Treating Eating Disorders: the Pitfalls and Perplexities

August 1, 2004

Patients with anorexia nervosa often attempt to deceive health care professionals because they do not want treatment for their disorder. Thus, physicians must maintain a high index of suspicion for signs and symptoms of AN due to its potentially fatal complications.

Patients with anorexia nervosa (AN) generally do not want to be treated and will try to deceive primary care physicians and mental health care professionals by denying the psychological aspects of their disorder and the seriousness of their behavior, warned Katherine Halmi, M.D., who has treated patients with eating disorders for more than 30 years.

Halmi, professor of psychiatry and director of the Weill Medical College of Cornell University's eating disorders program at New York-Presbyterian Hospital, Westchester Division (White Plains, N.Y.), recently cautioned attendees at the 2nd World Congress on Women's Mental Health that, despite individuals with AN being underweight, they are often able to convince primary care physicians and others that they do not meet AN diagnostic criteria.

"These patients will go onto Web sites and find out all kinds of information on how to trick the doctors. So if a physician asks them, 'Are you afraid of becoming fat?', most of them will answer 'no,'" said Halmi.

Consequently, if an individual comes in underweight and suffering from constipation, amenorrhea and bradycardia, the physician should consider the possibility of AN, regardless of what the patient says, Halmi advised. Also, it may be necessary to interview the patient's family members to learn that the patient is severely dieting or exercising excessively. Laboratory findings will help clarify the diagnosis, but they differ somewhat depending on whether the individual has the restricting or binge-eating/purging type of AN (see Table for possible findings on physical examination with AN or bulimia nervosa [BN]).

"If [patients] are just restricting food intake and not binging and purging, their laboratory tests will show a low white blood count, and mildly depressed T3, and a low fasting blood sugar," said Halmi.

If they have the binge/purge type of AN, Halmi said, the patients might have other problems such as electrolyte abnormalities (e.g., hypokalemia) and elevated serum amylase concentration due to enlarged salivary glands.

Patients with AN may also have electrocardiographic abnormalities, she said, which are "far more serious, since these patients can die during a cardiac arrest or arrhythmia."

If they are taking syrup of ipecac to induce vomiting, patients with AN may suffer irreversible myocardial damage and can die of congestive heart failure.

Because patients with AN do not want to be treated, few participate in treatment trials. According to Halmi, there are fewer than 15 randomly assigned, controlled studies for AN treatment in the medical literature.

"[Individuals with AN] will not enter treatment trials, and when they do, they will frequently drop out after two or three weeks or they develop medical complications and have to be withdrawn," she said. "In all of literature, [there are] only five randomized, placebo-controlled trials with antidepressants [e.g., Kaye et al., 2001], and only one of these had more than 40 patients." Additionally, most of the trials were conducted in a hospital environment designed to encourage weight gain among the patients, which is not really a good environment to determine the effectiveness of the drug.

Some studies have sought to compare nutritional counseling with cognitive-behavioral therapy (CBT) for AN, Halmi said. In a study by Pike et al. (2003), 33 patients with AN posthospitalization were randomly assigned to one year of outpatient CBT or nutritional counseling. The overall treatment failure rate (relapse and dropping out combined) was significantly lower for CBT (22%) than for nutritional counseling (73%).

"The message here is that CBT is the treatment of choice, and nutritional counseling should be made a part of the CBT," said Halmi.

However, she warned that patients with AN generally will not accept medication treatment alone, nor would she advise it. She cited an example of an unpublished collaborative study conducted by Stanford University, the University of Minnesota and Cornell University in which researchers randomly assigned patients with AN who were within 80% of their target weight to fluoxetine (Prozac), a manualized form of CBT or a combination of the two. Dropout rates among study participants were high: 23 patients (66%) receiving fluoxetine dropped out compared to 18 (49%) receiving CBT and 19 (53%) receiving the combination.

Moving the discussion from AN to BN, Halmi said the treatment of BN is complicated by the frequent presence of comorbid psychiatric conditions. She cited one study that found that 84% of a clinical sample of women with BN had a lifetime affective disorder, and 44% had a lifetime alcohol or substance abuse disorder (Bushnell et al., 1994). She also cited a study that looked at comorbidity in a sample of 59 female patients with BN (Brewerton et al., 1995). Three-quarters of the patients had a mood disorder (63% had major depressive disorder), 36% had an anxiety disorder and 20% had a substance abuse disorder. Bulimia nervosa has been found to be comorbid with Axis II disorders as well (Braun et al., 1994).

"Thus, one is never treating just binge/purge behavior," Halmi said.

Persuading patients with BN to enter treatment trials is much easier than persuading those with AN, according to Halmi. Patients with BN tend to enter trials and stay three or four months.

More than 30 placebo-controlled, randomly assigned drug studies have shown all antidepressants to be effective in reducing binge-eating/purging behavior, Halmi said.

"However, they only produce complete abstinence of binge/purge behavior in 20% to 30% of the bulimic population," she added. (For further discussion of 19 trials of antidepressants versus placebo for individuals with BN, see Bacaltchuk and Hay [2003]--Ed.)

More than 20 randomly assigned, controlled studies have shown CBT to be effective in reducing binge/purge behavior in 90% of patients with BN and producing abstinence from such behavior in about 30% to 40%.

Currently, the first line of treatment for BN is CBT, Halmi said, since after 20 weeks of treatment more patients stayed abstinent from their binge/purge behavior with CBT than those who were on medication alone.

More effective therapies need to be developed for AN and BN, Halmi emphasized. One potential new treatment for BN may be topiramate (Topamax), she said. Hoopes et al. (2003) conducted a randomized, double-blind, placebo-controlled trial to assess the efficacy and safety of topiramate in BN. The researchers concluded that topiramate was associated with significant improvements in both binge and purge symptoms in the study population.

Looking at relapse issues, Halmi described a treatment study of patients with BN who had received CBT (Halmi et al., 2002). This multisite study examined relapse in 48 patients with BN who had responded to CBT with complete abstinence from binge-eating and purging. Structured interviews and questionnaires were used to assess patients before and after treatment and at four months after treatment. At four months posttreatment, 44% had relapsed. Characteristics of those who relapsed included shorter duration of illness, fewer weeks of abstinence during treatment, a higher rating on severity of eating disorders at the end of treatment, less motivation for change and lower self-esteem.

Economic factors impacting relapse rates, according to Halmi, are the limits imposed by managed care organizations on hospital length-of-stays for patients with eating disorders. Studies from Australia and the University of Iowa analyzing patients' body mass index (BMI) at discharge from inpatient treatment found that patients discharged with a BMI <19 had a significantly greater risk of relapse compared with those who had a BMI ≥19. At the Weill Cornell program, Halmi said, the median discharge BMI from inpatient treatment fell from 19.5 in the 1980s to 17.5 in 1999. Meanwhile, the mean length of stay dropped from an average of 140 days in the early 1980s to 23 days in 1999.

During the 1980s, Halmi said the program had only five or six readmissions in the entire decade; now, one-third of admissions are readmissions. The limitations imposed by health maintenance organizations and managed care, she said, are often an "enormous detriment to the patients" and have resulted in higher relapse rates and more expensive care over the long term.

References:

References


1.

Bacaltchuk J, Hay P (2003), Antidepressants versus placebo for people with bulimia nervosa. Cochrane Database Syst Rev 4:CD003391 [update].

2.

Braun DL, Sunday SR, Halmi KA (1994), Psychiatric comorbidity in patients with eating disorders. Psychol Med 24(4):859-867.

3.

Brewerton TD, Lydiard RB, Herzog DB et al. (1995), Comorbidity of axis I psychiatric disorders in bulimia nervosa. J Clin Psychiatry 56(2):77-80.

4.

Bushnell JA, Wells JE, McKenzie JM et al. (1994), Bulimia comorbidity in the general population and in the clinic. Psychol Med 24(3):605-611.

5.

Halmi KA, Agras WS, Mitchell J et al. (2002), Relapse predictors of patients with bulimia nervosa who achieved abstinence through cognitive behavioral therapy. Arch Gen Psychiatry 59(12):1105-1109 [see comments].

6.

Hoopes SP, Reimherr FW, Hedges DW et al. (2003), Treatment of bulimia nervosa with topiramate in a randomized, double-blind, placebo-controlled trial, part 1: improvement in binge and purge measures. J Clin Psychiatry 64(11):1335-1341.

7.

Kaye WH, Nagata T, Weltzin TE et al. (2001), Double-blind placebo-controlled administration of fluoxetine in restricting- and restricting-purging-type anorexia nervosa. Biol Psychiatry 49(7):644-652.

8.

Pike KM, Walsh BT, Vitousek K et al. (2003), Cognitive behavior therapy in the posthospitalization treatment of anorexia nervosa. Am J Psychiatry 160(11):2046-2049.