Eating Disorders in Schizophrenia

Eating disorders in patients with schizophrenia have been underappreciated and poorly studied. Profiling characteristic phenotypic patterns will help clarify the distinctions among eating behaviors that are part of the spectrum of schizophrenia, those that represent distinct coexistent entities, and those that represent overlapping comorbidity.

Daily struggles with management strategies may lead physicians to raise certain questions about the prediction of illness courses. Among these questions are "Which symptoms are transient?" "Which are lasting?" and "Which should be prioritized in patient care?" There are tacit operational rules within medical practice that dictate which symptoms to focus on from a hierarchic list.

In this context, eating disorders in patients with schizophrenia have been underappreciated and poorly studied within the traditional diagnostic systems in mental illness. Even when they are recognized, there is often unsubstantiated optimism that eating-related symptoms will disappear as psychosis abates. On the other hand, the psychiatrist may have a speculative skepticism, regarding schizophrenia as encompassing an inability to comply with prescribed ways of living. This may reflect emphasis on overt behaviors in the clinical management of schizophrenia, with less concern about the cognitive drives behind them.

We have observed that patients with schizophrenia have disturbances in eating behaviors that make up an entire spectrum of eating disorders, including disinhibited as well as restrictive eating associated with underlying preoccupation with body image.1 Eating-related cognitive dimensions did not correlate with objective measures of severity of overt psychotic symptoms measured by the Positive and Negative Symptom Scale (PANSS),2 but instead, correlated strongly with patients' subjective perception of distress associated with psychiatric symptoms reported on the McLean Hospital 32-item Behavior and Symptom Identification Scale (BASIS-32).3

Determining whether eating disturbance is a comorbid condition in a subgroup of patients with schizophrenia or part of the broader schizophrenic spectrum disorder has important therapeutic implications for long-term outcome. Through functional profiling of phenotypic manifestations, we may be able to identify subgroups of schizophrenia with potential biologic overlap with eating disorders. This should be followed by a search for genetic correlates of variability.

Physiologic versus cognitive control of food

The coexistence of eating disturbance and schizophrenia is not a new phenomenon; in fact, it has been observed by many clinicians over the years. Kraepelin4 and Bleuler5 described disorganized and uncontrolled food intake as being characteristic of schizophrenia. Bruch,6 on the other hand, suggested that overeating in schizophrenia is an adaptive defense against stress, used for the maintenance of self-control; she referred to a monozygotic twin study in which Kallman observed greater weight as a favorable prognostic sign, since the heavier twin did not become psychotic. However, these early descriptions do not suggest cognitive awareness or deliberateness of behaviors.

Recent molecular examinations have assumed biologic differences in hunger, satiation, and satiety mechanisms in schizophrenia and have focused on metabolic signals. The assumption in most is that the homeostatic signals are offset by schizophrenia or by psychotropic medications. But this assumption is challenged by the heterogeneous patterns of weight and energy consumption changes observed in patients with identical diagnoses and medications.

The hypothalamic sum of physiologic feeding signals can easily be overridden by cortical signals related to emotions and environments. The salience of the motivation to eat; whether this actually results in enough drive to lead to active seeking of food; factors involved in eating termination; the actual hedonic value of the food; and how that changes future expectations of food reward and determines future feeding behaviors are all big chunks of the missing picture. Such cognitive processes may account in part for the individual variability in changes of food intake and body weight associated with the illness and its medications.


What kind of cognitive schemas drive eating and related behaviors in schizophrenia? Do patients have purposeful behaviors or is a compelling impulsivity accompanied by the presence of acknowledgment?

Distortions of body image and a deficient sense of self-effectiveness have been recognized as fundamental underlying features of schizophrenia.7 Such cognitive processes may predispose patients with schizophrenia to eating disorders. Many patients with schizophrenia will acknowledge body dissatisfaction and attempts at weight reduction, characterized by periods of restrictive eating, long-term unsuccessful diets that result in demoralization and cycles of binges and fasts. Some patients will deny body image preoccupations and concede the presence of less significant reasons for incessant calisthenics and rigid dietary restrictions, such as the need to fit into old clothes because there is no money for new ones.

Regarding uncontrolled intake, which has received more attention than food restriction in schizophrenia, some patients acknowledge eating large quantities of food after attempted periods of purposeful dieting, but without a subjective sense of loss of control over food intake. This is not a binge in its strict sense under DSM operational terms. Lack of insight is often regarded as a sign of psychosis. However, awareness or insight is not consistently present (nor are they all—or—nothing concepts) in other psychiatric illnesses, including eating disorders. In fact, one of the most distinctive characteristics of anorexia nervosa is a lack of appropriate concern about or denial of the seriousness of low body weight.8

Although it is a key drive for anorexic behaviors, the body conceptualization disturbance in anorexia nervosa is not well defined. There is considerable variation in the presence and severity of various observed behaviors and psychological states, such as denial of illness, motivation to change behavior, fear of changing behavior, and sense of adequacy and self-competence.9 Similarly, in bulimia nervosa, the International Classification of Diseases-10 criterion of "intrusive dread of fatness" and the DSM-IV criterion of "self-evaluation unduly influenced by body shape and weight" are not reflective of good insight. Eating disorders resemble schizophrenia in the fixation on thinness, perceptual distortions of somatic stimuli, perseveration over hips and thighs, and the bizarreness of beliefs and rituals related to eating.

To add further to the complexity, psychiatric symptoms fluctuate in qualitative as well as quantitative terms. When, during fluctuations in insight, a certain threshold of awareness is absent, accompanied by ego-syntonicity of excessive ideations and stereotypic behaviors, eating disorders have reached psychotic proportions. Indeed, for years, anorexia nervosa was considered a forme fruste of schizophrenia. Even now, in settings in which physicians are less familiar with eating disorders, such a complex of symptoms is much more likely to be diagnosed as schizophrenia.

Dimensional overlap and diagnostic challenges

One patient we saw when she was in her early 30s had been treated for schizophrenia since her late teens. She began starving herself at age 6, and she has chronic existential conflicts, disabling perfectionism, and fluctuating auditory hallucinations that gradually develop from heightened sensory perceptions and that are reactive to emotions and stressors. She empathically embraces her diagnosis of schizophrenia and the accompanying sick role in an enmeshed family.

Another woman we met in her 40s had been treated for schizophrenia since adolescence. She has never experienced hallucinations but has had fluctuating delusions of a paranoid/grandiose nature. One feature that has not changed throughout the course of her illness is her fixation on wanting to weigh 101 lb. She had 2 psychiatric hospitalizations because of refusal to eat. Her weight has fluctuated some 70 lb during adulthood, and she has now accepted her current weight of 130 lb at a height of 5 ft 7 in. However, she wakes up some mornings and sees her reflection as being extremely fat and wonders whether her "eyes are playing tricks on [her] mind." Despite her verbalized approval of her current weight, she hoards pictures of stick figures, continues to count calories, and dreams and fantasizes about life at 101 lb.

Both women appear to have their illnesses correctly diagnosed under operationally defined DSM criteria for schizophrenia. But should their management include primarily antipsychotic medications (Figure)? Other examples of dimensional overlap and diagnostic questions are seen in firstperson accounts of schizophrenia10 and descriptive cases of anorexia nervosa.11

The degree of body image distortion in some anorexic patients may appear to be of delusional proportion, but the disturbance is circumscribed to areas central to the primary eating disorder.12Although there is a high prevalence of multiple comorbid conditions in patients with eating disorders,13 it is fascinating that such phenomena are restricted to themes of body image and eating. For example, patients with eating disorders showed normal responses on the traditional Stroop tasks14 but had selective processing of information related to eating and weight on a modified version of the Stroop task.15

The distinction between eating disorders and schizophrenia is unclear because the brain itself does not naturally make such artificial diagnostic divisions. As Maudsley16 put it in 1879:

[I]nsanities are not really so different from sanities that they need a new and special language to describe them; nor are they so separated from other nervous disorders by lines of demarcation as to render it wise to distinguish every feature of them by a special technical nomenclature.

Metabolic implications

Whether eating disorders in patients with schizophrenia represent comorbid eating disorders or merely coexisting eating disturbances, the pathology influences metabolic health. Disinhibited eating was directly associated with metabolic changes such as increased triglyceride levels and body mass index.3 Cognitively restricted eating appears to be associated with disinhibited eating under various conditions.17 Within individual variability, we have also found and reported collective patterns of associations between medications and weight trends that are similar to those reported in the existing literature.18 However, such findings cannot adequately address how individual patient care can be optimized because medications appear to be only an additive factor to preexisting constitutional risks and multiple sociocultural stressors.19

Acceptance and self-affirmation

Foods serve diverse purposes emotionally and cognitively. Addressing the maladaptive and simultaneously adaptive defenses of disturbed eating is challenging. Low-salt, low-fat diets; weight loss; and exercise can be preached, but there is a danger that the distinction between the illness and the patient can become blurred as the fight against obesity, diabetes, hypertension, and dyslipidemia takes on emotional tones such as disparagement. In a recent case of a physician who refused educational remediation following patient complaints, the physician saw nothing wrong with telling the truth: "I told a fat woman she was obese. I tried to get her attention. I told her you need to get on a program, join a group of like-minded people and peel off the weight that is going to kill you."20

Healing words are valuable remedies. Conversely, inconsiderate words can be as wounding as any physical insult. Moreover, people who use food to combat anxiety and loneliness are apt to become depressed when dieting is enforced,21 and human relationships with food are difficult to alter with obesity lectures. Attempts to address any illness in patients with schizophrenia, or in any population for that matter, must consider each patient's emotional assets and frailties. Without addressing the underlying psychosocial conflicts that manifest in disordered behaviors, lasting, healthful changes are difficult to achieve.

In our experience with group therapy in schizophrenia with a primary emphasis on self-acceptance and subsidiary themes of health and eating behaviors, we have noticed changes in behaviors as well as improvements in blood pressures after about 6 months without measurable changes in cognitive schemas or weights.22 During this extended phase of contemplation and experimentation with alternatives, we have observed that praising any and all attempts as "good" is more effective in bringing about change than debating cognitive distortions and confronting resistance. Perhaps with such affirmations, focusing contemplation on self-care and healthy behaviors becomes therapeutic in itself, even in the absence of obvious changes in body weight or shape.


The heterogeneity of schizophrenia requires specific and individualized assessment and treatment approaches. With varied phenomenologic and clinical courses, identification and subsequent management of specific comorbid psychiatric syndromes are major determinants of outcome.23 Eating behaviors are complex reactive systems with multiple dimensions of phenomenologic variations, associated with various psychoneuroendocrine factors and dynamic in their progression. Further profiling of characteristic phenotypic patterns will help clarify the distinctions among eating behaviors that are part of the spectrum of schizophrenia, those that represent distinct coexistent entities, and those that represent overlapping comorbidity. These overlapping symptoms may then be studied for genetic causes. In addition, we should search for associations between behavioral phenotypes and genetic variability in psychotropic medication responses, including the effects on obesity. Such a need has recently assumed greater clinical importance with the increased prevalence of metabolic illnesses in schizophrenia that reflect both trait and state eating behaviors.

Dr Yum is affiliated with the New Jersey Veterans Affairs Health Care System, East Orange, NJ, and the Department of Psychiatry, University of Medicine and Dentistry of New Jersey-New Jersey Medical School in Newark. Dr Yum reports that he had support from the American Medical Association Foundation's "Fund for Better Health" public health grant for the group therapy discussed near the end of this article.

Dr Hwang is director of the Schizophrenia Program at the New Jersey Veterans Affairs Health Care System and associate professor of psychiatry at the University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School in Piscataway. Dr Hwang reports that he received project support from the American Medical Association.

Dr Halmi is professor of psychiatry and director of the Eating Disorder Program, Weill Medical College, Cornell University, New York. She reports that she has no conflicts to report regarding the subject matter of this article.



1. Yum SY. The starved brain: eating behaviors in schizophrenia. Psychiatr Ann. 2005;35:82-89.
2. Yum SY, Hwang MY. Eating disturbance in schizophrenia is not another symptom of psychotic phenomena and leads to metabolic morbidities. Paper presented at: Second Annual Research Symposium of the American Medical Association–Resident & Fellow Section; November 4, 2005; Dallas.
3. Yum SY, Hwang MY. Eating behaviors and associated psychopathology in patients with schizophrenia. In: Program and abstracts of the Annual Meeting of the American Psychiatric Association; May 23, 2005; Atlanta. Abstract NR54.
4. Kraepelin E. Dementia Praecox and Paraphrenia. Edinburgh: Thoemmes Press; 2002:87 [original published 1919].
5. Bleuler E. Textbook of Psychiatry. New York: Macmillan Co; 1924:149 [original published 1911].
6. Bruch H. Eating disorders and schizophrenia. In: Usdin G, ed. Psychoneurosis and Schizophrenia. Philadelphia: Lippincott; 1966:113-124.
7. Weckowicz TE, Sommer R. Body image and selfconcept in schizophrenia. J Ment Sci. 1960;106:17-39.
8. Casper RC, Heller W. "La douce indifference" and mood in anorexia nervosa: neuropsychological correlates. Progr Neuropsychopharmacol Biol Psychiatry. 1990;15:15-23.
9. Halmi KA. Classification, diagnosis and comorbidities of eating disorders: a review. In: Maj M, Halmi KA, Lopez-Ibor JJ, Sartorius N, eds. Eating Disorders. Chichester, England: Wiley; 2003:1-33.
10. Wagner PS, Spiro CS. Divided Minds: Twin Sisters and Their Journey Through Schizophrenia. New York: St Martin's Press; 2005.
11. Bruch H. The Golden Cage: The Enigma of Anorexia Nervosa. Cambridge, Mass: Harvard University Press; 2001 [original published 1978].
12. Halmi KA, Romano SJ. Anorexia nervosa: an overview. Prim Psychiatry. 2001;8:35-56.
13. Halmi KA, Eckert E, Marchi P, et al. Comorbidity of psychiatric diagnoses in anorexia nervosa. Arch Gen Psychiatry. 1991;48:712-718.
14. Ben-Tovim DI, Walker MK, Fox D, Yap E. An adaptation of the Stroop test for measuring shape and food concerns in eating disorders: a quantitative measure of psychopathology? Int J Eating Disord. 1989;8:681-688.
15. Cooper M, Todd G. Selective processing of three types of stimuli in eating disorders. Br J Clin Psychol. 1997;36 (pt 2):279-281
16. Maudsley H. Preface. In: Pathology of Mind. London: Macmillan; 1879.
17. Yum SY, Hwang MY, Yum SK. Predictors of disinhibited eating in restrained eaters in patients with schizophrenia. Obesity Res. 2005;13(Sept suppl):A71.
18. Hwang MY, Yum SY, Hamayun A. Antipsychotic treatment of obesity and metabolism in schizophrenia. In: Program and abstracts of the Annual Meeting of the American Psychiatric Association; May 23, 2005; Atlanta. Abstract NR183.
19. Yum SY, Hwang MY, Yum SK. A biopsychosocial look at metabolic illness in schizophrenia. World J Biol Psychiatry. 2005;6(S1):164.
20. Rayno G. Complaint over blunt words on obesity frustrates doctor. Union Leader. Manchester, NH. August 22, 2005. Available at: html. Accessed May 15, 2006.
21. Stunkard AJ. The "dieting depression." Incidence and clinical characteristics of untoward responses to weight reduction regimens. Am J Med. 1957;23:77-86.
22. Yum SY, Hwang MY. Group cognitive therapy improved blood pressures in schizophrenia. To be presented before the 61st Annual Scientific Convention and Meeting of the Society of Biological Psychiatry; May 18-20, 2006; Toronto.
23. Hwang MY, Bermanzohn PC. Schizophrenia and Comorbid Conditions: Diagnosis and Treatment. Washington, DC: American Psychiatric Press; 2001.