The Adaptive Response of the Immune System to the Particular Malnutrition of Eating Disorders

Home , Anorexia (symptom)

European Journal of Clinical Nutrition (2002) 56, Suppl 3, S34–S37 ß 2002 Nature Publishing Group All rights reserved 0954–3007/02 $25.00 www.nature.com/ejcn ORIGINAL COMMUNICATION The adaptive response of the immune system to the particular malnutrition of eating disorders

E Nova1, S Samartı´n1,SGo´mez1, G Morande´2 and A Marcos1*

1Instituto de Nutricioń y Bromatologıá (CSIC), Edificio Instituto del Frıó, Madrid, Spain; and 2Servicio de Psiquiatrıá Infantil, Hospital del Ninõ Jesu´s, Madrid, Spain

Despite the seriously undernourished state of patients with anorexia nervosa (AN) and bulimia nervosa (BN), controversial findings have been published regarding some aspects of the immune system that are otherwise impaired in more typical types of malnutrition, such as protein-energy malnutrition. In general, adaptation processes seem to occur enabling immune function to be preserved during long periods of the illness. However, cell-mediated immunity is usually altered in AN and BN as reflected by lymphocyte subset counts and the response to delayed hypersensitivity tests. Regarding the helper=cytotoxic T cell ratio (CD4:CD8), an immunological marker of the nutritional status, the results of our studies on AN and BN patients showed that the duration of the eating disorder and the time when appropriate treatment is achieved are likely contributors to the alteration of this ratio. Despite these findings, it has been repeatedly pointed out that anorexic patients seem to be free of common viral infections at least until the most advanced stages of debilitation. Some hypotheses that could explain the lack of infection symptoms are reviewed. Cytokines and the altered acute phase response to infection, as well as cortisol and leptin, are considered to be potential factors involved in the adaptation processes occurring in these syndromes. Further progress in the knowledge of the psychoneuroendocrine – immune interactions established in AN and BN will be relevant to the understanding of the aetiology and maintenance mechanisms of these pathologies. European Journal of Clinical Nutrition (2002) 56, Suppl 3, S34 – S37. doi:10.1038=sj.ejcn.1601482

Keywords: eating disorders; immune system; nutritional status; infection; cytokines

Features of eating disorders behaviour and attitude about food, sometimes accompanied Anorexia nervosa by self-induced vomiting and binge eating. These compen- Anorexia (AN) usually starts in the mid-teens and affects one satory mechanisms are present in the binge=purging subtype 15- y-old girl in every 150. Unfortunately, a very early start, of AN, as compared with the restricting subtype (APA, 1994). in childhood, is becoming more frequent, and occasionally it may start later, in the 30s or 40s. Nearly always, anorexia begins with the everyday dieting that is so much a part of Bulimia nervosa teenage life. Although technically the word anorexia means In bulimia nervosa (BN) the distinguishing feature is binge ‘loss of appetite’, sufferers from AN actually have a normal eating, which is the rapid consumption of a large quantity of appetite, but drastically control their eating. The pathophy- food in a short period of time, usually less than 2 h. Binge siological characteristics of AN patients, and those that are eating is commonly followed by purging. In the purging type currently used in their diagnosis are briefly the following: (1) of BN, vomiting and the abuse of laxatives and diuretics are a disturbed perception of body size and body image; (2) self- the compensatory mechanisms to avoid putting on weight starvation with significant weight loss; (3) amenorrhoea; (4) after bingeing. In the non-purging type of BN these mechan- physical hyperactivity and sleep disturbances; (5) bizarre isms include periods of diet restriction and physical exercise. Taken together, the binge eating and compensatory behaviour occur at least twice a week for 3 months (APA, 1994). It is very well known that the most common feature in AN *Correspondence: A Marcos, Instituto de Nutricioń y Bromatologıá and BN is an obsessive desire to lose weight together with an (CSIC), Edificio Instituto del Frıó, C=Ramiro de Maeztu, s=n, 28040 Madrid, Spain. exaggerated fear of becoming fat, leading to a particular type E-mail: [email protected] of malnutrition in both syndromes. Long-term follow-up Eating disorders and immunocompetence E Nova et al S35 studies indicate that AN is associated with serious morbidity anorexic patients. It has also been suggested the presence and mortality, although the prognosis for severe and refrac- of a stimulation factor in the serum of AN patients that tory BN is generally less favourable than for restrictive AN favours the maintenance of a normal lymphocyte transfor- (Marcos, 1997). mation response (Bessler et al, 1993).

The immune system in eating disorders: conflicting Cell-mediated immunity results When analysing immune parameters it is necessary to bear It is well established that malnutrition and infection have in mind the complex interactions and reciprocal control mutually aggravating effects. Nutritional deprivation, such among the immune system, the endocrine system and the as protein-energy malnutrition, often causes immunodefi- central nervous system (Marcos, 2000). Malnutrition may ciency, leading to increased frequency and severity of infec- have an impact on these interactions and may impair the tion, thymus atrophy and wasting of peripheral lymphoid communication between these systems. The neurochemical tissue (Chandra & Kumari, 1994). This outcome is reflected disorders in eating disorders may perpetuate pathologic in the significant impairment of several aspects of immunity, eating behaviours and might be responsible for several asso- including cell-mediated immune responses, production of ciated psychiatric symptoms including stress, anxiety and secretory immunoglobuline A, phagocyte function, comple- depression (Brambilla, 2001). ment system and cytokine production. However, regarding Our research group has carried out a few studies on the eating disorders, the studies on the impact of this particular immune status of patients suffering from AN or BN. From kind of malnutrition over the immune system have pro- our studies, it is clear that this type of malnutrition affects duced controversial findings. Thus, even though some of the cellular immunity first rather than humoral immunity. In immune impairments of AN are similar to those observed in fact, leukocyte, lymphocyte and T-cell counts have been simple malnutrition, they are less frequent and less severe, shown to be depleted in restricting type AN patients (mean and the immune function seems to be better preserved than age of 15, mean body mass index (BMI) ¼ 15.7 kg=m2) during would be expected, considering the highly defective nutri- a 1 y follow-up, while B cells are not modified in comparison tional status of the patients (Silber & Chan, 1996). with controls. However, an improvement of T cell levels has Another aspect of great interest is the fact that anorexic been found when patients are under treatment during hos- patients are surprisingly free of infectious complications pitalization, although these values are decreased back after despite their seriously undernourished state, at least until discharge. Nevertheless, cell-mediated immune function the late stages of debilitation. During long periods of their evaluated by the response to the delayed hypersensitivity illness, underweight AN patients seem to be protected skin test (both the number of positive responses and the against common viral infections (Mustafa et al, 1997). score or the sum of all diameters from the positive responses) is shown to be depleted in AN patients during the whole period of the study, especially during the hospitalization Why are patients with eating disorders less prone period (Figure 1, Marcos et al, 1997a). This outcome is an to infection than subjects under typical evidence of an increased risk of relapse during the first year malnutrition? after patients are discharged from hospital (Nova et al, 2001). Firstly, important differences exist between AN and starva- Similar results have been found in a group of 21 BN tion in terms of nutrient deficiencies. In starvation, the diet patients (mean age ¼ 19) and a 2 y duration of their disease is deficient in multiple vitamins and proteins, as well as upon the beginning of the study; patients BMI showing energy, but in AN the primary dietary inadequacy is of similar levels to the control group (Marcos et al, 1997b). It carbohydrates and fats. Protein intake is usually adequate in AN, and vitamin deficiencies are rare (Nova et al, 2001). Thus, a relatively high protein intake in AN patients, con- trary to more typical situations of nutritional deprivation, could contribute to the lack of infections (Marcos, 1997). However the protection seems to be lost on refeeding, as patients complain about recurrence of infection after gaining weight. It has been reported that starvation may suppress and refeeding may activate certain infections (Marcos, 2000). In this respect, it has been speculated that a marked reduc- tion in the percentage and absolute number of memory CD8 þ T cells found in AN patients could lead to a reduced frequency of lymphocytes capable of recall responses (Mus- tafa et al, 1997), and this would be related to a perceived lack Figure 1 CD4=CD8 and CD2=CD19 ratios in patients with AN in a 1 y of symptomatic common viral infections in underweight follow-up and in controls.

European Journal of Clinical Nutrition Eating disorders and immunocompetence E Nova et al S36 is important to stress that the subclinical malnutrition in spontaneous and elevated production of IL-6 and TNF-a, these patients is not evident only assessing their anthropo- contributing to weight loss, cachexia and osteoporosis in metrical parameters, but it is also necessary to evaluate their AN (Holden & Pakula, 1998). Cortisol secretion is also high immunocompetence. Vomiting practice also affect T cell in AN, which means that the feedback mechanism is not levels. working in this eating disorder. In relation with those spon- In both AN and BN patients, the CD4:CD8 ratio is very taneously elevated levels of pro-inﬂammatory cytokines, it likely to be affected by the duration of the eating disorder has been suggested that those cells stimulated to a certain and by the time when appropriate treatment is achieved. The extent to produce cytokines will respond poorly to further earlier the diagnosis and the more adequate the treatment stimulation (Figure 3). According to this, an impaired capa- are, the higher this ratio is (Marcos et al, 1993, 1997a,b). city to mount an acute-phase response to infection, under an These results can show the importance of an early diagnosis extra stimulus, could be expected. in order to avoid a severe situation of malnutrition. It is also Furthermore, a positive correlation between high cortisol important to stress that CD2:CD19 ratio is a good index to levels and an increased IL-1b secretion by cultured PBMC corroborate the hypothesis that cell-mediated immunity is from AN patients have been reported (Limone et al, 2000), usually more affected than humoral immunity (Marcos et al, which seems to mean that the normal relationship between 1997a), even after 1 y of treatment and in patients with an both parameters is deranged by the illness. On the other early diagnosis (Figure 2). hand, since IL-1b production by macrophages is regulated by a cortisol receptor in monocytes (Paez-Pereda et al, 1996), the high serum cortisol levels consistently found in AN patients Regulation of the immune system response to could result in an impaired IL-1b production in response to infection infectious processes. Then, a decreased release of this pyro- Role of cytokines gen could explain the inability of AN patients to get fever. Cytokines, as regulators of the immune system, play an Thus, the altered production of cytokines points out another important role in the relationship between nutrition and possible mechanism to explain the absence of infection infection. Cytokines are modulated by nutrients and their symptoms in patients with eating disorders (Figure 3). participation is essential in triggering certain mechanisms involved in infection processes. In cases of infection there is a cytokine-induced malnutrition, which results from the Role of leptin actions of proinﬂammatory cytokines such as TNF-a, IL-1 Leptin is a protein encoded by the ob gene, which is known and IL-6 (Figure 3). These cytokines are known to be able to to regulate appetite and energy expenditure, and thus to initiate an acute-phase reaction, which is stereotyped and control body weight. It is usually positively correlated with includes fever, loss of appetite, decreased food intake, cellular hypermetabolism and multiple endocrine and enzyme responses (Grimble, 1994). These cytokines are capable of activating the hypothalamic – pituitary – adrenal axis (HPAA) and to have a direct stimulatory effect on ACTH secretion. In turn, glucocorticoids alter the production of these cytokines in a feedback regulation. However, in the particular case of AN, patients do not show exactly this stereotyped reaction, especially that related to fever. Some authors have reported a

Figure 2 Number of positive responses and score values in response to Figure 3 Scheme of the abnormal functioning of cortisol and leptin on the delayed hypersensitivity skin test in patients with AN in a 1 y follow- proinﬂammatory cytokine synthesis and subsequent lack of infection up and in controls. symptoms in eating disorders.

European Journal of Clinical Nutrition Eating disorders and immunocompetence E Nova et al S37 BMI in obese, normal and slim individuals and also in AN Brambilla F (2001): Aetiopathogenesis and pathophysiology of bulimia nervosa: biological bases and implications for treatment. CNS and BN patients (Monteleone et al, 2000). In addition, Drugs 15, 119 – 136. exogenous leptin has been reported to stimulate phagocytic Chandra RK & Kumari S (1994): Effects of nutrition on the immune function and activate macrophages to produce proinflam- system. Nutrition 10, 207 – 210. matory cytokines, such as TNF-a, IL-6 and IL-12 in ob=ob Fantuzzi G & Faggoni R (2000): Leptin regulation of the immunity, inflammation and hematopoiesis. J. Leukoc. Biol. 68, 437 – 446. mice (Loffreda et al, 1998). Therefore, there is a novel func- Grimble RF (1994): Malnutrition and the immune response. 2. tion for leptin as an up-regulator factor of inflammatory Impact of nutrients on cytokine biology in infection. Trans. R. immune responses. Moreover, leptin production is acutely Soc. Trop. Med. Hyg. 88, 615 – 619. increased during infection and inflammation (Fantuzzi & Herpertz S, Norbert A, Richard W, Pelz B, Koepp W, Mann K, Blum WF, Senf W & Hebebrand J (2000): Longitudinal changes of Faggoni, 2000). Thus, an impairment in this acute increase circadian leptin, insulin and cortisol plasma levels and their in leptin production in AN patients could be related to the correlation during refeeding in patients with anorexia nervosa. lack of infection symptoms in these patients. Leptin con- Eur. J. Endocrinol 142, 373 – 379. centration is also decreased in BN patients, despite weight Holden RJ & Pakula IS (1999): Tumor necrosis factor-alpha: is there a continuum of liability between stress, anxiety states and anoexia values being similar to those in age-matched controls (Mon- nervosa? Med. Hypothesis 52, 155 – 162. teleone et al, 2000). Thus, factors other than body weight Limone P, Biglino A, Bottino F, Forno B, Calvelli P, Fassino S, Berardi may play a role in the determination of leptin changes in C, Ajmone-Catt P, Bertagna A, Tarocco RP, Rovera GG & Molinatti eating disorders (Figure 3). GM (2000): Evidence for a positive correlation between serum cortisol levels and IL-1beta production by peripheral mononuclear Prior to refeeding, the semistarvation state in AN patients cells in anorexia nervosa. J. Endocrinol. Invest. 23, 422 – 427. is associated with exceedingly low plasma leptin levels, a Loffreda S, Yang SQ, Lin HZ, Karp CL, Brengman ML, Wang DJ, Klein quantitative alteration in the circadian rhythm of leptin and AS, Bulkley GB, Bao C, Noble PW, Lane MD & Diehl AM (1998): cortisol levels and an alteration in the temporal coupling Leptin regulates proinflammatory immune responses, FASEB J. 12, 57 – 65. between cortisol and leptin. After weight gain, a leptin rise Marcos A, Varela P, Santacruz I, Munõz-Velez A & Morande´ G (1993): precedes the cortisol increase by 8 h, while no temporal Nutritional status and immunocompetence in eating disorders. A relationship is found between them in the semistarvation comparative study. Eur. J. Clin. Nutr. 47, 787 – 793. state (Herpertz et al, 2000). Marcos A (1997): The immune system in eating disorders: an over- view. Nutrition 13, 853 – 862. Marcos A, Varela P, Toro O, Lo´pez-Vidriero I, Nova E, Madruga D, Casas J & Morande´ G (1997a): Interactions between nutrition and Final remarks immunity in anorexia nervosa. A one year follow-up. Am. J. Clin. In conclusion, the outcomes so far suggest that the inter- Nutr. 66(Suppl), S485 – S490. Marcos A, Varela P, Toro O, Nova E, Lo´pez-Vidriero I & Morande´ G relationships among cortisol, leptin and cytokines, all of (1997b): Evaluation of nutritional status by immunological assess- them repeatedly found to be altered in eating disorders, ment in bulimia nervosa. Influence of BMI and vomiting episodes. may play an important role in the regulatory mechanism Am. J. Clin. Nutr. 66(Suppl), S491 – S497. triggered in individuals with AN or BN. These compensatory Marcos A (2000): Eating disorders: a situation of malnutrition with peculiar changes in the immune system. Eur. J. Clin. Nutr. 54 mechanisms may enable the patients to adapt to these (Suppl 1), S61 – S64. atypical situations of malnutrition and might also provide Monteleone P, Di Lieto A, Tortorella A, Longobardi N & Maj M a clue to explain some of the immunological findings, such (2000): Circulating leptin in patients with anorexia nervosa, as the lack of infections in these patients. Future findings bulimia nervosa or binge eating disorder: Relationship to body weight, eating patterns, psychopathology and endocrine changes. which increase our knowledge about the bi-directional com- Psychiat. Res. 94, 121 – 129. munication pathways between the neuroendocrine system Mustafa A, Ward A, Teasure J & Peakman M (1997): T lymphocyte and the immune system will help to perform further studies subpopulations in anorexia nervosa and refeeding. Clin. Immunol. in patients with eating disorders trying to define the role of Immunopathol. 82, 282 – 289. Nova E, Varela P, Toro O, Lo´pez-Vidriero I, Casas J, Cenãl MJ & those mediators in the pathogenesis of these syndromes. Marcos A (2001): Dietary intake and anthropometry in anorexia nervosa patients undergoing rehabilitation treatment. A one-year follow-up study. Eur. J. Clin. Nutr. 55, 547 – 554. References Paez Pereda M, Perez Castro C, Costas M, Nahmod VE, Stalla GK, American Psychiatric Association (1994): Diagnostic and Statistical Holsboer F & Arzt E (1996): Glucocorticoids inhibit the autoregu- Manual of Mental Disorders, 4th ed. Washington, DC: American latory induction of interleukin-1 in monocytes after endotoxin Psychiatric Association. stimulation. Neuroimmunomodulation 3, 227 – 232. Bessler H, Karp L, Notti I, Apter A, Tyano S, Djaldetti M & Weizman R Silber TJ & Chan M (1996): Immunologic fluorometric studies in (1993): Cytokine production in anorexia nervosa. Clin. Neurophar- adolescents with anorexia nervosa. Int. J. Eat. Disord. 19,415– 418. mac. 16, 237 – 243.

European Journal of Clinical Nutrition