RESIDENTIAL TREATMENT FOR ANOREXIA NERVOSA

The Germaine Lawrence School (GLS), a residential treatment center for adolescent girls, has been working with girls having life threatening psychosomatic illnesses, especially anorexia nervosa, for nine years.

When we started serving several girls with anorexia, staff were apprehensive. They were afraid of being responsible for a child with a life threatening condition. Could they keep her safe?
Staff also wondered what services they could provide that would help the girls recover from their disorder. After all, if the hospitals had not cured them, what did they have to offer? What could they do to put an end to the girl's need for hospitalization?

This paper examines the services GLS provided sixteen anorectic girls and their progress after discharge from GLS.

By studying files and and conducting an outcome study through interviewing ex-students and their parents, we attempt to resolve the two key issues :

1. Were anorectic students safe? And, if so, how were they kept safe?

2. Did GLS provide services that formed an effective treatment program for this population? What, if anything, did GLS have to offer adolescent girls with anorexia?

The Population

In most ways the anorectic girls served by The Germaine Lawrence School are quite similar to the majority of anorectics:

• they have starved themselves, losing at least 25% of their body weight
• they are preoccupied with being thin, having a radical distortion of their body image and believing they are fat
• they have developed obsessions and rituals concerning food preparation and eating
• they go to extreme lengths to control their weight including vomiting, laxative abuse, and diet pill abuse.

Compared to groups of anorectics studied elsewhere, however, our students seem to have more persistent problems.
As Table 1 shows, 95% of the girls in this study were hospitalized more than once prior to referral due to eating disorders; 56% had three or more hospitalizations.
The average number of hospitalizations was 3.4. In other studies (Dally, Hall, Minuchin and Morgan) the proportion of subjects being hospitalized only once was significantly lower, ranging from 20% to 50%.
No other study in our literature search described a population with so many multiple hospitalizations.

The anorectics served by The Germaine Lawrence School, then, represent the more difficult, more treatment-resistant portion of the general anorectic population.
They have serious, chronic illnesses that have not responded to outpatient therapy and multiple in-patient hospitalizations. Only after a series of hospitalizations are they referred to residential treatment.

Conceptualization of the Disorder

From our point of view, anorexia starts as a simple diet during a time of stress that develops into an obsession when it proves to be a successful way to reduce that stress.
For our students, however, the stressors they faced were quite intense and their eating disorders helped them cope with a variety of overwhelming problems including:

• emotional, physical, and sexual abuse
• severe retardation of age-appropriate autonomy due to intrusive, overly protective parents
• inability to gain approval from overly demanding parents, resulting in severely low self-esteem and a paralyzing lack of self-confidence

• for adolescents who have felt controlled by others, especially when that control has been connected to emotional or physical abuse, an eating disorder provides a vehicle for the girl to feel in control
• the obsession with caloric intake and exercise becomes such a complete focus that inner upset and turmoil is avoided or at least not perceived as acutely
• for some adolescents with a compromised sense of self-esteem, an eating disorder places her in the role of a sick, dependent child and allows her to avoid what she perceives as the overwhelming challenges of adult life
• in a society that values thinness and the ability to restrain caloric intake, the anorectic girl may feel virtuous and competent

Moreover, for a girl who has never been able to satisfy her parents' insatiable demands, an eating disorder can be the proving ground of a competency that approaches perfection: a life-threatening illness often gives the girl increased attention from one or both parents, and/or the illness becomes such a focus of parental concern that the parents, who had been in serious conflict, have an issue that they can unite around and tension between the parents decreases.

Anorexia, then, is yet another maladaptive coping mechanism resulting from some form of inadequate psychosocial development. The syndrome is just another way of responding to the inability to cope with problems that often intensify during adolescence.

Other responses, like substance abuse or delinquent behaviors, may lead to court involvement and punishment, obviously making the behavior less rewarding. Anorexia, however, leads to medical intervention which, for the seriously disturbed adolescent, may not alter the rewards of the behavior at all.

Hence for some girls, the benefits of remaining anorectic are so powerful, and the costs so minimal, that they may be unwilling or unable to give up the illness.

The Program

Students with eating disorders are conceptualized and treated as the rest of the population in most respects.
In general this means we view their anorectic symptoms as maladaptive behaviors resulting from impaired psychosocial development. Hence GLS makes their psychosocial development, rather than eating per se, the focus of their treatment.

The treatment program at GLS is centered on milieu therapy that promotes psychosocial growth through nurturance, consistency and structure (Hirshberg, 1986).

Specific services include a special education school, a highly structured dormitory schedule that includes recreational activities, individual therapy weekly, group therapy daily, and family therapy bi-weekly.

The behavior management system includes a point and level system through which students receive feedback and earn privileges. Individualized contingency contracts are used with most students to target problem behaviors and to promote the development of age appropriate skills.
For example, criteria for both discharge and certain important privileges include age appropriate behaviors such as interacting with peers, compromising with adults, and holding down a job.

We believe three aspects of our program that often promote psychosocial growth may be especially significant for our anorectic students.

First is increased physical and psychological separation from their families. Simply living away from their family for an extended period of time, may provide these girls with the space to develop more autonomy. Also, through family work, we can help the family develop appropriate boundaries that will assist the adolescent separate psychologically from her parents.
Specifically, we can help parents learn to make proper demands on their daughter, empowering parents to establish expectations and consequences, on the one hand, and blocking parents from intruding into areas that the child should be able to control on her own, on the other.

Another significant aspect of the program is our behavior management system that establishes consistent expectations, rewards and consequences. It requires increased age appropriate behavior to earn privileges.
With anorectics we use this approach to achieve two important objectives:

1. To decrease the rewards of the anorectic behaviors while creating costly consequences for continuing those behaviors

2. To promote the development of age appropriate autonomy, interests and competencies

For example, to change eating habits, we make powerful rewards available contingent upon attaining specific weights. For most girls with eating disorders, the most important reward is visiting home and, ultimately, discharge to home.
Thus, we often make weekend visits home contingent upon reaching a safe weight, an unusual practice with the rest of our population but one that usually is effective with anorectics.
Depending upon the eating pattern she chooses, she attains a weight that determines whether or not she visits home. This strategy avoids struggles about food at each meal, and gives the girl choices about what to eat with the knowledge that her choices will have important consequences for her.
As Minuchin suggests, this makes the child responsible for her own condition and she learns that she can use her power to influence her own situation and to control her own activities.

The final aspect of our program that is especially pertinent to our anorectic students is the highly structured schedule that keeps the girls interacting with peers and staff in a variety of pro-social activities.
This structure forces them to participate in activities, such as sports or work that they would avoid otherwise, that promote the development of age appropriate skills and competencies.
As skills develop so does self-esteem and confidence in one's ability to function autonomously. In addition, the social interaction of such activities develops social skills and peer acceptance.
Entry into the peer culture is another critical step towards autonomy and away from continued parental dependence.

Our residential treatment program, with its focus on psychosocial growth, seems to be well designed to help anorectics recover from their disorder.
Although students with anorexia are treated as other students in most ways, their eating disorder does demand some special measures.

We developed a "Structured Eating Program" (Appendix 1), modeled on hospital protocols, to provide a clear, consistent approach for students and staff.
This eating program is "a small part of the whole approach to the (student). It is used primarily to remove eating as an area of power struggle between the child on the one hand and the staff on the other. (Minuchin, 1983. p. 113)
Our relationship with the students' physicians are also very important. The risk and anxiety that comes with working with these girls can be greatly reduced when a good relationship exists between the treatment center and the physician. It is important that the physician establish a safe weight below which the physician will become responsible for monitoring the girl and hospitalize her if necessary.
When the relationship works effectively, the treatment center staff feel supported, knowing that if the anorectic student regresses to a medically dangerous point, her physician will take responsibility and ensure her safety.
Recently, after admission to GLS a girl needed to be re- hospitalized twice because she refused to eat. After the third admission to GLS her pediatrician agreed to tube feed her if she refused to eat (as would occur in the hospital) on an out-patient basis.
When the girl refused to eat she was taken to the hospital and given the choice of eating or being tube fed. She ate and has never refused to eat again. She has not been re-hospitalized again, and she has gained weight.

We believe that many re-hospitalizations and failed cases could be avoided with this level of cooperation from physicians.

Two Cases: History, Treatment, and Follow-up:

Debby
Debby grew up in a chaotic family in which her emotionally needy parents fought bitterly and left the three children emotionally deprived and competing for the little parental nurturance available.
Debby's early attempts to gain attention and approval by being the "good girl" failed; no matter how helpful she was in the house or how well she achieved in school, she was never good enough to gain the approval and nurturance she desired from her mother.

Approaching adolescence, Debby was extremely insecure and needy of attention and approval. She felt she had to be the best, perfect, to be accepted. She would do anything to please others, but she never received adequate support and approval in return. She developed no tolerance for failure or criticism. If she could not do something the best, she would give up impulsively.

Following in the footsteps of her oldest sister, she developed the behaviors of an acting-out adolescent - drinking, drug use, running away, verbal and physical aggression.
Her eating disorder, however, soon became her primary coping mechanism and her primary problem, replacing the more common forms of acting out behaviors. Her illness was very effective in getting what she wanted - more attention and emotional involvement from her parents.

Her father started visiting her regularly after having been absent for some time. In addition, controlling her eating was an arena in which she could be successful and competent.
She tried to be the best anorectic, obsessing on food, measuring her progress, and competing with the other anorectics.

Over time, controlling her impulses to eat became a metaphor for controlling all the dangerous types of impulses she had seen destroy her family.

Control of her eating created a field of autonomy for her within which adults were powerless. Thus, as her control perfected, Debby felt safer and more nurtured. She fought attempts by others to make her gain and maintain her weight.

After four years and eight hospitalizations, having weighed 76 pounds at a height of 5'7", Debby was referred to The Germaine Lawrence School by her physician who stated, "If you cannot help her, she will die."

Although she entered the program with a contract specifying issues that she wanted to work on and behavioral criteria for returning home, for the first few months of treatment Debby showed minimal investment and minimal motivation to change.

She continued to gain weight, but she made no apparent progress in her psychosocial development.
The battlefield for waging control struggles was no longer eating, but control was still the issue. Now she struggled against rules, and verbally abused staff when they set limits on her behavior.
Her acting-out behaviors returned, including substance abuse, running away, verbal abuse, destruction of property and suicidal gestures. During this period her strong attachment to her physician and the physical separation from her parents seemed to keep her from experiencing a relapse of her eating disorder.

Due to a serious escalation of dangerous and oppositional behavior, after eight months at GLS Debby was suspended and sent home to consider whether she wanted to remain at GLS.
At that point she chose to remain at GLS and she invested herself in the program. She began to use therapy to examine her problems and feelings. In the milieu she asked for contracts to help her control her behavior.

Her behavior improved both in school and in the dormitory. Six months after the suspension, Debby and her parents decided she could live at home safely and she was discharged to her mother's house.

Three years later, Debby is living with her father and is at a normal weight with normal eating habits. She has an active social life and a boyfriend. After getting her diploma, she has held a number of jobs including her present one as a nurse's aide which she has held for 18 months.
However, she is not free of problems. She still struggles with depression and is presently on an anti-depressant medication.

Amy
The child of two professionals with no perceived marital problems, Amy had an especially powerful role in her family.

Having an infantile need for exclusive bonds with her parents, she tantrummed intensely to maintain her unique status of the only child both after the birth of her younger brother and throughout her childhood.
She remained overly dependent upon her parents and was socially withdrawn and isolated. Also, she could not tolerate frailties and imperfections in others or in herself. She said one of the reasons for not eating was to punish and deny herself for not being good enough, for not deserving all of her advantages.

Amy's eating disorder started at the age of nine when she began refusing to eat starches and sweets so she could feel more independent and in control of her life.

During the next three and a half years her eating habits became increasingly bizarre and ritualistic. She would eat rotten foods from garbage cans, or food which she had saved and refused to eat until it showed signs of mold.
She said this was the only type of food she deserved. Her food intake was so inadequate that she stopped growing at the age of ten. Amy gained weight during her first two hospitalizations but each time she returned home, she returned to her old habits.

In September 1983, at the age of thirteen, she entered the hospital for the third time weighing 37 pounds, not having grown for over three years. The hospital staff recommended residential treatment rather than a return home at the end of the nine month hospital stay. Amy's parents agreed reluctantly and she was referred to GLS.

Amy did not enter GLS without a fight. She wanted to go home and by tantrumming and hurling insults at her parents (eg. "If you really loved me you'd let me come home.") tried to intimidate them into letting her.
Our first intervention was not to admit Amy. On the one hand we told her that we would not admit her until she agreed to come to work on specific problems.
On the other hand we supported Amy's parents and hospital staff to stand firm and not permit Amy's power play to change their stance.
After six weeks and five family interviews, Amy agreed to sign a contract that mainly focused on her psychosocial development (below) and she was admitted to GLS.

Amy's Admission Contract
I want to receive help at The Germaine Lawrence School so that I can live with my family successfully.
Specifically:

1. I want to learn to be more independent from my family and not always feel like being in the center

2. I want to learn to ask for attention and help more maturely (eg. not by getting sick or taking a tantrum)

3. I want my family to operate more as individuals emotionally and not share all of our moods as if we were one person

4. I want to learn how to assert myself in positive ways

5. I want to learn to accept limits and meet people half-way

6. I want to continue to learn to express my feelings better

7. I want to improve my eating patterns and hold my weight

8. I want to catch-up in school

9. I would like to grow and get stronger

1. I don't telephone my parents whenever I am upset and when I do telephone them, I do not become overly emotional

2. I hold my weight as specified by Dr. Spock

3. I do not throw temper tantrums

4. I share my feelings directly and verbally with my family, staff and peers

5. If one of my family gets upset, the others do not share that mood

6. I ask for what I want or need verbally

7. I maintain myself on Level 5 or 6

8. I try to understand my feelings, thoughts and behavior in therapy

9. I eat reasonable lunches and dinners and only eat limited after dinner snacks

10. I do not eat so much that I vomit

11. I earn my academic credits

12. I participate in activities and physical education classes

13. I focus on these issues and not talk about coming home

At first Amy was angered by how much responsibility and how little staff attention she got at GLS compared to the hospital.
However she adapted quickly and focused on making progress on her contract so she could go home.

In the first months she spent more time with staff than with peers, and phoned her family at least twice daily. She resented not being the center of adult attention as well as having to do things she did not do well.
During her placement Amy became more autonomous and less dependent on adults. As she gained weight her eating program was relaxed, giving her more control.
Although her eating habits remained very unusual, she took responsibility to gain two-thirds of a pound per month as required to maintain her privileges.

During her stay at GLS she gained eight pounds and actually grew - one and one-half inches - for the first time in over four years.

Part of Amy's progress resulted from improved peer relations. She was accepted by her peer group and appreciated for her humor, intelligence, and enthusiasm.
The positive regard she received from her peers helped Amy feel more confident of herself and less dependent on adults for attention and approval.

Amy left having made considerable progress separating from her family, developing a better sense of herself as an independent person, and maintaining her health. She still had some significant problems - she had not yet stopped vomiting - but her return home seemed like a fair risk.

Four years after discharge Amy is at a normal weight and has a normal menstrual cycle. She also has grown to her expected height. Through high school she achieved well academically and had an active social life, although she did not date. Her main interest has been participating in theatrical productions. Next Fall she will enter a major university's theatre arts program.

The Study

All students admitted to The Germaine Lawrence School with the diagnosis of anorexia nervosa were included in the study. They and their parents were first contacted by letter and then interviewed in person or on the telephone using a structured interview. In two cases interviews were not possible (one student was deceased, one hospitalized) but information was collected from clinical documents. Information collected included present weight, status of menstruation, medical and psychosocial history since discharge from GLS, and presence of eating disorder symptomatology.

Results of the Study: Outcomes

Three methods for evaluating recovery were used. First was weight recovery. Ten of the sixteen girls (62.5%) had returned to within at least 15% of the average for their age and height. Second was the need for rehospitalization. Eight girls or 50% had not required further hospitalization for their eating disorder.

Since the literature consistently concludes that weight recovery alone is not an adequate method for rating recovery from anorexia, the most common criteria are Morgan's General Outcome categories that consider both weight and menstrual cycle. A good outcome is defined as weight within 15% of average for six months and normal menstruation. An intermediate outcome is weight only intermittently risen to within 15% of average or over 15% above average and/or continuing menstrual disturbance. A poor outcome is when body weight is still 15% below average and menstruation is sporadic or absent. Using these criteria, seven or 44% had good outcomes, four or 25% had intermediate outcomes, and five or 31% had poor outcomes including one girl who died. See Table Two for a complete breakdown.

Table 2: Treatment and Follow Up of Sixteen Anorectic Students
Student Number	Months in Residence	Follow up interval (yrs.-mos.)	Weight % of Normal	Morgan	Re-Hosp.
1	9	4	98	Int.	Yes
2	33	7	95	Good	No
3	10	7	60	Poor	Yes
4	24	5-6	91	Good	No
5	42	0-6	79	Int.	Yes
6	3	3-6	64	Poor	Yes
7	9	2	90	Int.	No
8	5	4	117	Int.	No
9	9	2	57	Poor	Yes
10	9	4	95	Good	No
11	19	3-9	87	Good	Yes
12	13	3	100	Good	No
13	13	2-10	87	Good	No
14	21	2	95	Good	No
15	18	4	Deceased	Poor	Yes
16	7	3-3	N/A	Poor	Yes

Comparisons with Other Studies

Using Morgan's General Outcome categories as a basis for comparing results of our study with those of other studies (Table 3), we find that the recovery of our students falls within the range found in other studies. The averages of the other five studies shows 43% with good outcomes compared with 44% in the present study, 31% with intermediate outcomes compared to 25%, and 26% with poor outcomes compared to 31%.

Table 3: Comparison of General Outcome Results in Six Studies Morgan's General Outcome
Study	Good	Intermediate	Poor
Hall et al (1984)	37	37	26
Touyz & Beaumont (1985)	33	41	26
Morgan et al (1983
A. Maudsley	39	27	34
B. Bristol	58	19	20
C. St. George	45	30	22
Average	43	31	26
Present Study	44	25	31

Discussion

Given the statistically small numbers of students with eating disorders that GLS has served, any conclusions drawn from the data must be tentative. It is interesting, however, that the students in our study recovered at rates quite similar to patients in other studies even though a much higher percentage of our students had received multiple psychiatric interventions. Moreover, we believe that with better medical support, specifically out patient tube feeding, our results would improve significantly. The fact that GLS has recovery rates similar to other studies suggests that our residential treatment program may be an effective therapeutic intervention for many adolescents with eating disorders.

Another significant analysis is a comparison of outcomes of students with eating disorders and those without eating disorders. A previous follow up study of GLS students (Keifer, 1984) showed that less than 20% required further hospitalization compared to 50% of the anorectic students in the present study. In addition, about 50% of the students had good adjustments, and only 22% poor adjustments compared to 44% and 31% for the anorectic students. Again, small numbers make statistically significant comparisons impossible, however these results suggest that the girls with eating disorders may be more difficult to treat effectively than the rest of our population.

We also studied our ability to keep these adolescents safe. Although medical emergencies occurred in some cases due to weight loss, requiring rehospitalization, no student died while in our care. Our system of medical review and back up has worked effectively. The one student who did die from her eating disorder had been discharged home at a safe weight six months before her death.

Conclusions

Our findings suggest that even though anorectic adolescents are a difficult population, residential treatment centers can provide safe and effective treatment for them. Not only have we worked safely with a most difficult population of anorectics, but our hunch is that our program is more effective with many resistant anorectics than hospitals. We believe this may be the case because of our focus on psychosocial development in a setting that highly structures social interaction while providing relatively little structure concerning eating. Although it seems likely that hospitals are more effective at making anorectic patients gain weight, the eating regimen in hospitals may well do nothing to develop the more autonomous, age appropriate behavior that the patients need to recover from their disorder. In fact, in many ways the dependency the girls develop on the hospital staff to eat adequately mimics the dependency they have on their parents that is associated with the disorder (Minuchin 1973).

As we practitioners of residential treatment come to understand anorexia as yet another self destructive syndrome resulting from impaired psychosocial development, then it loses its mystery and we may realize that not only is there little new to fear from life threatening psychosomatic disorders, but that they are a natural part of our work. We may find that, not only are we capable of providing safe and effective treatment to these youngsters, but it is precisely residential treatment that is best designed to provide effective treatment. Due to its focus on psychosocial growth and its ability to provide long term treatment, residential treatment centers may prove to be the more effective placement choice for many severely disturbed adolescents. Perhaps the future of residential treatment in the mental health service delivery system is to provide long term in patient treatment, while hospitals focus on short term acute treatment and assessment. Certainly this is a hypothesis that deserves further study and consideration.