Review
Behavioral neuroendocrinology and treatment of anorexia nervosa

https://doi.org/10.1016/j.yfrne.2008.06.001Get rights and content

Abstract

Outcome in anorexia nervosa remains poor and a new way of looking at this condition is therefore needed. To this aim, we review the effects of food restriction and starvation in humans. It is suggested that body weight remains stable and relatively low when the access to food requires a considerable amount of physical activity. In this condition, the human homeostatic phenotype, body fat content is also low and as a consequence, the synthesis and release of brain neurotransmitters are modified. As an example, the role of neuropeptide Y is analyzed in rat models of this state. It is suggested that the normal behavioral role of neuropeptide Y is to facilitate the search for food and switch attention from sexual stimuli to food. Descriptive neuroendocrine studies on patients with anorexia nervosa have not contributed to the management of the patients and the few studies in which hormones have been administered have, at best, reversed an endocrine consequence secondary to starvation. In a modified framework for understanding the etiology and treatment of anorexia nervosa it is suggested that the condition emerges because neural mechanisms of reward and attention are engaged. The neural neuropeptide Y receptor system may be involved in the maintenance of the behavior of eating disorder patients because the localization of these receptors overlaps with the neural systems engaged in cue-conditioned eating in limbic and cortical areas. The eating behavior of patients with anorexia nervosa, and other eating disorders as well, is viewed as a cause of the psychological changes of the patients. Patients are trained to re-learn normal eating habits using external support and as they do, their symptoms, including the psychological symptoms, dissolve.

Introduction

The aim of studies on patients with anorexia nervosa is to improve their condition, but there are as yet no neuroendocrine studies and few, if any other studies that have contributed the reaching this goal [21]. A new way to approach the problem is needed.

It was suggested long ago that the hormonal changes in anorexia nervosa are non-specific effects of starvation [62] and so far, endocrine interventions have had no effect on the behavioral condition of the patients. At the present state of knowledge, therefore, the cause–effect relationship between hormones and behavior in anorexia nervosa is unclear and needs to be reconsidered. We will expand on this topic in light of recent results suggesting that some so-called “orexigenic” peptides influence appetitive, rather than consummatory aspects of ingestive behavior [4], [49], [99]. This is a departure from the prevailing paradigm, according to which hormones from stored fat and pancreas, along with glucose, lipids and amino acids from ingested macronutrients feed back on brain networks to excite or inhibit eating such that body weight remains stable over time [15]. There are many excellent recent reviews on this topic, e.g. [39], [101], on the neuroanatomical engagement during eating, e.g. [58], [72] and of the endocrine changes, e.g. [114] and the medical aspects of anorexia nervosa, e.g. [110]. Reviewing these fields again would be redundant.

According to the conventional view, anorexia nervosa emerges as a symptom of an unknown underlying mental disorder and we have recently discussed the problems associated with this view in detail [149], [150], [151], [152], [176]. No effective treatment has been developed based on this point of view [160]. The little support that may exist for the conventional view is in stark contrast to the overwhelming evidence that starvation causes psychological change and even mental disorders. Because the conventional view still permeates thinking about the cause and treatment of anorexia, we will review the evidence that starvation causes mental problems in some detail.

We will discuss the effects of enforced starvation and subsequent re-feeding first and then suggest that the symptoms of anorexia nervosa are reversible, often physiological consequences of starvation. We will then review some aspects of the life of Australian Aboriginal hunter-gatherers, which is conspicuously similar to that of starved or semi-starved people and patients with anorexia nervosa. On the basis of these descriptions, we will re-evaluate the concept of body weight homeostasis and suggest that maintenance of a low body weight for a long period of time along with a high level of physical activity and retention of the capacity to eat large meals provide an example of a situation in which body weight is homeostatically regulated. We will then describe the way in which hormones and behavior interact in this situation; the human homeostatic phenotype. The description is admittedly selective and aims at proposing a new paradigm that does not conform to the conventional concept of homeostatic regulation. This step seems necessary, because the conventional view is incompatible with the neuroendocrinology of anorexia nervosa.

It is commonly thought that there are many risk factors and that anorexia nervosa has an unknown etiology, but this hypothesis has not resulted in effective interventions. We have considered some other problems associated with this view elsewhere [149], [150], [151], [152] and will therefore not re-consider these in the present context. Instead, we have suggested that anorexia nervosa develops from the two known risk factors: reduced food intake and enhanced physical activity [20]. We have also suggested that a reduction in food intake would cause an increase in the release of dopamine in the ventral striatal terminals of the mesolimbic dopamine neurons via both adrenocortical hormone feedback on these terminals and activation of the dopaminergic cell bodies in the ventral tegmentum of the mesencephalon [20], [95]. Release of corticotrophin-releasing hormone (CRH) in the hypothalamus is the primary mediator of both these mechanisms [20]. The mesolimbic dopamine system is associated with the short-latency experience of reward [105]. In addition, release of CRH in the brain activates the noradrenergic neurons of the locus coeruleus in the brainstem [87]. These neurons project to forebrain areas that are engaged whenever an individual needs to pay attention to ongoing events [130], [135]. Hence, we suggested that anorexia nervosa develops because it is initially rewarding to eat less food and that anorexic behavior is subsequently maintained through conditioning to the situations that provide the reward [20]. Methods to treat patients based on this point of view were set up and the treatment was evaluated in a randomized controlled trial [17]. In this trial, 16 patients were allocated to treatment and 16 other patients were allocated to no-treatment using a randomized procedure. The number of patients needed in each group in order to demonstrate a significant effect was estimated based on preliminary results [18] and the possibility that patients may drop out was also taken into consideration. Very few exclusion criteria were used. Treatment was found to have a major effect on a series of strict outcome criteria; patients had to show normal eating patterns and perceive a normal level of satiety, have a normal body mass index (BMI), normal laboratory test values and normal levels of psychiatric symptoms, they had to be able to say that food and body weight were no longer a problem and they had to be back at school or professional activities [17]. All these criteria had to be fulfilled before a patient was considered in remission. Thus, a treatment based on a neurobiologically plausible framework was demonstrated to be effective. Furthermore, the treatment brought an estimated 75% of a group of 168 patients into remission and only 10% of a group of 83 patients relapsed within a year [17].

Finally, we will refine the framework for understanding anorexia nervosa delineated above and suggest how patients with anorexia nervosa and other eating disorders should be treated, based on this view.

Section snippets

The response to enforced starvation

During the last months of the Second World War, the supply of food to the Western Netherlands was severely restricted. The consequences of this enforced starvation have been extensively studied and the description below relies on the evidence presented in the first report of this phenomenon [29], unless otherwise noted.

Before the enforced starvation, the people of the Netherlands had not suffered from lack of food as their energy intake had been about 2000 kcal/day during most of the war.

The response to re-feeding

By the time the Germans had capitulated on May 7, 1945, 141,636,186 kg of food had been stored in southern Holland and the plan was to provide 2000 kcal/day to a population of 4.2 million. It was estimated that 400,000 people were in need of special treatment for starvation and that 30,000 needed acute treatment in a hospital. About 279,000 were actually treated and 3000 of these were treated in hospitals. Two hundred seventy-five (9%) of these died in hospital, most of them during the first

Maintenance of a BMI of 17 for 40,000 years

Australian Aborigines lived in small communities and maintained a relatively constant life style for perhaps more than 40,000 years [24]. It has been estimated that the size of this population of hunter-gatherers remained stable at about 300 000, with many mechanisms to maintain stability, e.g., prolonged breast feeding along with the associated lactational amenorrhea and infanticide of as many as 50% of the newborn children. Inter-tribal warfare and perhaps infections also contributed to

The human homeostatic phenotype

Body weight is most often considered to be the outcome of the activity of neuroendocrine networks that “control” eating behavior in order to keep body weight constant. This paradigm originates with Claude Bernard’s idea that the internal milieu of an organism must be kept stable, and if it is not, the organism will be entirely dependent upon fluctuations in the external environment [22].

Descriptive versus predictive neuroendocrinology

Most studies on the relationship among the neuroendocrine system, eating behavior and body weight rely on the Bernard–Cannon concept of homeostasis as adapted to neural function [158]. As mentioned above, the model suggests that brain networks excite or inhibit eating behavior in response to feedback from endocrine and metabolic signals in order to keep body weight constant, e.g. [39], [101]. Behavior is thought of as an outcome of neural control in this model; many publications are entitled:

The neuroendocrinology of the human homeostatic phenotype: the neuropeptide Y-paradox

As described above, the human homeostatic phenotype is one with a constant low BMI, a low body fat content, a high level of physical activity and retention of the capacity to eat more than required for maintenance of the low BMI. The neuroendocrinology of this phenotype is determined by the content of body fat. The lower the fat content the lower the level of leptin in the blood [81]. Leptin feeds back on the brain and causes a decrease in the synthesis and release of neuropeptide Y (NPY) and

Resolution of the NPY-paradox, part 1: NPY stimulates appetitive ingestive behavior

When an animal eats less food and loses weight, the synthesis and release of NPY increase in the brain [12]. The immediate behavioral response in this situation is hoarding, not eating [31], [133], [141]. The first indication that NPY might have a role in appetitive ingestive behavior was provided by an observation that intracerebral infusion of NPY does not stimulate consummatory ingestive behavior tested by intraoral infusion as described above [145].

Resolution of the NPY-paradox, part 2: NPY facilitates activity-based-anorexia

A high level of physical activity is an important behavioral expression of the human homeostatic phenotype. The goal of the activity is the acquisition of food. In the present analysis, it follows that the activity equals appetitive ingestive behavior. It also follows that treatment with NPY should stimulate physical activity in a model of anorexia nervosa. Such a model is available for 46 years [153]. In that model, rats have access to running wheels and, if allowed continuous access to food,

The capacity to eat a large amount of food

Research on meal initiation and termination has typically studied situation “when food is freely available” [161] or after brief periods of deprivation and as pointed out above these models may not be applicable to the human homeostatic phenotype. An important aspect of that phenotype is the capacity to eat large amounts of food when the BMI is low. At the present state of knowledge, there is therefore limited information on the factors that allow an individual to eat much more food than

A modified framework for anorexia nervosa

Twelve years ago, in a preliminary framework for understanding the etiology and treatment of anorexia nervosa, we suggested that anorexic behavior is maintained by conditioning to the situations that provided the reward that was initially experienced during the reduction of food intake. The overlapping neural engagement in food intake and reward has since then been studied in detail [171]. We also suggested that the noradrenergic projections from the locus coeruleus to the cortex are engaged in

Treatment of anorexia nervosa

“How can we change the eating habits of today’s children? I think that these questions require more thought than worrying about people trying to clone themselves.” Sydney Brenner[26].

Our answer to Brenner’s question is: through external support. In this review, we have argued that a situation in which humans maintain a constant body weight is when there is a demand for a high level of physical activity to obtain food. Anorexic patients are captured in this situation and there is no compelling

Conclusions

This review has attempted to switch attention from the conventional model of homeostatic regulation of body weight, which is based on tests of animals with continuous effortless access to food, to one in which the acquisition of food demands an effort. The conventional model does not mimic the normal human condition. Instead, we have argued that the human homeostatic phenotype is one in which body weight is controlled mainly by external factors; the most important is the availability of food.

Acknowledgments

We thank the SDU, the Netherlands, for permission to reproduce the data in Fig. 1, the National Health Service, The Information Centre, and the Department of the Environment, Food and Rural Affairs and Dr. F. Floud for permission to reproduce the data in Fig. 4, Elsevier for permission to reproduce Fig. 5, Fig. 11, Fig. 12, Fig. 13, Fig. 15, Blackwell Publishing for permission to reproduce the data in Fig. 6 and the American Physiological Society for permission to reproduce Fig. 7, Fig. 8, Fig.

References (176)

  • J.H. Fallon et al.

    Catecholamine innervation of the basal forebrain. II. Amygdala, suprarhinal cortex and entorhinal cortex

    J. Comp. Neurol.

    (1978)
  • M.L. Glucksman et al.

    The response of obese patients to weight reduction: a clinical evaluation of behavior

    Psychosom. Med.

    (1968)
  • H.J. Grill et al.

    Evaluating taste effects on ingestive behavior

  • S. Grinspoon et al.

    Effects of estrogen and recombinant human insulin-like growth factor-I on ghrelin secretion in severe undernutrition

    J. Clin. Endocrinol. Metab.

    (2004)
  • J. Hebebrand et al.

    The role of leptin in anorexia nervosa: clinical implications

    Mol. Psychiatry

    (2007)
  • A.W. Illius et al.

    The evolution of the control of food intake

    Proc. Nutr. Soc.

    (2002)
  • H.P. Jedema et al.

    Corticotropin-releasing hormone directly activates noradrenergic neurons of the locus coeruleus recorded in vitro

    J. Neurosci.

    (2004)
  • S. Kanayama et al.

    Influence of food deprivation on intestinal cholecystokinin and somatostatin

    Gastroenterology

    (1991)
  • J.M. Kaplan et al.

    Simultaneous display of sexual and ingestive behaviour by rats

    J. Neuroendocrinol.

    (1992)
  • D.A. Klein et al.

    Physical activity and cortisol in anorexia nervosa

    Psychoneuroendocrinology

    (2007)
  • E. Koya et al.

    Molecular reactivity of mesocorticolimbic brain areas of high and low grooming rats after elevated plus maze exposure

    Brain Res. Mol. Brain Res.

    (2005)
  • K.S. LaBar et al.

    Hunger selectively modulates corticolimbic activation to food stimuli in humans

    Behav. Neurosci.

    (2001)
  • Lecklin et al.

    Receptor subtypes Y1 and Y5 mediate neuropeptide Y induced feeding in the guinea-pig

    Br. J. Pharmacol.

    (2002)
  • American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, fourth ed., Washington, DC,...
  • A.A. Ammar et al.

    NPY-leptin: opposing effects on appetitive and consummatory ingestive behavior and sexual behavior

    Am. J. Physiol. Regul. Integr. Comp. Physiol.

    (2000)
  • A.A. Ammar et al.

    Locus coeruleus noradrenergic lesions attenuate intraoral intake

    Neuroreport

    (2001)
  • F.S. Arana et al.

    Dissociable contributions of the human amygdala and orbitofrontal cortex to incentive motivation and goal selection

    J. Neurosci.

    (2000)
  • S.A. Bacanu et al.

    Linkage analysis of anorexia and bulimia nervosa cohorts using selected behavioral phenotypes as quantitative traits or covariates

    Am. J. Med. Genet. B Neuropsychiatr. Genet.

    (2005)
  • M. Bachet, Etudes des troubles causés par la dénutrition dans un asile d’aliénés, Arnette, Paris,...
  • J.P. Baird et al.

    Effects of neuropeptide Y on feeding microstructure: dissociation of appetitive and consummatory actions

    Behav. Neurosci.

    (2006)
  • J.P. Baird et al.

    Effects of hindbrain melanin-concentrating hormone and neuropeptide Y administration on licking for water, saccharin, and sucrose solutions

    Am. J. Physiol. Regul. Integr. Comp. Physiol.

    (2008)
  • R.L. Batterham et al.

    Gut hormone PYY(3-36) physiologically inhibits food intake

    Nature

    (2002)
  • B. Beck

    Neuropeptide Y in normal eating and in genetic and dietary-induced obesity

    Philos. Trans. R. Soc. Lond. B Biol. Sci.

    (2006)
  • A.P. Beckerman et al.

    Foraging biology predicts food web complexity

    Proc. Natl. Acad. Sci. USA

    (2006)
    F.G. Benedict, W.R. Miles, P. Roth, H.M. Smith, Human Vitality and Efficiency Under Prolonged Restricted Diet, ,...
  • F.G. Benedict et al.

    Human Vitality and Efficiency Under Prolonged Restricted Diet

    (1919)
  • S.C. Benoit et al.

    Insulin and leptin as adiposity signals

    Recent Prog. Horm. Res.

    (2004)
  • S.C. Benoit et al.

    The role of previous exposure in the appetitive and consummatory effects of orexigenic neuropeptides

    Peptides

    (2005)
  • C. Bergh et al.

    A new treatment of anorexia nervosa

    Lancet

    (1996)
  • C. Bergh et al.

    Anorexia nervosa, self-starvation and the reward of stress

    Nat. Med.

    (1996)
  • N.D. Berkman et al.

    Outcomes of eating disorders: a systematic review of the literature

    Int. J. Eat. Disord.

    (2007)
  • C. Bernard

    An Introduction to the Study of Experimental Medicine 1865

    (1957)
  • C.L. Birmingham et al.

    Randomized controlled trial of warming in anorexia nervosa

    Int. J. Eat. Disord.

    (2004)
  • J. Blainey

    Triumph of the nomads, A History of Aboriginal Australia

    (1993)
  • C.V. Bouten et al.

    Body mass index and daily physical activity in anorexia nervosa

    Med. Sci. Sports Exerc.

    (1996)
  • S. Brenner

    Humanity as the model system

    Science

    (2003)
  • H. Bruch

    Perceptual and conceptual disturbances in anorexia nervosa

    Psychosom. Med.

    (1962)
  • C.M. Bulik et al.

    Selection of eating-disorder phenotypes for linkage analysis

    Am. J. Med. Genet. B Neuropsychiatr. Genet.

    (2005)
  • G.C.E. Burger et al.

    Starvation and Malnutrition in Western Netherlands, The Hague General State Printing Office

    (1948)
  • E.J. Button et al.

    A reappraisal of body perception disturbance in anorexia nervosa

    Psychol. Med.

    (1977)
  • M. Cabanac et al.

    Rats eating and hoarding as a function of body weight and cost of foraging

    Am. J. Physiol.

    (1989)
  • Cited by (0)

    View full text