Electroconvulsive Therapy in an Adolescent with Severe Major Depression and Anorexia Nervosa

Chief Complaint and Presenting Problem

J. is a 17-year-old girl in twelfth grade who was referred to the emergency department for suicidal ideation and weight loss.

History of Present Illness

J. was reported to be a “picky eater” as a child. She was not overweight in childhood but did report comments from peers that she should “work on her abs.” At the age of 15 she first felt fat and wanted to eat “healthy.” J. is an accomplished dancer and was dancing 2–4 hours per day, 5 days per week. She started using laxatives at age 15, with most recent use three months before referral. She first binged at age 16 at a frequency of once every other week. She reported a sense of loss of control, and would eat about 2000 kilocalories in one sitting. At her height of 168 cm (66 inches), her subjective ideal weight was 41 kg (90 lbs.), a goal body mass index (BMI) of 14.5 kg/m². Her maximum reported weight was reported as 55 kg (120 lbs.) at age 14, and her minimum was 42 kg (93 lbs.) at age 16, a BMI of 15.0 kg/m². Thinking about food and weight was constant, and she weighed herself every other day. She skipped two meals per day and restricted to 500 kilocalories daily, mostly eating fruit, vegetables, and granola bars. She had no history of vomiting. J. had been amenorrheic for three months prior to admission.

J. reported that she first became depressed at age 14 during the transition from middle school to freshman year of high school. J. found this transition to be difficult. At this time there were also frequent verbal altercations at home between her parents. She described low mood and social isolation, and started cutting her wrists, thighs, and abdomen with a razor blade about once per week. At age 15 during tenth grade her mood worsened, and she developed periods of passive death wishes and suicidal ideation, although she had no intent or plan at that time. Resumption of the dance season brought temporary improvement; however, at age 16 during eleventh grade J. experienced frequent crying spells and worsening social isolation, as well as a return of suicidal ideation. Her cutting behaviors increased in frequency to daily. She described the pain from cutting as a distraction from emotional pain. In the two months prior to admission, J. reported that her depression, energy, and social isolation had significantly worsened, and that she felt overwhelmed trying to balance school, work, and dance. She also reported intensified thoughts of self-harm.

J. had no history of mania, hypomania, or psychotic symptoms.

J. was brought to the emergency department by her parents after another peer reported to the school guidance counselor that there were suicidal thoughts posted on J.'s online web log (“blog”). The counselor gave these posts to her parents and asked that J. be evaluated by a psychiatrist before returning to school.

Past Psychiatric History

J. has had no past psychiatric evaluation or treatment.

Developmental History

J. was the product of a planned pregnancy, and mother received prenatal care. There were no in utero exposures or known complications during pregnancy. She was born at 41 weeks via Caesarean section. J. reached all developmental milestones on time and was described as “a happy child.”

Education History

J. did well in school, earning mostly As and Bs, and took several advanced placement classes. There was no history of repeated grades, failures, suspensions, or special educational services.

Social History

J. lived at home with her parents and younger sister. She had no history of significant romantic relationships and had not been sexually active. J. worked part-time in the summer as a dance instructor. There was no history of alcohol, tobacco, or any other drug use. There was no history of abuse, neglect, or trauma.

Family History

J.'s parents are married. They both completed high school. Mother works in human resources, and father has never worked outside the home. The parents' relationship was described by J. as “strained.” J. has a sister, 15, with whom she gets along very well.

There is no known significant psychiatric or medical history in the immediate or extended family.

Medical History

J. has a history of dyspepsia and borderline anemia. She had been followed by a pediatrician and her immunizations were up to date. She had no history of general medical hospitalizations, surgery, seizures, or head injury. She has no allergies. Her medications on admission were ranitidine 150 mg and a multivitamin daily.

Mental Status Examination

J. was a well-groomed, thin, fair skinned Caucasian female wearing a hospital gown and dark rimmed glasses. She appeared younger than her stated age. She was alert and fully oriented. She walked with a steady gait and station. She was calm, cooperative, and made fair eye contact. There was no psychomotor agitation, retardation, or abnormal movements. Speech was normal in rate, volume, tone, coherence, and articulation. There was no formal thought disorder present. Mood was described as “bad.” She reported no anxiety. Her affect was dysphoric and restricted with minimal brightening. She denied passive death wishes, suicidal ideation, or homicidal ideation. There were no hallucinations, delusions, or other indications of thought disorder. Insight and judgment were fair.

Hospital and Treatment Course

J. was initially admitted to the child and adolescent psychiatry unit; however, she was transferred to the eating disorders program three days after admission since it appeared that her mood disorder could only be successfully treated if her weight was restored. The eating disorders program is a mixed adolescent and adult, inpatient and partial hospital, behaviorally-based program focused on weight restoration, interruption of eating disordered behaviors, and relapse prevention. Treatment is provided by a psychiatrist-led, multidisciplinary team that includes nursing, social work, occupational therapy, and dietary. Patients on the eating disorders service attend three group therapy sessions per day. In addition, patients with significant mood comorbidity attend evening specialty mood disorder groups. Group methodologies include cognitive-behavioral, dialectical-behavioral, psychoeducation, and open-ended interpersonal groups.

J.'s weight on admission was 41 kg (91 lbs.), a BMI of 14.8 kg/m². Her goal weight range was 120–124 lbs. Laboratory results were unremarkable. She was diagnosed with major depressive disorder and anorexia nervosa, restricting type. During this initial 26 day admission she was treated with fluoxetine 20 mg daily, and was transitioned to the day hospital when she reached a weight of 112.9 lbs. (BMI 18.2 kg/m²). During her 27 day course at the day hospital, fluoxetine was increased to 40 mg. She was discharged to outpatient follow-up at a weight of 118.8 lbs. (BMI 19.2 kg/m²).

Shortly after discharge she began restricting calories, abusing laxatives, and cutting her forearms. She was readmitted to the inpatient service two months later. She presented at 103.8 lbs. (BMI 16.8 kg/m²) and resumed the weight gain protocol. She was eventually discharged three months later at 126.1 lbs. (BMI 20.4 kg/m²). While J.'s body image remained disturbed, she was adherent to the eating disorders treatment protocol, and her weight remained stable throughout the rest of her course.

During the first admission due to J.'s scorn for her parents and their difficulties with each other, living arrangements with another family were considered but were not instituted.

J.'s subsequent admissions focused on her mood disorder, which worsened at this time. The second admission occurred six weeks after discharge from the partial hospital, roughly three months after J. was initially admitted to the inpatient service, and she was an inpatient or in partial hospital for a further 24 weeks. Fluoxetine was increased to 60 mg, and trazodone 75 mg was added for insomnia. Up to 10 mg of aripiprazole was used for augmentation. Having little effect, aripiprazole was discontinued one month later. Nortriptyline was added to fluoxetine 60 mg and titrated to a maximum level of 106 ng/mL. Lithium augmentation was initiated and titrated to a level of 0.8 mEq/L. Additional treatment with levothyroxine 25 mcg was also briefly introduced. J. was unable to tolerate nortriptyline and lithium due to tachycardia at 141 beats per minute, so was only on this regimen for four weeks. J. continued to experience significant depressive symptoms, stating she felt worthless, wished to die, and did not believe her condition would improve. Self-injury, usually in the form of scratching herself, severely escalated, necessitating the use of mitts and wrist restraints. One incident included the use of a shaving razor to cut both wrists. After an orthopedic surgery consult, her wounds were closed. There was no tendon or vascular injury.

Given the severity and refractoriness of J.'s depression and self-harm, electroconvulsive therapy (ECT) was initiated after two independent recommendations from child psychiatrists. After a medication taper, J. began right unilateral ECT treatments, and subsequently transitioned to the day hospital. Venlafaxine XR was initiated. The patient was discharged and readmitted for self-injury or suicidal behavior two additional times.

Bilateral ECT was eventually required to bring J.'s mood closer to its baseline. Venlafaxine was slowly increased to 300 mg, though this was complicated by resting sinus tachycardia. Fluphenazine was added for augmentation and increased to 2 mg. She underwent a total of 31 ECT treatments (10 right unilateral followed by 21 bilateral) over 13 weeks. (See Table 1 for hospital course summary.)

J. was gradually weaned from the day hospital and demonstrated decreased depressive symptoms and improved ability to prevent, though not extinguish, urges to self-injure. She was ultimately able to attend graduation, as well as a trip to Walt Disney World with her family.

J. has currently been out of the hospital for 10 weeks and is attending community college. Perhaps the most striking change in J.'s presentation during ECT, in addition to reduction in suicidal ideation and urges to self-injure and improved ability to manage these urges, was the improvement of her attitude toward her parents. After the switch to bilateral ECT, J. began to brighten when her parents visited; she hugged them spontaneously, and began to identify her parents as a source of support. Parent education about eating disorders and depression improved her parents' ability to cope with the stress of J.'s illness.

Brief Formulation

In summary, J. is a 17-year-old Caucasian girl referred by her parents for emergency psychiatric evaluation after the discovery of suicidal statements on her online blog. Onset of mood symptoms began three years prior to admission with progressively depressed mood, social isolation, eating disorder behaviors, and cutting.

There was no family history of psychiatric illness to suggest a genetic risk for depression. While the patient's malnutrition likely exacerbated her depressive symptoms at the time of initial presentation, the patient's weight was not a factor for most of her course of treatment for her mood disorder. There were no additional medical factors that were thought to play a role in her depression. Secondary amenorrhea and dyspepsia complicated the presentation of anorexia; both are common features of anorexia nervosa. Secondary amenorrhea is thought to be hypogonadic in anorexia nervosa, and dyspepsia is very common, both in the starved state and during refeeding. Sinus tachycardia was present and may have been a side effect of venlafaxine or nortriptyline; it was not thought to be related to refeeding or volume depletion, since it persisted for long after weight restoration (Verri et al. 1998; Waldholtz & Andersen, 1990).

J. underwent a protracted hospital course, including four inpatient admissions and partial hospitalizations over nine months. Initial treatment with an SSRI and atypical neuroleptic, followed by a tricyclic with lithium augmentation, was unsuccessful. ECT was subsequently undertaken with an SNRI augmented by a typical neuroleptic. A prolonged ECT course of 31 treatments was necessary. Axis II issues of borderline personality features were raised, but given the severity of the Axis I diagnoses, no clear determination could be made at that time.

Conflicts with her parents at home were significant. She performed well in school and dance but felt overwhelmed by perceived pressure from her parents to excel. Restriction of eating, over-exercising, laxative abuse, and self-injury developed in the setting of this pressure.

Multi-Axial Diagnoses

Discussion

This case represents a very challenging, but not uncommon, simultaneous presentation of two clinically impairing problems, major depressive disorder and anorexia nervosa, in which the major depressive disorder is thought to be treatment refractory. It appears, in retrospect, that the roots of J.'s eating disorder may have been laid down in early childhood, culminating in onset of clinically significant symptoms in early adolescence. Not surprisingly, J.'s mood disorder symptoms also emerged in early adolescence, perhaps precipitated by puberty, transition to high school, and conflicts with parents. Suicidal thoughts, social isolation and self-injurious behavior followed. The first order of business was to disentangle J.'s mood symptoms from her eating disorder symptoms while maintaining her safety, and as a result treatment was initiated for the anorexia.

After her body weight was stabilized over the next 9 months, J. received treatment with what would appear to be adequate doses of a serotonin reuptake inhibitor, fluoxetine, followed by augmentation with an atypical neuroleptic, aripiprazole, without improvement in depressive symptoms. A switch was subsequently made to nortriptyline, a tricylic, and thus a second antidepressant class, and augmentation with lithium followed by levothyroxine. Again, there was no evidence that this combination was beneficial, and consultation for ECT was initiated. Improvement in J.'s mood symptoms finally resulted after 31 treatments (21 bilateral) in combination with venlafaxine up to 300 mg.

Several interesting issues and questions are raised by this case. Treatment resistant depression (TRD) is far from rare in adolescents. In a recent review by Drs. Maalouf and Brent, TRD is defined as the “lack of adequate clinical response to an appropriate dose of evidence-based treatment” (Maalouf et al. 2011). Adequate treatment is defined as optimal dosage, adherence, and duration; evidence-based treatments include pharmacotherapy, cognitive behavioral treatment (CBT), and interpersonal therapy (IPT).

History indicates that J. probably underwent adequate duration of pharmacotherapy, defined as 8–12 weeks with at least 20 mg fluoxetine or equivalents. J. had had at least two different antidepressants, and also had a trial of CBT. One could argue that a trial of a second SSRI might have been indicated before switch to a tricyclic, but J.'s symptoms were quite severe. However, tricyclic antidepressants have a poorer track record in treatment of pediatric depression, and are generally more poorly tolerated in terms of adverse effects. Another potential issue is that it is not entirely clear if J.'s adherence to her medication regimen had been ascertained, as adolescents may not always comply fully with recommended treatment. The risk of non- or poor adherence to psychotropic medication would also likely be higher for those with eating disorders, such as J., who might be particularly concerned about weight gain.

Past studies have suggested that only about 60% of adolescents will respond to initial treatment with a selective serotonin reuptake inhibitor (SSRI). Results of the TORDIA (Treatment of SSRI- Resistant Depression in Adolescents) trial in 334 adolescents age 12–28 years indicated that at the 12 week endpoint, cognitive behavioral treatment plus switch to another SSRI or to venlafaxine resulted in greater improvement in mood symptoms than medication switch alone, but there was no difference in response rate to either a second SSRI or venlafaxine (Brent et al. 2008). At 24 month follow up of this sample, 39% had achieved remission, but initial treatment group was not predictive. Remission rate was higher and achieved more quickly in those who had shown, among other factors, clinical response by week 12, and in those with less severe depression, hopelessness, anxiety, suicidal ideation, and family conflict (Emslie et al. 2010). By 72 weeks, although 61% had achieved remission, more than one third did not, and one quarter of those who had achieved remission had undergone a relapse (Vitiello et al. 2011).

Risk factors for TRD include high rates of psychiatric and medical comorbidity and family conflicts, which J. clearly manifests. Comorbidity between anorexia and depression is well established; it is possible that J.'s poor nutritional status, borderline anemia, and apparent lack of psychotherapy may also have been contributors (Maalouf et al. 2011).

Another very interesting question posed by the case is the indication for ECT in adolescents. Although ECT is an effective and generally safe treatment for adults, there is little data on its use in youth. The American Academy of Child and Adolescent Psychiatry's (AACAP) Practice Parameters recommended consideration of ECT when “there is a lack of response to two or more trials of pharmacotherapy, or when the severity of symptoms precludes waiting for a response to pharmacological treatment” (Ghaziuddin et al. 2004). State legal guidelines must be followed; some states have minimal age restrictions. Many states require consensus between at least two independent child and adolescent psychiatrists, which did take place with J. Baseline neurocognitive (memory) assessment is recommended prior to treatment.

AACAP recommends unilateral electrode placement except in cases in which there is severe symptomatology, such as refusal to eat or drink, severe suicidality or psychosis. Close post ECT monitoring is recommended. Given that there is no evidence to date that ECT prevents relapse in the future, it should be viewed as an acute treatment (Ghaziuddin et al. 2004).

Finally, another interesting and complex aspect of the case is the question of Axis II features, given the cutting and self-injurious behavior. J.'s description of her cutting behavior as “relief from emotional pain” is certainly suggestive of borderline personality features. That said, she is not yet 18 and is suffering from a major Axis I disorder, and thus it would be difficult to unequivocally suggest personality disorder features at this time with these considerations. On the other hand, 47% of the sample of adolescents in the TORDIA study reported a previous history of non-suicidal self-injury (NSSI) and suicide attempts (SA), and the 24 week study incidence was 11% for NSSI and 7% for SA, suggesting that NSSI is not rare in treatment resistant depression, and may be a significant predictor of future NSSI behavior and suicide attempts (Asarnow et al. 2011).