اثر سیتالوپرام در بی اشتهایی عصبی: یک مطالعه مقدماتی

Introduction: Anorexia nervosa (AN) still lacks a defined treatment. Since fluoxetine proved effective in weight-restored anorexics, this pilot study evaluates the efficacy of another SSRI, citalopram, in restricting-type AN. Experimental procedures: Fifty-two female anorectic outpatients were randomized in the citalopram (n=26) and waiting list (n=26) as a control group. Efficacy was assessed using Eating Disorder Inventory-2, Eating Disorder Inventory-Symptom Checklist, State-Trait Anger Expression Inventory, Beck Depression Inventory, Symptom Checklist-90 and Structured Clinical Interview for DSM-IV Axis II Disorders. Results: Thirteen patients dropped-out, thus 19 patients received citalopram and 20 remained in the control group. After 3 months of treatment, the citalopram group showed a decrease on BDI and SCL-90 Depression subscale and an improvement of baseline obsessive compulsive features on SCL-90, EDI-2 impulsiveness and Trait-anger on STAXI. Weight gain was similar in the two groups. Discussion: These preliminary results support the efficacy of citalopram in anorectics. Citalopram seems to improve depression, obsessive–compulsive symptoms, impulsiveness and Trait-anger.

Anorexia nervosa (AN) is a complex disorder that occurs predominantly in young women (American Psychiatric Association, 1994). It is believed to be determined by many factors but its etiology is uncertain. As a consequence there is still no proven or unequivocal treatment for this disorder (Ferguson et al., 1999; Herzog et al., 1992). According to some authors (Jimerson et al., 1996) adjunctive pharmacotherapy in the treatment of hospitalized patients has few benefits. Other authors have reported as useful clomipramine (Crisp et al., 1987), cyproheptadine hydrochloride (Halmi et al., 1986; Goldberg et al., 1979; Vigersky and Loriaux, 1977) and monoamine oxidase inhibitors (Johnson et al., 1983; Hudson et al., 1985). Recently, numerous studies using selective serotonin reuptake inhibitors (SSRIs) have been conducted, but the results are inconsistent. Strober et al. (1997) demonstrated that fluoxetine, in addition to other therapies, did not significantly improve outcome. An open-trial study has, however, demonstrated the effectiveness of fluoxetine on weight, obsessive thoughts, and depression (Gwirtsman et al., 1990). Further studies to explain these different results suggested that the effectiveness of fluoxetine depends on the clinical condition of the patient. Kaye and co-workers found that in weight-restored AN patients fluoxetine improved outcome and reduced relapse (Kaye et al., 1991; Kaye, 1997). Attia et al. (1998) demonstrated the ineffectiveness of fluoxetine in underweight patients. A retrospective study further demonstrated that all SSRIs are ineffective in malnourished underweight patients with AN (Ferguson et al., 1999). This hypothesis is also supported in a review by Kaye et al. (1999). It is as yet unclear whether fluoxetine or other SSRIs can be used before complete weight restoration as it is difficult to derive consistent responses from the above-mentioned studies. Until recently only a few reports in the literature have discussed the use of other SSRIs in the treatment of AN (Calandra et al., 1999; Raitasuo et al., 1998; Bergh et al., 1996). Further explorations are justified because these compounds are known to influence the depressive and obsessive aspects of AN, which are extremely important from clinical and genetic viewpoints (Lilenfeld et al., 1997; Casper, 1990; Kaye et al., 1998; Collier et al., 1997; Enoch et al., 1998). Besides, as demonstrated by studies on the cerebrospinal fluid 5-hydroxyindoleacetic acid (5-HIAA) factor (Kaye et al., 1998) and on the temperamental trait Harm Avoidance (Cloninger et al., 1993; Fassino et al., in press), some pathological behaviors common in AN could be linked to serotonergic activity. Citalopram is the most recent SSRI. It is characterised by a more selective pharmacological profile with respect to other compounds of this class, and it is safe (Bezchlibnyk-Butler et al., 2000; Tan and Levin, 1999; Feighner and Overo, 1999). The aim of this pilot study was to examine the efficacy of citalopram in the treatment of outpatients suffering from restricting-type AN. Primary outcome measures were eating symptoms and psychopathological symptoms were considered. Moreover, one objective was to verify whether changes noted after pharmacological treatment are independent from possibly confounding variables such as age, years of disease, body mass index (BMI), and personality disorders.

The use of citalopram in restricting-type AN outpatients demonstrated promising results, also because these improvements were not influenced by important factors such as duration of the disease, age, weight, or presence of personality disorders. Furthermore these results also provide insight into some psychopathological traits that have not received much consideration to date. On the other hand there are three main limitations to our study. First, the effect on depression complicates the analysis because it is not clear whether some of the clinical improvements are secondary to an effect of citalopram on depression. Second, our study was not designed for very low weight anorectic patients, because it was conducted in an outpatient department. Even though we have found no differences in the outcome improvement between patients with higher BMI and patients with a lower one, the absence of severely underweight anorectics in our sample does not allow to generalize our results. For example subjects with a mean BMI over 16 could respond to antidepressant medication in a different way from the very low weight anorexics (Ferguson et al., 1999; Kaye et al., 1991; Attia et al., 1998; Kaye et al., 1998). Last, a methodological weakness of this study is that its design was not double-blind, placebo-controlled. Such a design would have decreased the halo effect. In the present study the researcher might have a tendency to overestimate the improvements obtained in the citalopram group in anger and impulsiveness, owing to the improvement of both depressive and obsessive symptomatology due to the assumption of the drug. Of course, the halo effect decreases the confidence in the results, though the perceivers’ overestimation of the results obtained might have been partly reduced by the use of self-administered questionnaires. Notwithstanding this methodological bias we believe that some interesting considerations on the use of citalopram in anorexia nervosa emerge from the present study, especially considering that up to now in literature there are not yet specific studies on this subjects, not even with an open design. The use of a no treatment control group in the present study, with respect to an open trial, allows at least the distinction of the drug effect from some non-specific factors which might contribute to healing, first of all, the natural illness course. Moreover, as regards the control group, subjects were visited as outpatients and administered questionnaires in the same way of the citalopram group; this might have elicited a placebo response or ‘response to the healing situation’ as a consequence of the symbolic impact of treatment setting (Papakostas and Daras, 2001). This placebo effect might in part balance the placebo effect which has certainly contributed to the improvement of patients in the citalopram group, thus increasing the possibility to compare of the two groups. In conclusion, stronger evidence will be available with randomized, placebo controlled trials, evaluating the efficacy of citalopram in a larger and more comprehensive sample. Such trials are in fact still the best way to perform a clinical trial sustaining the efficacy of a drug (Montgomery, 1999). If the preliminary results of this pilot study will be confirmed citalopram may be particularly beneficial in the therapy of AN for its benefits on depression and obsessive–compulsive symptoms and its action on some aspects of the pathogenic core of this disease, such as anger and impulsiveness.