To Name :
To Email :
From Name :
From Email :
Comments :

Evidence-Based Reviews

Drug therapy algorithms target autism’s problem behaviors

Having an evidence-based game plan can help you manage maladaptive behaviors that are holding back children or adults with autism.

Vol. 2, No. 4 / April 2003

Aggression, irritability, repetitive phenomena, or hyperactivity can interfere with the education and treatment of patients with autism.1 To assist clinicians in selecting drug therapies to control these behaviors, we developed a medication strategy based on controlled studies and our experience in an autism specialty clinic. In this article, we:

  • offer an overview with two algorithms that show how to target maladaptive behaviors in patients of all ages with autism
  • and discuss the controlled clinical evidence behind this approach.

Targeting behaviors

A multimodal approach is used to manage autistic disorder. Although behavioral techniques are useful and may decrease the need for drug therapy, one or more medications are often required to target severe maladaptive behaviors. Although we recommend that you consider all interfering behaviors when selecting medications, it is important to:

  • initially focus on symptoms—such as self-injurious behaviors and aggression—that acutely affect the patient’s and caregiver’s safety
  • consider the patient as being pre- or postpubertal, as developmental level may affect medication response.


Source: Reprinted with permission from McDougle CJ, Posey DJ. Autistic and other pervasive developmental disorders. In: Martin A, Scahill L, et al (eds). Pediatric psychopharmacology: principles and practice. New York: Oxford University Press, 2002. Copyright © 2002 by Oxford University Press, Inc. Aggression or self-injury. When drug therapy is needed to control severe aggression or self-injurious behavior, we recommend a trial of an atypical antipsychotic for prepubertal patients (Figure 1). Alternately, you may consider a mood stabilizer such as lithium or divalproex for an adult patient with prominent symptoms (Figure 2). However, mood stabilizer use requires monitoring by venipuncture for potential blood cell or liver function abnormalities.

Atypical antipsychotics are associated with adverse effects including sedation, weight gain, extrapyramidal symptoms, tardive dyskinesia, and hyperprolactinemia. Therefore, you may need to monitor weight, lipid profile, glucose, liver function, and cardiac function when using this class of agent.

Repetitive phenomena can be addressed with selective serotonin reuptake inhibitors (SSRIs), although SSRIs may be activating in children and tend to be better tolerated in older adolescents and adults.

Anxiety. In children with anxiety, we suggest a trial of mirtazapine—because of its calming properties—before you try an SSRI. If both the SSRI and mirtazapine are ineffective, try an α2-adrenergic agonist. A baseline ECG is recommended before starting an α2-adrenergic agonist—particularly clonidine, which has been associated with rare cardiovascular events. Overall, these agents are well tolerated.

Hyperactivity, inattention. Anecdotal evidence indicates that stimulants may worsen symptoms of aggression, hyperactivity, and irritability in individuals with autism. Therefore, we suggest you start with an α2-adrenergic agonist when target symptoms include hyperactivity and inattention. Consider a stimulant trial if the α2-adrenergic agonist does not control the symptoms.

An atypical antipsychotic is generally recommended for autistic individuals with treatment-resistant hyperactivity and inattention, anxiety, and interfering repetitive behaviors.

Drug therapy. As summarized below, controlled evidence for behavior-based drug therapy of autism (Table.) suggests a role for antipsychotics, antidepressants, mood stabilizers, and psychostimulants in managing interfering behaviors.


Risperidone. Two controlled studies have reported the effect of risperidone on autism’s related symptoms.

McDougle et al conducted a 12-week, double-blind, placebo-controlled trial of risperidone, mean 2.9 mg/d, in adults with pervasive developmental disorders. 2 Repetitive behavior, aggression, anxiety, irritability, depression, and general behavioral symptoms decreased in 8 of 14 patients, as measured by Clinical Global Impressions (CGI) scale ratings of “much improved” or “very much improved.” Sixteen subjects who received placebo showed no response. Transient sedation was the most common adverse effect.

More recently, the National Institutes of Mental Health-sponsored Research Units on Pediatric Psychopharmacology (RUPP) Autism Network completed a double-blind, placebo-controlled study of risperidone, 0.5 to 3.5 mg/d (mean 1.8 mg/d), in 101 children and adolescents with autism.3 After 8 weeks, 69% of the risperidone group responded, compared with 12% of the placebo group. Response was defined as:

  • at least a 25% decrease in the Irritability subscale score of the Aberrant Behavior Checklist (ABC)
  • and CGI ratings of “much improved” or “very much improved.”

Risperidone was effective for treating aggression, agitation, hyperactivity, and repetitive behavior. Adverse effects included weight gain (mean 2.7 kg vs. 0.8 kg with placebo), increased appetite, sedation, dizziness, and hypersalivation.

Olanzapine. A 12-week, open-label study of olanzapine, mean 7.8 mg/d, was conducted in eight patients ages 5 to 42 diagnosed with a pervasive developmental disorder.4 Hyperactivity, social relatedness, self-injurious behavior, aggression, anxiety, and depression improved significantly in six of seven patients who completed the trial, as measured by a CGI classification of “much improved” or “very much improved.” Adverse effects included weight gain (mean 18.4 lbs) in six patients and sedation in three.

Malone et al conducted a 6-week, open-label comparison of olanzapine, mean 7.9 mg/d, versus haloperidol, mean 1.4 mg/d, in 12 children with autistic disorder. 5 Five of six subjects who received olanzapine responded, compared with three of six who received haloperidol, as determined by a CGI classification of “much improved” or “very much improved.” Anger/uncooperativeness and hyperactivity were reduced significantly only in the olanzapine group. Weight gain and sedation were the most common adverse effects with both medications. Weight gain was greater with olanzapine (mean 9 lbs) than with haloperidol (mean 3.2 lbs).

Quetiapine. Only one report addresses quetiapine in children and adolescents with autism.6 Two of six patients completed this 16-week, open-label trial of quetiapine, 100 to 350 mg/d (1.6 to 5.2 mg/kg/d). The group showed no statistically significant improvement in behavioral symptoms. Two patients achieved CGI ratings of “much improved” or “very much improved.” One subject dropped out after a possible seizure during treatment, and three others withdrew because of sedation or lack of response.

Ziprasidone. McDougle et al conducted a naturalistic, open-label study of ziprasidone, 20 to 120 mg/d (mean 59.2 mg/d), in 12 subjects ages 8 to 20. Nine were diagnosed with autism and three with a pervasive developmental disorder not otherwise specified (NOS). Six subjects responded across 14 weeks of treatment, as determined by ratings of “much improved” or “very much improved” on the CGI.7

Transient sedation was the most common adverse effect. Mean weight change was −5.83 lbs (range −35 to +6 lbs). Despite ziprasidone’s reported risk of causing abnormal cardiac rhythms, no cardiovascular adverse effects were reported.

Clozapine. Three case reports but no controlled studies have addressed clozapine in autistic disorder:

  • Zuddas et al used clozapine, 100 to 200 mg/d for 3 months, to treat three children ages 8 to 12. 8 Previous trials of haloperidol were ineffective, but the patients’ aggression, hyperactivity, negativism, and language and communication skills improved with clozapine, as measured by the Children’s Psychiatric Rating Scale (CPRS) and the Self-Injurious Behavior Questionnaire.
  • Chen et al used clozapine, 275 mg/d, to treat a 17-year-old male inpatient. 9 Aggression was significantly reduced across 15 days, as determined by a 21-question modified version of the CPRS. Side effects included mild constipation and excessive salivation.
  • Clozapine, 300 mg/d for 2 months, reduced aggression in an adult patient and was associated with progressive and marked improvement of aggression and social interaction, as measured by the CGI and the Visual Analog Scale (VAS). No agranulocytosis or extrapyramidal symptoms were observed across 5 years of therapy. 10

Clozapine’s adverse effect profile probably explains why the literature contains so little data regarding its use in patients with autism. This drug’s propensity to lower the seizure threshold discourages its use in a population predisposed to seizures. Autistic patients also have a low tolerance for frequent venipuncture and limited ability to communicate agranulocytosis symptoms.


Clomipramine—a tricyclic antidepressant with anti-obsessional properties—has shown some positive effects in treating autistic behaviors. Even so, this compound is prescribed infrequently to patients with autism because of low tolerability. In one recent double-blind, placebo-controlled, crossover study, 36 autistic patients ages 10 to 36 received:

  • clomipramine, 100 to 150 mg/d (mean 128 mg/d)
  • haloperidol, 1 to 1.5 mg/d (mean 1.3 mg/d)

or placebo for 7 weeks. Among patients who completed the trial, clomipramine and haloperidol were similarly effective in reducing overall autistic symptoms, irritability, and stereotypy. However, only 37.5% of those receiving clomipramine completed the study because of adverse effects, behavioral problems, and general lack of efficacy.11




Behaviors improved

Significant adverse effects


Risperidone 2

Aggression, irritability

Mild transient sedation

Conducted in adults

Risperidone 3

Aggression, irritability

Weight gain, increased appetite, sedation, tremor, hypersalivation

Largest controlled study to date in children with autism

Clomipramine 11

Irritability, stereotypy

Tachycardia, tremors, diaphoresis, insomnia, nausea

Conducted in children and adults

Fluvoxamine 12

Aggression, repetitive phenomena

Nausea, sedation

No published controlled pediatric data; unpublished pediatric data unfavorable

Clonidine 24

Hyperactivity, irritability, stereotypies, oppositional behavior, aggression, self-injury

Hypotension, sedation, decreased activity

Small number of subjects

Clonidine 25

Impulsivity, self-stimulation, hyperarousal

Sedation and fatigue

Small number of subjects

Methylphenidate 28

Hyperactivity, irritability

Social withdrawal and irritability

Adverse effects more common at 0.6 mg/kg/dose

• Double-blind, placebo-controlled studies

Fluvoxamine. Only one double-blind, placebo-controlled study has examined the use of an SSRI in autistic disorder.12 In this 12-week trial, 30 adults with autism received fluvoxamine, mean 276.7 mg/d, or placebo. Eight of 15 subjects in the fluvoxamine group were categorized as “much improved” or “very much improved” on the CGI.

Fluvoxamine was much more effective than placebo in reducing repetitive thoughts and behavior, aggression, and maladaptive behavior. Adverse effects included minimal sedation and nausea.

Unpublished data of McDougle et al suggest that autistic children and adolescents respond less favorably than adults to fluvoxamine. In a 12-week, double-blind, placebocontrolled trial in patients ages 5 to 18, only 1 of 18 subjects responded to fluvoxamine at a mean dosage of 106.9 mg/d. Agitation, increased aggression, hyperactivity, and increased repetitive behavior were reported.

Fluoxetine. Seven autistic patients ages 9 to 20 were treated with fluoxetine, 20 to 80 mg/d (mean 37.1 mg/d) in an 18-month longitudinal open trial. Irritability, lethargy, and stereotypy improved, as measured by the ABC. 13 Adverse effects included increased hyperactivity and transient appetite suppression.

In another open-label trial of 37 autistic children ages 2 to 7, response was rated by investigators as “excellent” in 11 and “good” in another 11 who received fluoxetine, 0.2 to 1.4 mg/kg/d. However, aggression was a frequent cause for drug discontinuation during the 21-month study. 14

Sertraline. In an open-label trial, Steingard et al treated nine autistic children ages 6 to 12 with sertraline, 25 to 50 mg/d for 2 to 8 weeks. Clinical improvement was observed in irritability, transition-associated anxiety, and need for sameness in eight of the children. 15


Source: Reprinted with permission from McDougle CJ, Posey DJ. Autistic and other pervasive developmental disorders. In: Martin A, Scahill L, et al (eds). Pediatric psychopharmacology: principles and practice. New York: Oxford University Press, 2002. Copyright © 2002 by Oxford University Press, Inc.In an open-label trial, 42 adults with pervasive developmental disorders were treated for 12 weeks with sertraline, 50 to 200 mg/d (mean 122 mg/d). Twenty-four (57%) achieved CGI ratings of “much improved” or “very much improved” in aggressive and repetitive behavior.16

Paroxetine. Fifteen adults with profound mental retardation (including seven with pervasive developmental disorders) received paroxetine, 10 to 50 mg/d (mean 35 mg/d) in a 4-month open-label trial. Frequency and severity of aggression were rated as improved after 1 month, but not at 4 months.17

Mirtazapine. In an open-label trial, 26 patients ages 3 to 23 with pervasive developmental disorders were treated with mirtazapine, 7.5 to 45 mg/d (mean 30.3 mg/d). Aggression, self-injury, irritability, hyperactivity, anxiety, depression, and insomnia decreased in 9 patients (34%), as determined by ratings of “much improved” or “very much improved” on the CGI. Adverse effects included increased appetite and transient sedation. 18

Mood stabilizers

Lithium. No recent studies of lithium in pervasive developmental disorders are known to exist, but several case reports have been published:

Did you miss this content?
An under-recognized epidemic of elder abuse needs your awareness and action