Homeostatic and Non-homeostatic Pathways Involved in the Control of Food Intake and Energy Balance


  • Hans-Rudolf Berthoud

    Corresponding author
    1. Neurobiology of Nutrition Laboratory, Pennington Biomedical Research Center, Louisiana State University System, Baton Rouge, Louisiana.
      Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808. E-mail: berthohr@pbrc.edu
    Search for more papers by this author

Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808. E-mail: berthohr@pbrc.edu


A neural network sensitive to leptin and other energy status signals stretching from the hypothalamus to the caudal medulla has been identified as the homeostatic control system for the regulation of food intake and energy balance. While this system is remarkably powerful in defending the lower limits of adiposity, it is weak in curbing appetite in a world of plenty. Another extensive neural system that processes appetitive and rewarding aspects of food intake is mainly interacting with the external world. This non-homeostatic system is constantly attacked by sophisticated signals from the environment, ultimately resulting in increased energy intake in many genetically predisposed individuals. Recent findings suggest a role for accumbens–hypothalamic pathways in the interaction between non-homeostatic and homeostatic factors that control food intake. Identification of the neural pathways that mediate this dominance of cortico-limbic processes over the homeostatic regulatory circuits in the hypothalamus and brainstem will be important for the development of behavioral strategies and pharmacological therapies in the fight against obesity.


Obesity is now recognized as a major health problem in the United States, where >20% of the adult population is classified as clinically obese (BMI > 30 kg/m2) and about one half the population as overweight (BMI = 25 to 30 kg/m2) (1). Perhaps most disturbingly, childhood obesity is rapidly increasing in many countries. Strong correlations between BMI and development of type 2 diabetes, cardiovascular disease, gall bladder disease, osteoarthritis, sleep, and mental disorders have been clearly shown (1), and the fact that type 2 diabetes is now increasingly detected in obese teenagers should be most alarming.

Although important genes responsible for monogenic childhood obesities have been discovered recently (2), the rapidly changing environment in the post-industrial era and its associated lifestyle are increasingly recognized as the primary causes of obesity for the large majority of the population (3). Each of the many factors by which lifestyle and environment influence energy balance interacts with specific sets of susceptibility genes, the variations of which determine the physiological impact of the particular factor (4). Consequently, some individuals respond strongly (prone) and others weakly or not at all (resistant) to a given factor.

The initiation and maintenance of ingestive behavior is co-determined by metabolic and non-metabolic factors. Among the latter, environmental cues, as well as reward, cognitive, and emotional factors, play an important role, particularly in human food intake in the modern world. These non-homeostatic factors are processed mainly in cortico-limbic structures such as the prefrontal cortex, amygdala, and ventral striatum. Thus, an important issue is to understand how and where metabolic and non-metabolic factors are integrated to drive food intake.

Homeostatic Regulation of Energy Balance: Importance of Internal Signals and the Hypothalamus

Undoubtedly the most influential discovery for obesity research in the last decade was that of leptin, leptin receptors, and its downstream signaling pathways. This hormone, secreted from adipose tissue, promised to be the long-sought feedback signal regulating adipose tissue mass and body weight through concerted effects on food intake and energy expenditure. Clearly, the absence of leptin signaling in the brain is among the most potent stimulators of food intake, as shown by genetic lesions such as the ob/ob mouse and leptin-deficient humans and the fact that leptin treatment in such cases drastically reduces food intake. It is now clear that leptin has an important neurotrophic action on the development of feeding-related neurocircuitry in the hypothalamus (5, 6).

Hypothalamic peptide circuits are believed to be an essential part of the homeostatic energy balance regulatory system (7). Neurons in the mediobasal and lateral hypothalamus acting as integrative metabolic sensors generate output signals that drive the endocrine, autonomic, and behavioral systems, controlling energy intake and expenditure in a coordinated fashion. Many of the downstream pathways of these metabolic sensors in the hypothalamus have been identified. The paraventricular nucleus provides an interface to at least the pituitary endocrine axis and autonomic outflow through the medulla and spinal cord. Other specific cell groups in the lateral hypothalamus also connect with autonomic outflow systems in the brainstem.

Thus, there is an impressive neural regulatory feedback system to guarantee short- and long-term energy homeostasis. However, the overwhelming majority of obese humans have paradoxically high circulating leptin levels and very few respond favorably to exogenous leptin (8), suggesting a state of leptin resistance (9). Leptin treatment as an adjunct to moderate food intake restriction shows more promising effects, probably by counteracting the behavioral and metabolic adaptations that accompany weight loss attempts (10).

It is clear that leptin and all of the other negative feedback signals are unable to prevent the development of obesity of many humans in the modern environment and in certain animal models using human-style cafeteria and highly palatable diets. A number of explanations have been offered for the state of leptin resistance. There are two fundamentally opposing views. One view sees leptin resistance as a condition that naturally evolved (9), the other, more like a pharmacologically treatable disease state brought about by either dietary (11), early life experience, and/or genetic factors (12).

Non-homeostatic Eating: Importance of Food Reward and Cortico-limbic Structures

Particularly in humans, the initiation of a meal often starts as a purely cognitive/executive decision from the cortex in the absence of any depletion signal. Thus, even in the presence of satiety and replete energy stores, it seems to be easy for the cortex and limbic system to overpower the hypothalamus into an ingestive mode. De Castro and Plunkett (13) used computer simulation to show that food intake levels change depending not only on the internal milieu but also on the external environment, that environmental changes result in new levels of intake, and that inherited individual differences in responsiveness to environmental factors can markedly influence the levels obtained.

With regard to food intake, the major components of this non-homeostatic system are cephalic and intestinal feedforward mechanisms, the abundance of food cues in the modern environment, and the easy availability (low physical effort and cost) of palatable, energy dense foods (snacks) in a socially enhanced environment. Palatability and pleasantness are arguably the most powerful determinants of food intake. Experiencing or feeling pleasure is regarded as one of the human emotions that has been difficult to define both in psychological and neurological terms. Most researchers agree that emotions evolved from mechanisms that made animals engage in behaviors with a beneficial outcome (14). Specifically applied to food intake, the positive emotion or pleasure of tasting sweet (sugars and certain amino acids) or creamy (fat-rich) foods may have evolved to guarantee sufficient intake of varied foods and high-energy foods (15). Although we can follow gustatory processing from the brainstem through the secondary taste area in the insular cortex to a tertiary taste area in the orbitofrontal cortex, it is not well understood how the unconditioned or predicted reward value of pleasurable taste and flavor guides ingestive behavior. Berridge and Robinson (16) have outlined the potential psychological components that constitute reward into learning, liking, and wanting.

Mogenson et al. (17) was the first to recognize that the nucleus accumbens with its afferent and efferent connections may provide an interface between motivation and behavioral action, and their basic idea has been further developed more recently (18). Specifically, Zahm (18) has presented an integrative neuroanatomical perspective and proposed a convincing conceptual framework implicating nucleus accumbens circuitry in general adaptive responding. Significant projections from the nucleus accumbens to mainly the lateral hypothalamus have also been shown (19, 20).

In a comprehensive series of studies, the group of Kelley has shown that blockade of AMPA and kainate glutamate receptors (21), activation of α-γ-aminobutyric acid receptors (22, 23, 24), or activation of μ-opioid receptors (25, 26, 27, 28) in specific areas of the nucleus accumbens stimulates robust feeding responses. While the γ-aminobutyric acid agonist muscimol elicited intake of both high-carbohydrate and high-fat food equally (22), the μ-opioid agonist (d-Ala2-N-methyl-Phe-Gly-ol5)-enkephalin (DAMGO)1 preferentially stimulated intake of high-fat food, even in rats that had a slight preference for the high-carbohydrate diet (26), and of sucrose solution but not water (15). When an instrumental response was required to obtain sucrose pellets in non–food-deprived rats to estimate incentive motivation to eat, intra-accumbens DAMGO but not muscimol injections induced a marked increase in the breakpoint on a progressive ratio schedule of reinforcement (29). Maldonado-Irizarry et al. (21) further showed that microinjection of muscimol into the lateral hypothalamus completely blocked chow intake elicited by accumbens injection of the AMPA antagonist 1,2,3,4-tetrahydro-6-nitro-2,3 dioxobenzo [f] quinoxaline-7-sulfonamide and high-fat intake by antibody-coated bacteria (Acb) injection of DAMGO (25).

Based on these observations, we hypothesized that accumbens-hypothalamus projections might engage the hypothalamic peptidergic systems known to be involved in homeostatic appetite control and that this might be an important pathway for the “cognitive” and “emotional” brain to override homeostatic regulation (Figure 1). First, we found that in the lateral hypothalamus, particularly the perifornical region, orexin, but not melanin-concentrating hormone neurons, were activated and in the arcuate nucleus, neuropeptide Y neurons were activated, and proopiomelanocortin/cocaine- and amphetamine-regulated transcript neurons were deactivated by accumbens manipulation with muscimol (30). Using anterograde tracing from behaviorally characterized injection sites in the nucleus accumbens, we found considerable penetration of most of the hypothalamus with traced axon profiles in “eaters” but much fewer axons in “non-eaters.” Anterogradely labeled axons from the nucleus accumbens made close anatomical contacts with orexin, melanin-concentrating hormone, and cocaine- and amphetamine-regulated transcript neurons. Furthermore, voracious intake of a high-fat diet in presatiated rats after accumbens injection of the μ-opioid agonist DAMGO depended on Y1R and OX1R transmission, as tested with selective antagonists (Zheng and Berthoud, unpublished observations).

Figure 1.

Schematic diagram depicting possible interactions between “cognitive” and “emotional” brain with the “metabolic” brain. Behavioral and anatomical observations suggest that accumbens→hypothalamus projections are involved in reward-driven food intake by modulating hypothalamic feeding circuits thought to be crucial for homeostatic control of energy balance. Indirect projections through the ventral pallidum and projections from the amygdala and relevant cortical areas may also be important. Hypothalamus→accumbens projections are also present and may provide modulation of feeding motivation by metabolic state signals.

These findings suggest a role for accumbens–hypothalamic pathways in the interaction between non-homeostatic and homeostatic factors that control food intake. Identification of the neural pathways that mediate this dominance of cortico-limbic processes over the homeostatic regulatory circuits in the hypothalamus and brainstem will be important for the development of behavioral strategies and pharmacological therapies in the fight against obesity.


This study was supported by National Institute of Health Grants DK47348 and DK071082.


  • 1

    Nonstandard abbreviation: DAMGO, (d-Ala2-N-methyl-Phe-Gly-ol5)-enkephalin.