Use of atypical antipsychotics in individuals with anorexia nervosa

Practice Points:

1. Select atypical antipsychotics can be considered in individuals with anorexia nervosa (AN) and ongoing cognitive distortions or obsessive preoccupations regarding eating, weight gain, food, and/or body image that interfere with engagement in treatment and weight restoration.

2. Aripiprazole has growing evidence to support its use, with interest in the effects of D₂ partial agonism in targeting key features of AN. Aripiprazole has demonstrated benefit for eating disorder (ED) thoughts and is well-tolerated.

3. Olanzapine is the most studied atypical antipsychotic in those with AN. Evidence is mixed regarding the effects on weight and ED symptoms. High rates of side effects (eg, sedation, dyslipidemia, increased appetite, and hyperglycemia) contribute to high rates of medication discontinuation.

4. Risperidone and quetiapine have not demonstrated consistent benefit in the treatment of AN. Quetiapine is frequently associated with transient side effects, including sedation, difficulty concentrating, and orthostasis, and risperidone is associated with hyperprolactinemia.

Introduction

Anorexia nervosa (AN) is a life-threatening mental health condition characterized by severe cognitive distortions and obsessive preoccupations regarding eating, weight gain, food, and/or body image, which often prolong treatment and delay recovery.^1-4 While eating disorder (ED)-focused Family-Based Therapy (FBT) and weight restoration are mainstays of treatment among individuals with AN, interest exists in evaluating the role of atypical antipsychotics to target cognitive distortions and obsessive preoccupations.^5-8

While treatment guidelines do not provide clear recommendations regarding antipsychotic choice, the timing of initiation, or duration of treatment, atypical antipsychotics are among the most commonly prescribed psychotropic medications in individuals with AN.^2-4 Olanzapine, risperidone, quetiapine, and aripiprazole are described in the literature, with mixed efficacy and safety profiles.^{4 ,8-11} Through an illustrative case review, this manuscript will highlight practical and evidence-based pearls to apply to this important patient population.

Illustrative Case

“I don’t deserve to eat. I hate my body.”

CM is an 18-year-old cisgender female admitted to the hospital with concerns about abdominal discomfort, constipation, calorie restriction (600-800 cal/d), excessive exercise (3 hr/d), fear of gaining weight, and negative thoughts of self-image. CM reports significant food and calorie restriction that started 8 months ago, precipitated by social stressors, including online bullying and perceived enhancement of athletic performance as a competitive skier. She eats a small breakfast and dinner daily, identifying “carbs, dessert, and soda” as fear foods and “veggies and watermelon” as safe foods. Her last menstrual cycle was more than 4 months ago. Upon further assessment, you learn that she has lost 17 kg in the past 3 months. Her body mass index (BMI) is currently 17 kg/m² , and she was 20 kg/m² 3 months ago. Currently, she is at 55% of her ideal body weight (IBW) (41 kg) and in the 30%ile based on weight.

CM’s initial treatment is as follows: (1) monitor for electrolyte imbalances, fluid retention, dehydration, and refeeding syndrome; (2) a meal plan starting at 1750 kcal/day, increase by 250 kcal/day as tolerated; (3) give oral supplement at 100% of calories for uneaten foods; (4) supervise meals and snacks and provide meal support; (5) goal weight gain 0.2 kg/day; and (6) initiate supportive medications, including a daily multivitamin, thiamine, zinc, and polyethylene glycol.

“My eating disorder is telling me I’m disgusting.”

After acute medical stabilization, CM is transferred to the day treatment ED program for FBT and structured support with nutrition. Despite ongoing reports of intrusive, perseverative ED thoughts, CM’s weight trend is positive in the initial weeks of treatment. Notably, her BMI is 19 kg/m² and 70% of IBW (50 kg) and 45%ile based on weight. As treatment progresses, CM endorses symptoms of anhedonia, sleep disturbance, fatigue, restlessness, guilt, depressed mood, and anxiety, for which escitalopram 5 mg by mouth daily is initiated. At the time of antidepressant initiation, BMI is 21 kg/m² and 80% of IBW (55 kg) and 50%ile based on weight.

One month into ED treatment, CM reports ongoing body image distress, fear of weight gain, and self-injurious behavior related to recently learning her weight. Of note, food-avoidant behaviors have increased, with CM reporting a heightened difficulty eating fear foods, completing meals, and engaging in FBT. Escitalopram was increased weekly to a dose of 20 mg by mouth daily to target ongoing symptoms of depression and anxiety. CM denies side effects at this dose and does report some improvement in depressed mood. Overall, CM’s weight trend has been neutral, with a recent negative trend. Her BMI is 20 kg/m² and is 75% of IBW (52 kg) and 47%ile based on weight. Given concerns for heightened food-avoidant behavior, challenges engaging in FBT, and a negative weight trend, the treatment team decided to start aripiprazole 2.5 mg by mouth daily, targeting cognitive distortions and rigidity regarding food and body image.

Evidence-Based Discussion

Eating Disorder Guidelines

The goals of ED treatment include weight restoration, return to a healthy growth trajectory, normalizing eating behaviors, and establishing a positive relationship with food and body image.^{3 ,4} In select cases, psychotropic medications can be useful adjuncts to weight restoration and FBT.^{3 ,4} It is estimated that more than 50% of individuals with AN are prescribed psychotropic medications, with selective serotonin reuptake inhibitors (SSRIs) and atypical antipsychotics among those most prescribed.^{2 ,3,12} While SSRIs are often prescribed to target co-occurring symptoms of depression, anxiety, or obsessive-compulsive or trauma-based disorders, atypical antipsychotics are often prescribed to support weight gain and target cognitive distortions and obsessive preoccupations regarding eating, weight gain, food, and/or body image.^8-11

Despite high rates of atypical antipsychotic prescribing in clinical practice, ED treatment guidelines do not provide specific recommendations regarding their use.^{3 ,4} Instead, ED treatment teams must make decisions based on updated primary literature (Table), patient-specific factors, and formulary availability.

TABLE.

Available literature for atypical antipsychotic use in anorexia nervosa

Reference	Study Design	Study Population	Intervention	Results
Aripiprazole
Frank et al5 2016	Case series	N = 4 females with AN Mean age: 13 yr	Aripiprazole 2-5 mg/d Duration: 3-12 mo	Improved cognitive flexibility, reduced ED preoccupations surrounding weight gain Improvements noted within 1 week of initiation (self-report) Increase in weight Well tolerated
Frank et al6 2017	Retrospective, case-control study	N = 106 females with AN (n = 22 aripiprazole, n = 84 controls) Mean age: 14 yr	Aripiprazole 1-5 mg/d; mean dose 3.6 mg/d Duration: 40 d (mean)	Aripiprazole associated with greater increase in BMI (6% higher) and BMI percentile (20% higher) compared with control group Improvement in ED cognitions Well tolerated
Tahıllıoğlu et al13 2020	Retrospective study	N = 11 females with AN; n = 9 previously treated with an AAP 11-17 yr	Aripiprazole 2.5-15 mg/d Duration: 20-28 mo	Significant reductions in obsessive thoughts related to food/body image (CGI-S score improvement from 6.09 to 1.82) Improvement in BMI Well tolerated, compared with previous AAP trial
Trunko et al14 2011	Case series	N = 8 females with AN (n = 5) or BN (n = 3) 15-55 yr	Aripiprazole 5-15 mg/d Duration: 4-41 mo	Decrease in fear of eating, obsessive thoughts related to food and body image, cognitive rigidity (self-report and clinician observation) Increase in BMI and oral food intake Improvement in comorbid anxiety, depression (rating scales not described) Well tolerated
Olanzapinea
Attia et al15 2011	Double-blind, randomized controlled trial	N = 23 individuals with AN (n = 1 male); n = 11 olanzapine, n = 12 placebo Mean age: 27.7 yr	Olanzapine 2.5-10 mg/d; mean dose 8 mg/d Duration: 8 wk	Significant improvement in BMI No difference in ED cognitions (BSQ, EDI, PANSS, YBC-EDS) No difference in comorbid mood symptoms (BAI, BDI) Good tolerability (74% completed the study). Most common side effect was sedation
Attia et al16 2019	Double-blind, randomized controlled trial	N = 152 individuals with AN; n = 75 olanzapine, n = 77 placebo Mean age: 29 yr	Olanzapine 2.5-10 mg/d; mean dose 7.7 mg/d Duration: 16 wk	Significant improvement in BMI No difference in ED obsessive thoughts (YBOCS) or ED cognitions (EDE)
Barbarich et al17 2004	Uncontrolled, nonrandomized, open-label trial	N = 17 individuals with AN Mean age: 20.5 yr	Olanzapine 2.5-7.5 mg/d; mean dose 4.7 mg/d Duration: 6 wk	Significant increase in weight Improvement in some mood (BDI) and ED cognitions (YBC-EDS scores) No improvement in anxiety (STAI) and obsessions/compulsions (YBOCS)
Bissada et al18 2008	Double-blind, randomized controlled trial	N = 34 females with AN 23-30 yr	Olanzapine 2.5-10 mg/d Duration: 10 wk	Significant improvement in BMI and earlier achievement of target weight Decrease in ED thoughts (YBOCS)
Bosanac et al19 2003	Retrospective study	N = 14 females with AN 19-49 yr	Olanzapine mean dose 9.7 mg/d Duration: 25 d (mean)	Significant improvement in BMI
Brambilla et al20 2007	Double-blind, placebo-controlled, randomized controlled trial	N = 30 females with AN 24-26 yr	Olanzapine 2.5 mg/d for 1 mo, then 5 mg/d for 2 mo Duration: 3 mo	No significant difference in weight Small improvement in obsessive thoughts (YBOCS-ED)
Brambilla et al21 2007	Double-blind, randomized controlled trial	N = 20 patients with AN Mean age: 23 yr	Olanzapine 2.5 mg/d for 1 mo, then 5 mg/d for 2 mo Duration: 3 mo	No significant difference in weight
Himmerich et al22 2017	Retrospective study	N = 12 individuals with AN 18-60 yr	Olanzapine dose not reported Duration: 8 wk	Increase in weight (4.6 kg) Increase in GGT among those taking olanzapine
Kafantaris et al23 2011	Double-blind, placebo-controlled pilot study	N = 20 females with AN 12-22 yr; median age 17 yr	Olanzapine 8.5 mg/d (mean) Duration: 10 wk	No difference in percent median body weight No change in eating attitudes and behaviors (EDE), obsessive thoughts (YBC-EDS), and resting energy expenditure Trend in increased fasting glucose and insulin (week 10) Sedation and fear of weight gain common reasons for olanzapine refusal/drop out
Leggero et al24 2010	Prospective study	N = 13 females with AN-R 10-16 yr; mean age: 13.7 yr	Olanzapine 3.75-12.5 mg/d; mean dose 4.1 mg/d Duration: 6 mo	Significant improvement in BMI Significant improvement in global functioning (CGAS), eating attitudes (EAT-26), anxious-depressive symptoms (CBCL), and hyperactivity (SIAB) 54% considered responders at 6 mo (EAT-26) Well-tolerated overall. Increase in liver enzymes (n = 2) that resolved with an olanzapine dose reduction
Malina et al25 2003	Retrospective study	N = 18 individuals with AN Mean age: 22 yr	Olanzapine 4.7 mg/d (mean) Duration: 17 weeks	Based on a retrospective questionnaire (nonstandardized instrument), self-reported reductions in obsessive thoughts and about body image and fear of being fat, and reduced anxiety during meals
Marzola et al11 2015	Retrospective chart review	N = 75 individuals with AN • n = 23 aripiprazole • n = 27 olanzapine • n = 25 no atypical 25 yr (mean)	Olanzapine 6.1 mg/d (mean) Aripiprazole 9.13 mg/d (mean) Duration: 5 wk (mean)	All groups demonstrated improved food obsessions, compulsions, preoccupations, and rituals (YBC-EDS); significant among aripiprazole group (P = .005) Aripiprazole more effective in reducing eating-related preoccupations and rituals (YBC-EDS), P = .017. Aripiprazole more effective in reducing purging episodes compared with olanzapine, P = .09 No difference in BMI, anxiety (HAM-A), or excessive exercise among treatment groups
Mondraty et al26 2005	Randomized, controlled, open-label study	N = 15 individuals with AN • n = 8 olanzapine • n = 7 chlor-promazine 25 yr (mean)	Olanzapine 5-15 mg/d; mean dose 10 mg/d Chlor-promazine 25-100 mg/d; mean dose 50 mg/d Duration: 46-53 d	No significant difference in weight in either group Small improvement in ED cognitions and ruminations in olanzapine group (EDI-2) Sedation most common side effect in both groups
Norris et al27 2011	Retrospective, matched-groups comparison study	N = 86 females with AN or EDNOS • n = 43 olanzapine • n = 43 controls 10-17 yr	Olanzapine 5 mg/d (mean) 50% on concurrent SSRI or SNRI Duration: total treatment duration unclear	No difference in weight gain Effect of olanzapine on ED cognitions could not be assessed because of confounders (ie, greater illness severity, acuity, and psychiatric comorbidities) Among those treated with olanzapine, 56% had at least one side effect with olanzapine. Sedation was the most common (40%), followed by lipid abnormalities (29%)
Powers et al28 2002	Uncontrolled, nonrandomized, open-label trial	N = 18 individuals with AN (n = 2 males) 14-56 yr; mean age: 27 yr	Olanzapine 10 mg/d Duration: 10 wk	Most patients gained weight (78%), and 17% achieved their IBW Sedation was the most common adverse effect (65%)
Pruccoli et al29 2022	Observational, naturalistic, case-control study	N = 118 individuals with AN • n = 37 olanzapine ≤ 5 mg/d • n = 29 olanzapine > 5 mg/d • n = 52 controls 15 yr (mean)	Olanzapine low dose ≤ 5 mg/d or full dose >5 mg/d Duration: 5 mo	Greater improvement in depressive symptoms with low-dose olanzapine (BDI-II) Low-dose olanzapine better tolerated, with a longer duration of treatment
Spettigue et al30 2018	Non-randomized, open-label, study	N = 32 individuals with AN • n = 14 olanzapine • n = 18 no antipsychotic 11-17 yr; mean age: 15.5 yr	Olanzapine 5.3 mg/d (mean) Duration: 12 wk	Olanzapine associated with significant increase in weight (P = .012) No difference in mood, anxiety, ED cognitions/behaviors (EDEQ-A, EDI-3) 32% experienced asymptomatic increase in serum prolactin with olanzapine (as early as week 2); 32% discontinued olanzapine, as recommended by the treating physician, because of AEs (ie, dyslipidemia, hyperprolactinemia, elevated liver functions tests)
Quetiapine
Bosanac et al31 2007	Open-label clinical trial	N = 7 females with AN Mean age: 33.3 yr	Quetiapine 50-800 mg/d; mean dose 522 mg/d Duration: 8 wk	Small improvement in BMI Improvement in restrictive eating (EDE restraint subscale score) No difference in ED thoughts (EDE-12) No difference in mood or obsessive thoughts (MADRS, YBOCS) Dose-related sedation observed
Court et al32 2010	Naturalistic, open-label, randomized-controlled trial	N = 33 individuals with AN (n = 1 male) Mean age: 22 yr	Quetiapine 50-400 mg/d; mean dose 322.5 mg/d Duration: 12 wk, with 52-wk open-label follow-up	No significant difference in ED thoughts (EDI-2) Most common side effects included sedation, fatigue, difficulty concentrating, and orthostatic dizziness
Powers et al33 2007	Open-label pilot study	N = 19 individuals with AN (n = 1 male) Mean age: 26.8 yr	Quetiapine 150-300 mg/d Duration: 10 wk	Significant decrease on total mean PANSS scores (56.2 to 48.4, P = .24). No difference in PANSS positive or negative sub scales Significant decrease in EDI-2 scores No difference in total YBOCS or HAM-D Small change in weight (mean increase 0.73 kg) Most common AEs: dizziness, fatigue, constipation, and dry mouth
Powers et al34 2012	Double-blind, randomized, placebo-controlled clinical trial	N = 21 individuals with AN (n = 1 male) Mean age: 34 yr	Quetiapine 178 mg/d (mean) Duration: 8 wk	No difference in hours occupied by ED preoccupations and rituals (YBC-EDS). No difference in ED thoughts (EDI-2) No difference in BMI
Risperidoneb
Hagman et al35 2011	Double-blind, placebo-controlled study	N = 40 females with AN Mean age: 16 yr	Risperidone 0.5-4 mg/d titrated weekly to mean dose of 2.5 mg/d Duration: 9 wk (mean)	No significant differences in drive for thinness (week 11), body dissatisfaction (week 7), and body image distortion (EDI-2, week 7) No difference in time to reach 90% of IBW Risperidone associated with significant elevations in serum prolactin at week 7 (P = .001)
Pruccoli et al36 2023	Naturalistic, placebo-controlled study	N = 120 individuals with AN (n = 42 risperidone) Mean age: 16 yr	Mean starting dose 0.36 mg/d; mean maximum dose 0.86 mg/d Mean duration: 117 d (range: 3-365 d)	No difference in BMI or ED symptoms (EDI-3) No difference in mood (BDI-II) Retention rate of 50% at 3 mo and 9.5% at 1 yr among risperidone users No difference in rehospitalization rate at 1 yr

AAP = atypical antipsychotic; AEs = adverse effects; AN = anorexia nervosa; AN-R = anorexia nervosa restricting type; BAI = Beck’s Anxiety Inventory; BDI = Beck’s Depression Inventory; BDI-II = Beck’s Depression Inventory-II; BMI = body mass index; BN = bulimia nervosa; BSQ = Body Shape Questionnaire; CBCL = Child Behavior Checklist; CGAS = Children’s Global Assessment Scale; CGI-S = Clinical Global Impressions Scale; EAT-26 = Eating Attitudes Test-26; ED = eating disorder; EDEQ-A = Eating Disorder Examination Questionnaire–Adolescent; EDI = Eating Disorders Inventory; EDI-2 = Eating Disorders Inventory-2; EDI-3 = Eating Disorders Inventory-3; ED-NOS = eating disorder not otherwise specified; EDE = Eating Disorder Examination; EDE-12 = Eating Disorder Examination-12; GGT = gamma-glutamyl transpeptidase; HAM-A = Hamilton Anxiety Rating Scale; HAM-D = Hamilton Depression Rating Scale; IBW = ideal body weight; MADRS = Montgomery-Åsberg Depression Rating Scale; PANSS = Positive and Negative Syndrome Scale; SIAB = Structured Inventory for Anorexic and Bulimic Syndromes; SNRI = serotonin-norepinephrine reuptake inhibitor; SSRI = selective serotonin reuptake inhibitors; STAI = Spielberger State-Trait Anxiety Inventory; YBC-EDS = Yale-Brown-Cornell Eating Disorder Scale; YBOCS = Yale-Brown-Cornell Obsessive Compulsive Scale.

^aCase reports not included, given availability of controlled studies.

^bCase reports including < 3 individuals were not included.

Aripiprazole

Aripiprazole has gained interest in the treatment of individuals with AN given its favorable side effect profile, positive effects on ED cognitions (Table), and unique mechanism of action.^{7 ,37} As a D₂ partial agonist, dopamine receptor downregulation is thought to contribute to reward system responsiveness, habituation to refeeding, decreased conditioned fear response, and reduced cognitive rigidity and body image distortion.^{5 ,13,37} While studies are limited by small sample sizes and lack of rigorous study methodology, aripiprazole has demonstrated improvements in ED cognitions and weight gain in 3 case series^{5 ,13,14} and 1 retrospective case-control study.³⁸ Among a group of 9 patients who were previously treated with an antipsychotic, aripiprazole was associated with an increase in medication compliance, improved medication tolerability, and improved ED outcomes.¹³ Reasons for prior antipsychotic discontinuation included increased appetite with olanzapine and galactorrhea, sedation, or increased appetite with risperidone.¹³ With studies reporting good tolerability and low rates of adverse effects (eg, sedation, nausea), low-dose aripiprazole (≤5 mg/d) is a leading choice in clinical practice.

Olanzapine

Olanzapine is the most well-studied antipsychotic among individuals with AN (Table) and is frequently used in clinical practice. Among 7 double-blind, randomized, controlled trials,^{15 ,16,18,20,21,23,26} olanzapine was associated with mixed results for weight gain and ED cognitions. Increases in fasting blood glucose and sedation were common reasons for treatment discontinuation.^{23 ,26} Ten open-label or retrospective studies^{11 ,17,19,22,24,25,27-30} demonstrated variable improvements in weight but with mixed effects on ED symptoms. Of note, high rates of side effects were reported with sedation, dyslipidemia, hyperprolactinemia, and elevations in liver function tests (LFTs) among the most common reasons for treatment discontinuation. In 1 retrospective study, aripiprazole was associated with greater improvements in ED symptoms and tolerability compared to olanzapine.¹¹ Altogether, the use of olanzapine should be considered carefully, given the mixed results and the risk of adverse effects.

Quetiapine

Overall, quetiapine has not demonstrated significant improvements in ED cognitions, weight gain, or mood (Table).^{33 ,34} While 1 open-label study demonstrated some improvements in weight and ED symptoms,³¹ subsequent studies have not demonstrated benefit.^32-34 In the only double-blind, placebo-controlled study,³⁴ no difference in ED symptoms or BMI compared with placebo was observed. Sedation, orthostasis, constipation, dry mouth, and difficulty concentrating were noted at the beginning of treatment but were generally transient.^31-33 While the use of quetiapine could be considered in this patient population to support concurrent symptoms (eg, bipolar disorders), quetiapine should not be routinely considered for AN-associated cognitions.

Risperidone

Two studies suggest a lack of benefit for risperidone in those with AN, with no improvement in ED cognitions and weight gain (Table).^{35 ,36} In a double-blind, placebo-controlled study,³⁵ risperidone did not demonstrate improvements in core ED symptoms or time to reach 90% of IBW. While overall well-tolerated, it is notable that risperidone was associated with a significant elevation in serum prolactin by week 7 of treatment. Additionally, a naturalistic, placebo-controlled study³⁶ did not demonstrate a difference in BMI, ED symptoms, mood, or rehospitalization rates among those prescribed low-dose risperidone. Overall, risperidone is an undesirable choice, given the lack of robust literature and current data, suggesting a lack of benefit in individuals with AN.

Recommended Strategies for Appropriate Medication Use in ED

Antipsychotics may be used to augment treatment in patients with severe ED cognitive symptoms, thought distortions, or obsessive preoccupations that interfere with engagement in treatment. Exact timing of initiation is not well-understood but likely can be considered early in weight restoration.

Olanzapine is the most widely studied antipsychotic, but clinicians should consider the benefits versus risks, including sedation, dyslipidemia, hyperprolactinemia, and elevations in LFTs. Aripiprazole has growing evidence, particularly in adolescents. Low doses should be used to optimize D₂ partial agonism effects. Based on identified improvements in ED cognitions, aripiprazole could be considered a reasonable alternative to olanzapine with lower rates of side effects. Quetiapine and risperidone should be avoided for ED symptoms and weight gain.

Routine monitoring should include evaluation for extrapyramidal side effects and tardive dyskinesia, lipid panels, fasting blood glucose/HgbA1c, waist circumference, and an electrocardiogram in some high-risk cases. Metabolic monitoring should occur at baseline (before antipsychotic initiation) and then at 3, 6, and 12 months. Weight and waist circumference should be measured at each office visit.^39-41 An Abnormal Involuntary Movement Scale (AIMS) should be completed every 6 months.³⁹

Consider short-term use (eg, 8-12 weeks), with ongoing evaluation for need as patient continues to engage in other evidence-based interventions (eg, FBT, weight restoration). The duration of use may also be influenced by other comorbidities (eg, mood disorder).

Conclusion

Atypical antipsychotics can be considered in some patients with AN, and difficult-to-treat ED cognitions or thought distortions that may interfere with participation in FBT and/or weight restoration. Olanzapine is well-studied, and there is growing literature to support the use of aripiprazole. Once an atypical antipsychotic is started, its ongoing need should be reevaluated with consideration to discontinue within 3 months of weight restoration. Routine monitoring for side effects is important to ensure safe use (eg, AIMS, metabolic monitoring).