels to fall (25). Similarly,
depletion of glycogen stores results in reduced insulin levels
as plasma glucose levels fall. At this point, leptin and
insulin levels no longer accurately reflect the status of
adipose stores, and their disproportionate decline disinhibits
central anabolic and inhibits catabolic systems. This se-
quence of metabolic events provokes arousal, hunger, and
foraging for food (5).
Question 4: What Systems Are Essential for
the Preservation and Restoration of Energy
Stores During and After Periods of Chronic
The systems engaged during chronic restriction of energy
stores are also those that operate during short-term depri-
vation. Reduced leptin and insulin levels, as well as reduced
substrate to fuel metabolic processes in organs and loss of
lean body mass, all contribute to a sustained reduction in
resting metabolic rate with establishment of a homeostatic
balance at a new lower level (22,26,27). The degree to
which this occurs is independent of the starting adipose
mass but is sustained until lost adiposity is restored
(22,27,28). Thus, even after losing 10% to 15% of their
body weight, obese individuals are left with a large adipose
stores but still maintain a net anabolic balance in neural
systems controlling energy homeostasis the brain (29–31).
When food is again readily available, this anabolic tone
stimulates hyperphagia, which, along with reduced energy
expenditure, is maintained until prerestriction adipose levels
are accurately restored (22,26,27). Such observations sug-
gest that there is some internal set-point that mediates this
very precise restoration of lost adipose stores even after
prolonged periods of deprivation. However, this set-point is
not fixed and can easily be moved upward by the relatively
slow accretion of calories (32). On the other hand, the
set-point rarely moves downward in the absence of some
intervention such as surgery, illness, or drug treatment. One
possible explanation for such a one-way movement is that
new neural pathways are established within the distributed
network of metabolic sensing neurons that permanently
resets the defended body weight at higher levels (33).
Question 5: Why Do Some Individuals Eat
Beyond Their Metabolic Needs?
This question lies at the heart of the problem of treating
obese individuals. Studies in rats selectively bred to develop
diet-induced obesity on a high-fat diet provide one answer
to this question. This rat is an excellent example of the
“perfect survivor.” When the caloric density of its diet is
increased, it fails to respond by reducing its intake despite
early increases in plasma leptin and insulin levels, which
should limit ingestion. It remains hyperphagic and eats
larger meals for several weeks until obesity is well estab-
lished (34,35). At that point, it avidly defends its new higher
level of adiposity against chronic caloric restriction
(22,26,27,36). This ability to eat beyond its metabolic needs
seems to be caused by an inborn elevation in its threshold
for detecting or responding to the inhibitory signals associ-
ated with rising insulin, leptin, and glucose levels (3,37–39).
These and a number of other “abnormalities” of neural
function are present before they become obese, and some
are normalized only when obesity develops (40). Their
elevated levels of adiposity are avidly defended, and their
brains become even more resistant to the inhibitory effects
of leptin and insulin caused by reduced transport across the
blood–brain barrier and further blunting of central signaling
pathways (26,32,36–39,41). Increased dietary fat content
further reduces their transport and central signaling (37,39).
It is likely that the blunted responses of diet-induced obese
rats to inhibitory feedback signals occur at the level of their
metabolic sensing neurons (42,43). Such findings suggest
that the diet-induced obese rat is genetically programmed so
that it can increase its adipose stores far above its metabolic
needs when energy dense, high-fat foods are abundant. This
is an excellent survival strategy when food in only inter-
mittently available but would promote the development of
obesity in such individuals in our modern human society.
Summary and Conclusions
Survival of the species is dependent on the development
of systems that drive the individual to seek and ingest food
and to conserve energy stores during times of low food
availability. However, those individuals capable of laying
down excess energy stores when food is readily available
have a competitive advantage during times of prolonged
famine. At least in the diet-induced obese rat, this is caused
by a raised threshold for sensing catabolic signals from the
periphery by specialized metabolic sensing neurons. Unfor-
tunately, this very trait that has allowed such individuals to
survive famine has become a potential seed of destruction.
As younger and younger children develop obesity, type 2
diabetes, and the metabolic syndrome, this may shorten
their life span sufficiently to prevent them from reproducing
and passing on their thrifty genotype. Clearly, we need to
develop better strategies for dealing with obesity and its
associated diseases to maintain a continued line of these
1. Diamond J. Guns, Germs and Steel. London: W.W. Norton,
2. Neel V. In diabetes mellitus: a “thrifty” genotype rendered
detrimental by “progress”? Am J Hum Genet. 1962;14:353–
Regulation of Energy Homeostasis, Levin
194SOBESITY Vol. 14 Supplement August 2006
3. Levin BE, Routh VH, Kang L, Sanders NM, Dunn-Mey-
nell AA. Neuronal glucosensing: what do we know after 50
years? Diabetes. 2004;53:2521–8.
4. Levin BE. Glucosensing neurons: the metabolic sensors of the
brain? Diabetes Nutr Metab. 2002;15:274–80.
5. Levin BE. The drive to regain is mainly in the brain. Am J
Physiol Regul Integr Comp Physiol. 2004;287:R1297–300.
6. Woods SC, Seeley RJ. Understanding the physiology of
obesity: review of recent developments in obesity research. Int
J Obes Relat Metab Disord. 2002;26(Suppl 4):S8–10.
7. Horvath TL, Diano S. The floating blueprint of hypothalamic
feeding circuits. Nat Rev Neurosci. 2004;5:662–7.
8. Hagan JJ, Leslie RA, Patel S, et al. Orexin A activates locus
coeruleus cell firing and increases arousal in the rat. Proc Natl
Acad Sci U S A. 1999;96:10911–6.
9. Hara J, Beuckmann CT, Nambu T, et al. Genetic ablation
of orexin neurons in mice results in narcolepsy, hypophagia,
and obesity. Neuron. 2001;30:345–54.
10. Lowry CA. Functional subsets of serotonergic neurones: im-
plications for control of the hypothalamic-pituitary-adrenal
axis. J Neuroendocrinol. 2002;14:911–23.
11. Cabanac M, Ferber C, Fantino M. Effects of dexfenflura-
mine on the feeding behavior of rats foraging in the cold for
palatable bait. Pharmacol Biochem Behav. 1989;32:1025–31.
12. Drazen DL, Wortman MD, Seeley RJ, Woods SC. Neu-
ropeptide Y prepares rats for scheduled feeding. Am J Physiol
Regul Integr Comp Physiol. 2005;288:R1606–11.
13. Day DE, Bartness TJ. Agouti-related protein increases food
hoarding more than food intake in Siberian hamsters. Am J
Physiol Regul Integr Comp Physiol. 2004;286:R38–45.
14. Fantino M, Cabanac M. Body weight regulation with a
proportional hoarding response in the rat. Physiol Behav.
15. Strader AD, Woods SC. Gastrointestinal hormones and food
intake. Gastroenterology. 2005;128:175–91.
16. Antin J, Gibbs J, Holt J, Young RC, Smith GP. Cholecys-
tokinin elicits the complete behavioral sequence of satiety in
rats. J Comp Physiol Psychol. 1975;89:784–90.
17. Bodosi B, Gardi J, Hajdu I, Szentirmai E, Obal F Jr,
Krueger JM. Rhythms of ghrelin, leptin, and sleep in rats:
effects of the normal diurnal cycle, restricted feeding, and
sleep deprivation. Am J Physiol Regul Integr Comp Physiol.
18. Otukonyong EE, Dube MG, Torto R, Kalra PS, Kalra SP.
High-fat diet-induced ultradian leptin and insulin hypersecre-
tion are absent in obesity-resistant rats. Obes Res. 2005;13:
19. Young JB, Landsberg L. Suppression of sympathetic ner-
vous system during fasting. Science. 1977;196:1473–5.
20. Ruderman NB, Saha AK, Vavvas D, Witters LA. Malonyl-
CoA, fuel sensing, and insulin resistance. Am J Physiol. 1999;
21. Lean ME, James WP. Metabolic effects of isoenergetic nu-
trient exchange over 24 hours in relation to obesity in women.
Int J Obes Relat Metab Disord. 1988;12:15–27.
22. MacLean PS, Higgins JA, Johnson GC, Fleming-Elder
BK, Peters JC, Hill JO. Metabolic adjustments with the
development, treatment, and recurrence of obesity in obesity-
prone rats. Am J Physiol Regul Integr Comp Physiol. 2004;
23. Beuers U, Jungermann K. Relative contribution of glyco-
genolysis and gluconeogenesis to basal, glucagon- and nerve
stimulation-dependent glucose output in the perfused liver
from fed and fasted rats. Biochem Int. 1990;21:405–15.
24. Migliorini RH, Garofalo MA, Kettelhut IC. Increased sym-
pathetic activity in rat white adipose tissue during prolonged
fasting. Am J Physiol. 1997;272:R656–61.
25. Ricci MR, Fried SK. Isoproterenol decreases leptin expres-
sion in adipose tissue of obese humans. Obes Res. 1999;7:
26. Levin BE, Dunn-Meynell AA. Defense of body weight
against chronic caloric restriction in obesity-prone and -resis-
tant rats. Am J Physiol. 2000;278:R231–7.
27. MacLean PS, Higgins JA, Johnson GC, et al. Enhanced
metabolic efficiency contributes to weight regain after weight
loss in obesity-prone rats. Am J Physiol. 2004;287:R1306–15.
28. Leibel RL, Rosenbaum M, Hirsch J. Changes in energy
expenditure resulting from altered body weight. N Eng J Med.
29. Leibel RL, Hirsch J. Diminished energy requirements in
reduced-obese patients. Metabolism. 1984;33:164–70.
30. Levin BE, Dunn-Meynell AA. Chronic exercise lowers the
defended body weight gain and adiposity in diet-induced
obese rats. Am J Physiol. 2004;286:R771–8.
31. Levin BE, Dunn-Meynell AA. Sibutramine alters the central
mechanisms regulating the defended body weight in diet-
induced obese rats. Am J Physiol. 2000;279:R2222–8.
32. Levin BE, Dunn-Meynell AA. Defense of body weight de-
pends on dietary composition and palatability in rats with
diet-induced obesity. Am J Physiol Regul Integr Comp
33. Levin BE. The obesity epidemic: metabolic imprinting on
genetically susceptible neural circuits. Obes Res. 2000;8:342–
34. Farley C, Cook JA, Spar BD, Austin TM, Kowalski TJ.
Meal pattern analysis of diet-induced obesity in susceptible
and resistant rats. Obes Res. 2003;11:845–51.
35. Levin BE, Dunn-Meynell AA, Ricci MR, Cummings DE.
Abnormalities of leptin and ghrelin regulation in obesity-
prone juvenile rats. Am J Physiol Endocrinol Metab. 2003;
36. Levin BE, Keesey RE. Defense of differing body weight
set-points in diet-induced obese and resistant rats. Am J
37. Clegg DJ, Benoit SC, Reed JA, Woods SC, Levin BE.
Reduced anorexic effects of insulin in obesity-prone rats and
rats fed a moderate fat diet. Am J Physiol Regul Integr Comp
38. Levin BE, Dunn-Meynell AA. Reduced central leptin sensi-
tivity in rats with diet-induced obesity. Am J Physiol Regul
Integr Physiol. 2002;283:R941–8.
39. Levin BE, Dunn-Meynell AA, Banks WA. Obesity-prone
rats have normal blood-brain barrier transport but defective
central leptin signaling prior to obesity onset. Am J Physiol
Regul Integr Physiol. 2004;286:R143–50.
Regulation of Energy Homeostasis, Levin
OBESITY Vol. 14 Supplement August 2006195S
40. Levin BE, Routh VH. Role of the brain in energy balance and
obesity. Am J Physiol. 1996;271:R491–500.
41. Kaiyala KJ, Prigeon RL, Kahn SE, Woods SC, Schwartz
MW. Obesity induced by a high-fat diet is associated with
reduced brain insulin transport in dogs. Diabetes. 2000;49:
42. Levin BE, Govek EK, Dunn-Meynell AA. Reduced glucose-
induced neuronal activation in the hypothalamus of diet-in-
duced obese rats. Brain Res. 1998;808:317–9
43. Song Z, Routh VH. Differential effects of glucose and lactate
on glucosensing neurons in the ventromedial hypothalamic
nucleus Diabetes. 2005;54:15–22.
Regulation of Energy Homeostasis, Levin
196S OBESITY Vol. 14 Supplement August 2006