Introduction: Patients with autism spectrum disorders (ASD) and attention-deficit/hyperactivity disorder (ADHD) show more symptoms of ADHD. Since there are more adverse events caused by psychostimulants compared to non-psychostimulants, the use of a non-psychostimulant such as atomoxetine might prove more beneficial for younger patients and/or those with comorbid ADHD. Objective: The aim of this thesis is to determine the efficacy and tolerability of atomoxetine in ASD patients presenting with ADHD, by examining (a) differences in ADHD symptoms for participants receiving atomoxetine versus those receiving placebos, and (b) risk differences in adverse events between these participants. Methods: An electronic search of both PubMed.gov and ClinicalTrials.gov were conducted. To be deemed eligible, studies had to (a) be randomized, double-blind, placebo-controlled trials comparing atomoxetine with a placebo, (b) administer atomoxetine for at least 1 week, and (c) include data on either ADHD outcomes or adverse events. Effect sizes for ADHD outcomes were calculated using Cohen's d, whereas risk differences were calculated for adverse events. For each of these two meta-analyses, effect sizes were aggregated across studies using a random effects method. Results: Overall ADHD outcomes were better for participants who received atomoxetine than for participants who received placebo, =0.297. Participants who received atomoxetine also demonstrated better outcomes in terms of attention and hyperactivity-impulsivity symptoms, =0.345 and 0.393, respectively. Though there were more adverse events for patients taking atomoxetine than placebo, the results were not statistically significant. Discussion: This thesis extends the findings of previous meta-analyses of pharmacological treatments for ASD and ADHD, while addressing the concerns raised in the critique of existing meta-analyses presented in this thesis, e.g., limited studies, length of treatment weeks, and dichotomization of data. It provides evidence that atomoxetine improves ADHD symptoms, with an overall frequency of adverse events that did not sufficiently differ from placebo beyond chance.

Thesis Completion




Thesis Chair/Advisor

Cheng, Zixi


Bachelor of Science (B.S.)


Burnett School of Biomedical Sciences


Biomedical Sciences

Degree Program

Biomedical Sciences



Access Status

Open Access

Release Date


Included in

Pharmacology Commons