Effective Treatments for Anorexia and Food Refusal in Children

Hannah Vaughan

"Eating Disorders in children and adolescents represent potentially life-threatening, debilitating conditions that impede physical, emotional, and behavioral growth and development. If treated soon after onset, childhood and adolescent eating disorders have a relatively good prognosis; however, if not treated, they may become chronic conditions by adulthood with devastating and sometimes irreversible medical, behavioral, and emotional consequences (Robin, Gilroy, and Dennis, 1998, pp421)."

Feeding problems and eating disturbances in toddlers and early school age children are not particularly rare. They occur in 25%-40% of the population (Kerwin, 1999). However, severe eating disturbances requiring treatment occur in only 3%-10% of young children (Kerwin, 1999). The most common eating disturbance in young children involves some form of food refusal, which some believe may be related to anorexia (Robin et. al., 1998).

Very little is known about eating disorders in pre-pubertal children, especially those who are otherwise healthy and have no apparent developmental disabilities (Singer et. al., 1992). According to Singer et. al., "Some authors have conceptualized eating disorders in pre-pubertal children as early manifestations of anorexia nervosa, but these children rarely met DSM-III-R diagnostic criteria" (Singer et. al, 1992, pp 847). Although children often do not meet the technical requirements to be diagnosed with Anorexia Nervosa, they often fall into the category of eating disorders not otherwise specified (Robin et. al., 1998). And, the health risks for these children can be are just as serious as those risks for patients with a clear diagnosis.

Children who refuse food have similar health risks to people diagnosed with anorexia. Food deprivation, for any reason, can cause physical problems (Kerwin, 1999). Some physical factors that are affected by poor diets include the following: anatomical abnormalities, sensory perceptual impairment, motor dysfunction, oral motor dysfunction and respiratory cardiac and gastrointestinal problems (Kerwin, 1999). Additionally, eating disturbances may place the child at great risk for aspiration, malnutrition, invasive medical procedures, admission to an inpatient unit for treatment of the problem, and limitations in social emotional and educational functioning and development (Kerwin, 1999).

Early treatment of eating disorders is extremely important in caring for the health of children and adolescents. There are many treatments currently available to treat a variety of eating disorders. Some that will be discussed include anti-depressant medication, hospitalization, individual psychotherapy, family therapy, classical and operant conditioning, and cognitive restructuring. The purpose of this paper is to explore different treatments for anorexia and food refusal in children, and determine which treatments are most effective.


Children and Adolescents with Anorexia can be hospitalized in pediatric or psychiatric units equally effectively (Robin et. al., 1998) as long as the impatient care environment is safe, firm, caring, highly structured, and not punitive. The main goal of hospitalization is nutritional stabilization. Inpatient treatments tend to focus more on physical health than complete psychiatric recovery (Robin et. al., 1998).

One effective strategy used in inpatient treatment is a signed contract (Robin et. al., 1998). This contract lays out all of the rules and policies of the clinic. The rules are written in an authoritative style. They are clear and concise, yet detailed and address policies regarding exercise, eating, and visitation, letting the patient know precisely what is expected of them while in treatment (Robin et. al., 1998).

Hospitalization is often favored over out-patient programs because of the amount of control and consistency that hospital settings can offer (Robin et. al., 1998). Also, patients are separated from their family setting, which relieves them from family tension and conflict that may be intensifying the eating disturbance. According to Robin et. al., one limitation of hospitalization programs is that they are expensive and third party reimbursement agents are inclined to remove patients as soon as their immediate physical ailment is treated, usually one-two weeks after admission (1998). Robin et al argue that patients admitted for eating disorders should remain hospitalized until they are eating and gaining weight regularly and have reached 85-95% of their ideal body weight (1998). In most patients this process takes 4-6 weeks (Robin et. al., 1998). Inpatient treatments can be very effective, when the patient is able to remain in the clinic until the treatment is completed.


While hospitalized, patients must be treated for their immediate physical problems. Parenteral nutrition (TPN) is an effective but costly method of treating starvation (Maloney, Pettigrew and Farrell, 1983). TPN forces protein, glucose, and calories into the bloodstream, avoiding the stomach (Maloney et al., 1983). Maloney et. al.. claim that anorectic patients prefer TPN to nasogastric forced tube feedings because they are scared of the full-stomach feeling (1983). TPN is also advantageous from the physicianís perspective because the calories cannot be expelled through vomiting.

There are some risks to TPN. Maloney states, "the central venous lines are threaded through the skin below the clavicle into a large vein returning blood to the heart"(1983, pp56). Tampering with the line can cause massive bleeding, an aire embolus, or an infection. TPN is an effective medical procedure for treating the adverse physical effects of starvation, however, it has some risks. TPN is most effective when followed by psychotherapy and family therapy (Maloney et. al, 1983).


Partial hospitalization, a day treatment program, has been a successful alternative to full hospitalization for re-feeding in children and adolescents (Robin et al., 1988). In one program, patients attended a treatment facility from 8am-10pm. In this program, adolescents were supervised after meals for one hour to prevent vomiting and between meals to prevent excessive exercise (Robin et al., 1988). Parents were encouraged to be involved in the treatment and supervision. As the patients began gaining weight consistently, they were discharged from the facility and participated in outpatient sessions three times a week. The results of a follow up nine months later show, "84% of patients had reached and retained their ideal body weight; 89% resumed menstruation; 59% overcame body image distortions; and 88% stopped ritualistic exercise" (Robin et al., 1988, pp429). These results are encouraging and lend support the Partial Hospitalization as an effective treatment method.


According to Wiener, recent studies show a strong correlation between eating disorders and depressive disorders (1984). Therefore, some have attempted to treat some eating disorders with anti-depressant medication. Several case studies and open trials using imipramine or other TCAís to treat patients with anorexia indicate that these drugs are effective. But since the studies are not double-blind or placebo-controlled, the results can not be generalized. Wiener suggests that the treatment of eating disorders with anti-depressants might only be useful when the patient also meets the criteria for major depressive disorder (1984). Wiener concludes, "medication alone is rarely a sufficient treatment for childhood disorders and should be prescribed in conjunction with psychosocial interventions within a biopsychsocial framework (Wiener, 1984, pp 836)".


Maloney et al. describe a patient who was hospitalized for severe anorexia (1983). At the beginning of her medical treatment, she was too irritable and fatigued to participate in individual or family therapy. However, she did agree to one-minute bedside visits by the psychiatrist. After the first month of treatment, she agreed to five- to ten-minute individual sessions each day. She refused family therapy until the third month of hospitalization, but then agreed to 30 minute individual and family therapy sessions. She was released soon after and two-years after hospitalization, the patient is successful in social and academic endeavors and maintains a normal weight for her height and age (Maloney et al., 1983).

Through this case study, Maloney et al. show that psychotherapy is important but can not be forced (1983). Psychiatrists must be sensitive to the physical conditions of their patients. Maloney et al. discuss the importance of traditional individual and family therapy to settle underlying interpersonal conflicts, but stress that psychotherapy must follow medical treatment. Psychotherapy without appropriate medical treatment may actually cause further problems and conflicts in the patient (Maloney, 1983).

Individual psychotherapy is important in treatment of eating disorders because these sessions focus on underlying conflicts including "ego strength, coping skills, individuation from the nuclear family, confusion about identity, and other interpersonal issues regarding physical, social, and emotional growth, and the relationship of these issues to eating, weight expectations, and body image" (Robin et. al. 1998, 430). Through interpretation and support, the therapist sets up a situation that encourages the adolescent to accept herself. Self-acceptance and a stronger ego help expel the feelings of depression, distrust, and poor identity that have perpetuated eating disturbances (Robin et al., 1983).

Family Therapy has also been demonstrated an effective method of treatment for many children (Robin et al., 1998). Family therapy typically involves three stages. First, parents are asked to take charge of the childís eating and work together to come up with strategies for re-feeding the child. Then, as the parents gain control, emphasis is placed on balanced food intake and regular weight gain. Finally, control over eating is returned to the patient and the focus of discussion switches to parent interactions and the individuation of the child (Robin et. al., 1998).

One study contrasted Ego-Oriented Individual Therapy (EOIT) and Behavioral Family Systems Therapy (BFST). Individual therapy showed slower change in restoring normal body weight, but showed positive changes in eating attitudes and depression that were equivalent to the results of BFST (Robin et. al, 1998). Another study on family therapy versus individual therapy compared four subgroups of anorexic patients: early onset and short duration; early onset and long duration; late onset and long duration; and bulimic subtype (Robin et. al, 1998). Eighty patients were followed after they were discharged from an inpatient re-feeding program. Those in the early onset, short duration category showed more favorable outcomes with family therapy than with individual therapy. Individual therapy showed more favorable outcomes than family therapy on weight gain for the late onset anorexia group. There were no significant differences in treatment for the other two groups. Patients in the early onset subgroups were more likely to drop out of individual therapy and patients in the late onset subgroup were more likely to drop out of family therapy (Robin et. al, 1998).


Operant Conditioning has often been used to treat anorectic-like eating problems in children. However, there are few studies that examine classical conditioning procedures (Siegel, 1982). One case study by Lawrence Siegel, shows how both operant and classical conditioning techniques have been used to effectively treat an eating disorder in a six year old.

Parents of the 6-year-old child sought treatment because of their sonís extremely limited diet and lack of appropriate weight gain for his height and age. Ever since an accidental overdose of paragoric, for diarrhea treatment, when he was two, the childís diet has been extremely limited (Siegel, 1982). He ate only soft foods and would gag and vomit if any new food were placed in his mouth for even a second. As he aged, his diet remained extremely limited and he became nauseated at the dinner table when family members were eating. The eating disturbance was the only way that this child was manifesting problems. He was socially adjusted and academically above average (Siegel, 1982).

Treatment began with the use of operant procedures (Siegel, 1982). For a week, the childís parents kept a record of the foods eaten at each meal to assess the types of foods and eating patterns of the child. Then, for two weeks, the child was required to eat a small piece of food with protein with his typical dinner of cereal. If the child ate the protein, he was allowed to watch T.V. and play with his siblings in the evening after dinnertime. If he did not eat the protein, he was required to stay in his room after dinner until bedtime (Siegel, 1982). These first three weeks were successful.

Then, for two weeks, the child was required to eat two different foods with dinner (Siegel, 1982). He was only able to do this successfully on two days. The rest of the evenings he vomited or gagged and refused to finish the meal. During the 6th week of treatment, the child was given a second chance later in the evening to return to the table and finish dinner (Siegel, 1982). This was unsuccessful; the child ate only his bowl of cereal during the week (Siegel, 1982). Because these operant procedures were no longer facilitating treatment, they were combined with classical conditioning procedures.

In the 7th week, the child was asked to simply smell the dinner food that the rest of the family was eating (Siegel, 1982). He was successful at this task and his behavior was reinforced by spending the evening with his siblings. Then for two weeks, he was required to touch a different piece of food to his tongue each night at dinner. He successfully completed this task too. For two more weeks, he was required to place a piece of food in his mouth and chew it for several seconds (Siegel, 1982). He was to do this while watching television. He was allowed to spit the food out if he chose to, but sometimes he accidentally swallowed it. Television served as a distracting activity to eliminate the childís anxieties about chewing the food (Siegel, 1982). This was successful.

In the 12th and 13th weeks of treatment, the child was to begin swallowing the food (Siegel, 1982). He was also allowed to watch television during this time. He was successful at this and showed no signs of gagging or vomiting. During the final 4 weeks of treatment, the child was required to eat an increasing number of table foods and the amount of cereal that he ate gradually decreased, until it was completely eliminated as a dinnertime food (Siegel, 1982).

Overall, this study shows that operant procedures alone are not always successful in treating children with food disturbances. However, the combination of operant procedures with classical conditioning procedures appears quite effective (Siegel, 1982). Similar studies agree that classical conditioning is an important component for effective treatment. One study of a two-year-old who refused food, demonstrated that eating (unpleasant task) while bathing (pleasant task) was effective (Delgado, et. al., 1993).

Another study agrees that combining operant and classical conditioning procedures is important. Luiselli and Gleason studied sensory reinforcement and texture fading procedures in a deaf, visually impaired 4-year-old (1987). The child was reinforced for specific eating behaviors with contingent presentation of light and rocking behaviors until she could sit at the table and feed herself. Additionally, solid foods were gradually consumed using texture fading. Increasingly, the child was given thicker foods, greater variety of food types, and more items at each meal. Operant and classical conditioning procedures were effective methods of treating this child.


An article by Singer et. al. discusses several boys treated with cognitive-behavioral therapy (1992). The boys developed severe eating disorders that resulted in significant nutritional deficiencies and/or growth retardation. Some of the treatment was operant conditioning. In addition to operant conditioning procedures, it was believed that the boys would benefit from cognitive therapy (Singer, 1992). Often cognitive therapy is not employed when it is hypothesized that the patients are too young to understand (Singer, 1992).

One child, after choking on a hot dog gradually began limiting his food until he eliminated all solid foods from his diet (Singer, 1992). His behavioral treatment began with a hierarchical list of foods that were increasingly difficult for him to eat. These foods were introduced to him sequentially and he received positive reinforcement for increased variety and amount of food consumed. The parents were trained in contingency management. By the end of treatment, this child was eating a variety of age appropriate foods and two years later maintained normal eating behaviors and body weight (Singer, 1992).

Another child, after being hospitalized for pneumonia at 18 months, refused most solid foods consistently for 5 years (Singer, 1992). His diet when brought in for treatment consisted of only ice cream, mashed potatoes, pretzels, rolls, muffins, chocolate milk, and ginger ale (Singer, 1992). His cognitive treatment included education about the importance of eating enough food and a variety of it. His behavior treatment consisted of a list of "feared" foods and "mastered" foods. Each day he had the opportunity to eat some foods on his "mastered" list and was required to eat an increasing number of foods from his "feared" list (Singer, 1992). He also received stickers and was able to wear a badge if he ate at least 50% of each food on his dinner plate (Singer, 1992). Upon release from the hospital, his weight had increased and five months later, he continued to increase the variety and amount of foods in his diet (Singer, 1992).


Both Physical health and Mental Health are important when treating eating disorders. But often the providers of physical health and mental health services work in isolation. Maloney et. al. believe that its very important for professionals to collaborate when treating eating disorders in adolescents and children (1983). Robin et. al. and The Society for Adolescent Medicine agree, "children and adolescents with eating disorders require management by an interdisciplinary team consisting of physicians, nurses, dieticians, and mental health professionals" (1998, pp426).

Collaboration is especially important in determining the treatment process. Maloney et al. suggest that some patients have immediate physical needs that cannot be ignored and so psychological therapy must wait (1983). However, some patients require less of a medical attention and are better able to tolerate psychological treatment earlier. Communication between professionals is imperative for determining the proper treatment for the patient. Some psychologists make the situation worse for the patients by demanding too early that they begin treatment (Maloney et al., 1983).


Many of the above treatments have been shown effective in various aspects of eating problems. Hospitalization, Partial Hospitalization, and TPN all succeeded in restoring the patientsí body weight. Individual Psychotherapy succeeded in teaching patientsí self-acceptance. Family Therapy was successful in changing patientsí attitudes towards eating. Behavior Therapy was successful in modifying eating patterns and increasing the amounts and varieties of food eaten. Cognitive Therapy succeeded in modifying eating patterns through education about nutrition and restructuring of schemas. Each treatment appears effective, in its own way.

The results of all of the studies that I read were generally positive and discussed the effectiveness of particular approaches. However, it is important to realize that all of these studies have limitations and therefore, the results may not be as promising as they appear. Some general problems in studying the treatments for eating disturbances include, "inconsistent definitions of eating problems, inferior experimental designs, small sample sizes, inadequate or missing control groups or conditions, absence of standardized outcome measures, and incomplete descriptions of interventions" (Kerwin, 1999, pp195).

In all of the studies I reviewed, the sample size was one. It is very hard to find research done with more than one patient. Perhaps, this is because each patient has different symptoms and the patients are not all categorized together under any specific heading. This classification difficulty results from a lack of clear definitions of eating disturbances. Because severe eating disorders in children are not particularly common, appearing in only 3-10% of the population (Kerwin, 1999), it is also possible that there arenít enough cases at any one time to do a multiple-subjects study. Either way, I only looked at one or two studies per treatment philosophy. And, studying only one or two people is not helpful for making predictions about the general public. The results from these studies can not be generalized.

Another limitation of the studies I reviewed is that it is often hard to separate the various treatment procedures from each other. In several of the studies, it appeared that more than just one factor was influencing the onset and recovery of eating disorders. For example, during hospitalization, children are separated from their parents, which may serve as negative reinforcement for not eating. Additionally, hospitals are highly structured environments and this may factor into the recovery of a child. In Maloney et al.Ďs study, the treatment was a combination of medical procedures and psychotherapy and Singerís study combined both behavioral and cognitive approaches. In these studies, it is hard to tell if the recovery was a result of one factor or both.

The treatments discussed above are common procedures utilized in combating eating disturbances in children. Some treatments have been more thoroughly researched than others. The area in which there is most research is behavioral treatments. Perhaps, I should have limited my review to behavioral treatments in young children with eating disturbances. Then, I could have explored operant and classical conditioning procedures in more detail and generated more information on which specific behavioral interventions proved most beneficial. The topic I chose may have been too broad. However, I did learn about different treatment procedures and have formulated an educated opinion on which treatments are most effective.


Delgado, S.V., Emde, R.N., Pope, K.K. (1993). An atypical eating disorder in a 2 year

old female. Bulletin of the Menninger Clinic 57(2). 242-251.

Kerwin, M.L.E. (1999). Empirically supported treatments in pediatric psychology:

Severe feeding problems. Journal of Pediatric Psychology 24(3). 193-214.

Luiselli, J.K., Gleason, D.J. (1987). Combining sensory reinforcement and texture fading

procedures to overcome chronic food refusal. Journal of Behavior Therapy and Experimental Psychiatry 18(2). 149-155.

Maloney, M.J., Pettigrew, H., Farrell, M. (1983). Treatment sequence for severe weight

loss in anorexia nervosa. International Journal of Eating Disorders 2(2). 53-58.

Robin, A.L., Gilroy, M., Dennis, A.B. (1998). Treatment of Eating Disorders in Children

and Adolescents. Clinical Psychology Review 18(4). 421-446.

Siegel, L.J. (1982). Classical and operant procedures in the treatment of a case of food

aversion in a young child. Journal of Clinical Child Psychology 11(2). 167-172.

Singer, L.T., Ambuel, B., Wade, S., Jaffe, A.C. (1992). Cognitive Behavioral treatment

of health impairing food phobias in children. Journal of the American Academy of Child and Adolescent Psychiatry 31(5). 847-852.

Winer, J.M. (1984). Psychopharmocology in childhood disorders. Psychiatric Clinics of

North America 7(4). 831-843.



Psychology Department

The Health Psychology Home Page is produced and maintained by David Schlundt, PhD.

Vanderbilt Homepage | Introduction to Vanderbilt | Admissions | Colleges & Schools | Research Centers | News & Media Information | People at Vanderbilt | Libraries | Administrative Departments | Medical 

  Return to the Health Psychology Home Page
  Send E-mail comments or questions to Dr. Schlundt


Search: Vanderbilt University
the Internet

  Help  Advanced

Tip: You can refine your last query by searching only the results by clicking on the tab above the search box

Having Trouble Reading this Page?  Download Microsoft Internet Explorer.